2023 College and University Educators Workshop

FASKIANOS: We are going to get started now for the final session, so if everybody could come in. Again, I want to thank you all for coming to the workshop. We have one final session and then the reception afterwards. It’s been great to have all of you here. We are going to send out a survey that I hope you’ll give us feedback on what we can—what we did well, what we can do better, any other suggestions, how we can help support the important work you’re doing.

As you saw in the program, CFR Academic. We offer webinars for students. I know many of you participated. We have a series for you all. Additionally, we do go to professional gatherings to talk about CFR resources and foreign affairs. So if you are connected—I just spoke to Loren [Henderson] about her association. So if you have a connection, we are trying to get the word out and be a resource. We would love to come to your associations to do presentations at the conferences, and we can also help set up, travel allowing, et cetera, a fellow to come speak for that kind of thing.

Additionally, for all of our fellows’ books we’ve got teaching notes. Richard mentioned for his book we have teaching notes. So they’re online. And we have a whole page of offerings on the website and other things with briefers, backgrounders, and things—all different kinds of content. Print, video, timelines, infographics, all sorts of formats for the kind of students. There are many kinds of learnings, different formats. So you can plug and play.

So with that, I’m going to invite our final panel onto the stage. And Zaid Zaid is going to moderate this discussion. Very excited to have him, of Cloudflare. And then we can gather in the reception room for a glass of wine. So enjoy the last hour, and look forward to talking to you after this. Thank you. (Applause.)

ZAID: Hi. Good afternoon. So my is Zaid Zaid. I’m the head of U.S. public policy at Cloudflare. And I’m here with Lauren Kahn, who’s a research fellow here at CFR, focusing on defense innovation and impact—and the impact of emerging technology in international security. And also with Marc Rotenberg, who’s the president and founder of the Center for AI and Digital Policy.

So we’re going to—I want to drill down a little bit on the emerging technology, because it’s a broad topic. I know we’re going to talk about AI specifically a little bit in more detail. But can we start, Lauren, maybe with an overview of, like, just what are the different—what are the different topics in emerging AI? Maybe the top three to five different things in emerging technology—I’m sorry. I keep saying emerging AI. Emerging technology.

KAHN: Sure, absolutely. So I focus on a security lens, so I’ll talk about it in that perspective, I think. Emerging technologies, I would say, are—you know, it’s a catch-all term. We’re hearing a lot of terms about emerging technology, edge technologies, things like that. I usually like to think of it as the cohort of technologies we see now that are more nascent. But again, from things like artificial intelligence, quantum computing, things that are a little bit further down the line. But I also like to kind of categorize, at least from the defense and emerging technologies perspective, I think it’s a little bit helpful to ground some technologies that have been around for a while as well, that are either being used in new ways or reaching maturity.

So I kind of say there’s, like, maybe three buckets, right, where we have technologies that have been longstanding but are reaching maturity, and are using—are used more proliferated in different ways. Like, things like drones that we’ve been familiar with but, again, we’re seeing used in new ways on the battlefield, like in Ukraine and in Russia. And then I’ll say there is another category of technologies that are a little bit newer to the foray but, again, are not super new—things like loitering munitions, you know, that have increased degrees of autonomy but are not as familiar as drones, and things, again, like AI and quantum that are a little bit more nascent, we’re just starting to see the first kind of implications and the first applications of them in the battlefield and in international security contexts.

ZAID: And, Marc, what about you? What would you add?

ROTENBERG: Well, our focus is on artificial intelligence. But I would say about this topic, it turns out to be crosscutting because we think about AI in terms of national security and defense policy. We think about AI in terms of consumer products and ChatGPT. And we think of AI as an enabler across many sectors of the economy, from environmental protection, to medical innovation. So in that respect, it’s really quite interesting how many different domains AI touches.

ZAID: Great. So, Marc, want to start with you on a question about bias in AI. It’s not new. There have been bias in AI since the beginning. But it doesn’t seem to be getting that much better. What can we do to ensure that—as we go forward, that we take—that people are conscious of it, and that we try to take some of the bias out of the tools?

ROTENBERG: Well, I think the interesting challenge that AI poses for the bias problem and the protection of civil rights law or equal employment law is that increasingly decisions that are being made in automated systems are opaque. So traditionally we could look at a business, whether it was a bank making a determination about a loan or a realtor deciding which homes to show to a couple, and say: Well, there’s some paper records, there’s some business practice, at least something we can look at and try to determine if factors were being used that were inadmissible, like race, like gender, like nationality.

Today, we’re in an era of machine learning. And this is a lot of big data, statistical inference, that generates outputs that even the people who design the systems can’t tell you precisely why this person got the loan and someone else didn’t. And so there’s a lot of discussion about how to improve the transparency of AI decision making. But frankly, I think there’s a real sense in this moment that that problem hasn’t been solved. So if you said to me, how do we take out bias in a 1980s expert system that’s making loan decisions, I could say, well, just open the box and look at the factors and figure out of those are permissible factors. But using a machine learning system today, I actually don’t think we have a good answer.

ZAID: Do you have anything to add to that?

KAHN: Yeah. I agree entirely. I also agree that it’s—I don’t want to get too scary, but I think it’s even more of a bigger problem than that. Where it’s not necessarily just that the question here is about bias by itself, but about fair algorithms. And is bias enough to kind of make sure that you have fair algorithms? I don’t think that—I think that’s one piece of the puzzle, but even if we did solve bias it’s not solving the whole problem of making sure that algorithms, again, are transparent, that they’re explainable, and that we can kind of see the decisions that they’re making, and are making the decisions in the best interests of those using those systems.

ZAID: Great. So, Lauren, given your expertise in military innovation, in national security, can you tell us a little bit about how AI has been used in the war with Ukraine, and whether or not this is sort of the way of the future in terms of AI and military conflict.

KAHN: Absolutely. I would say, honestly, before then my work was very largely hypothetical, where there was perfect examples, I think. And I think the field is changing very rapidly. And so I think before it was—there was a lot of programs of record research, and kind of about how AI might be used in these kinds of systems. There are things like that DARPA put out, the Defense Research Agency put out about AlphaDogfight, about having AI-driven pilots that are able to kind of do simulations, and things like that. And that was largely kind of where the discussion ended. There’s been the Project Maven in the past as well, about using AI algorithms to look through and sift drone footage, for example, were really the two examples that people liked to express.

I think that’s totally been blown open now, and I think a lot more understanding of the implications and uses of AI that are commercial as well, or beyond, you know, the specific input on military systems. I think, for example, they’re using natural language technology to look at and translate radio transmissions in real time. They’re using facial recognition technology to identify combatants and those injured. There’s—and those are public technologies, like Clearview is the facial recognition technology. Palantir is greatly involved in creating a sort of glass battlefield about bringing in all sources of information so that those on the ground know what’s happening. A lot of combination with satellite data.

So we’re seeing a lot of these more emphasis on data, and bringing in things together that I think is novel and, again, wasn’t really seen beforehand. And so that states now are kind of watching and seeing how this can move forward. There’s been a lot of investment in AI, but again it’s been very general. I think so now that states are kind of focused and zeroing in, and really seeing how this capacity and these technologies can really help improve and generate efficiencies for forces.

ZAID: So if you’re thinking about some of the stuff that was only theoretical before the war, like, and—I mean, let’s—there’s been talk, a lot of talk, about what happens in Ukraine may impact how things happen in Taiwan—between Taiwan and China. So what are you looking at theoretically that you would say might apply to China and Taiwan that we’re not thinking about now?

KAHN: Yeah. Honestly, I think a lot of the discussion about the pacing challenge with China has become a lot about technological competition. And that has been further reduced, I think, to a lot of the discussion about the AI competition. Again, there’s been a lot of investment both from the United States perspective and the Chinese perspective. And a lot of individuals worrying what that means. A lot of people need to compare publications in AI field, and how much they’re spending, and things like that. And have, you know, touted the ability of China to kind of move more quickly.

I will be a little bit more hesitant on that. I mean, there’s been statements put out by the PLA about having an intelligent ties military that can use AI, and use cyber, and use amphibious assault all together in some miraculous way to potentially launch an assault on Taiwan. And while that’s all well and good, it is very challenging. I think from a U.S. defense perspective in particular, we’ve seen a lot of organizational and bureaucratic challenges, I think, when it comes to software in particular, and these technologies that are not necessarily weapons, but also commercial. So how does that regulate?

So I think just because—I don’t think that China’s immune to those challenges as well. And so I think that they’re watching. And I think that the United States is watching and learning about how these can be implemented. But I don’t think it’s a done deal, necessarily, yet.

ZAID: Great. Thanks.

Marc, I know you have looked at regulations in dozens of countries and how they’re regulating AI. Can you tell us a little bit about where sort of the state of regulation of AI is, where the United States is, where we need to go, what worries you?

ROTENBERG: Great. Well, first of all, thank you for the question. Secondly, warning, advertisement ahead. (Laughter.) We’re going to hand out for you this postcard, which has a QR code for our report, titled Artificial Intelligence and Democratic Values Index. I believe it’s the most comprehensive report of its type ever published. We looked at seventy-five countries in detail to evaluate their AI policies and practices. We have a twelve-factor metric to evaluate and score countries, and then to rank them. And we’re trying to identify elements in AI policies and practices that, frankly speaking, favor openness, pluralism, respect for rule of law, respect for fundamental rights.

I thought Richard’s point on the last panel was actually very important, in one sense. You don’t put everyone in two camps, and then expect to be able to negotiate with the people that you’ve put in the opposing camp. But in the AI realm, if you break things down a little bit, you actually get some interesting results. I mean, Lauren mentioned, for example, the importance of fairness and transparency, and accountability, by the way. So those are themes that you will find in national AI policies. And when we find them, and where we see them being implemented in law, we say that’s good. And interestingly, it’s not just the nation’s—you know, the so-called democratic nations that are trying to establish these principles. Because in the AI realm, it turns out that you actually need fairness, and accountability, and transparency to create systems that are human-centered, and reliable, and trustworthy.

So you can almost embed into your technical design certain political values that turn out, we would say, to be favorable. Now, there are also contrary applications. Clearview is probably one of the most controversial deployments of AI technology because it does involve mass surveillance. But to your question, Zaid, it’s very interesting to see not only in the proposed Artificial Intelligence Act of the European Union a prohibition on the use of AI for mass surveillance, you’ll find that also in the UNESCO recommendation on AI ethics adopted in 2021, and endorsed by 193 countries, including, by the way, China and Russia. So when you begin to drill down on these policies and practices, you begin to ask questions, which are most favorable for democratic societies, and which global frameworks could be most useful, you’ve got some very interesting results.

And that’s, by the way, is here. (Laughter.) So pick up the card. I’ve got another card too, but he hasn’t asked me that question yet. (Laughter.) So we’ll get to that card in a moment.

ZAID: Well, since you brought up Clear, I want to ask you about—the last time I took a flight, went to London, I didn’t show a human my passport at all. Even when I got to the gate, they scanned my face and they were like, OK, you can go and get on the plane. Didn’t look at my ticket, didn’t look at my passport. So is that technology the same technology that worries—that is controversial, I should say, for Clear, or?

ROTENBERG: Well, we would say that’s a context-dependent form of authentication, which you can still debate. And there are still people who would like the option to be able to opt out. As people wanted the option, by the way, to opt out of the whole-body scanners. Originally in U.S. airports, by the way, those scanners revealed the entire contours of the naked body of the passenger. I mean, they were terribly invasive. And I was actually involved the lawsuit to get those scanners taken out of U.S. airports. So the current technique for screening is satisfying the security goal and minimizing the privacy intrusion.

But to your point, I mean, it’s one thing to say, before a person boards an airplane we need to be sure we know who this person is. It’s quite another thing to say, if we have a hundred thousand people attending a rally in Washington, DC, standing on the mall, that we have the technical ability and will, in fact, identify every one of those hundred thousand people, and maybe, by the way, assign a score to them, an adverse score, because they’re participating in a protest against the national government. And this is precisely what some countries are already beginning to do.

KAHN: Can I just jump in? I think that—no, I think that touches on a great point about these technologies that I think are—that I think, you know, it talks about emerging technologies in general that’s so hard about them. Is that because they are general purpose, because they touch on so many different facets of life, the technology itself isn’t that different but the application matters greatly. And I think that makes it very challenging for regulation. But also, at the same time, makes why—why I focused on the military aspect. Actually, by narrowing it down has actually helped, and I think has brought more individuals and more consensus on what to do about some of these applications.

For example, the United States just published its—the State Department released a public statement on the responsible use of AI and autonomy in military systems and is hoping to get other states to sign on. And it seems those are already pretty initial agreement. And the United States, France, and the United Kingdom has already agreed to one of those kind of statements in prior contexts, about not having a machine learning algorithm, for example, controlling nuclear weapons decisions. So I think there’s already some established kind of protocol and understanding of what’s acceptable and what’s not. But again, it’s highly, I think, context dependent, and will make this challenging going forward as these technologies kind of evolve and we see new applications and new kind of contexts that it can be introduced into.

ZAID: So, Lauren, a year ago you wrote in Foreign Affairs that Washington’s biggest concern should be that it’s moving too slowly on integrating AI into our military. So a year later, what do you think about that? Do you think they are moving fast enough? Is there more work to do? How would you rate things? If you writing that same article again now, what would you say?

KAHN: I would probably say the same thing. I think—I think there’s been a big shift, I think especially since the war in particular, about kind of that there needs to be revamp. I think I’ve seen a lot of progress, for example, again, the political statement. There’s been an update to the autonomous weapons directive, which controls autonomy in weapons systems. There’s been efforts to reorganize the entire Defense Department’s kind of approach to AI in general. Before it was very siloed. You know, they didn’t have the left hand speaking to the right, especially about a technology that is so dependent on data and things like that, that wasn’t all sitting together in the Defense Department.

I think now that’s been redone with the Chief Digital AI Office. All these, like, we’re a jargon soup. But they’re now in charge, and have brought all of that together and have been prioritizing AI education, and have been focusing on getting the data and the machine learning algorithms in the same place, so to speak, and to start discussing and interacting more with stakeholders in Silicon Valley and the private sector as well, and those actually developing these technologies. So I’m hopeful. I think it’s moving in the right direction. Whether it’s quick enough, quote/unquote, remains to be seen. I guess we’ll see in the next few years. But I am hopeful. I am positive about the direction.

ZAID: Marc, so I admit I did not get through all 1,266 pages of your report on democracy and AI.

ROTENBERG: Neither did I, by the way. (Laughter.) We have a big research team. (Laughter.)

ZAID: So we talked a little bit about regulation, but that question is different from democracy. So tell me about—tell us about what AI means for democracy, for voting, for authoritarian control, for everything. You’ve touched on some of those things already, but what would you say?

ROTENBERG: That’s actually a very difficult question to answer. When I started this report, which was a big empirical study, I had first done a reference book looking at all the AI policies I could find. And then someone told me, there are now about eight hundred AI ethics policies. I said, I don’t have time for that. So that’s when I started doing empirical work. (Laughs.)

I was genuinely interested in how we measure AI’s impact on democracy, but I was also interested in how we get to AI policy. So in our metrics, for example, we ask questions like: Are there opportunities for meaningful public participation in the development of a national AI strategy? That’s a pure process question. It doesn’t tell you what the outcome is going to be. Related question: Are the country’s AI policies readily available to the public? Because we simply assume that if people—I mean, I think we should assume that if people are going to meaningfully engage, they will need access to the relevant material, so that they can participate in that conversation.

So there were process-based dimensions to our evaluation. And again, it produces some surprising results. So we haven’t yet got into the generative AI discussion yet. That’s ChatGPT. But the Chinese government announced last month a proposed regulation on generative AI. And they have given the public till May 10 to comment. And we have a page devoted to the public voice. And if you would like to comment on that proposed regulation, the government has given you the opportunity to do so. Now, is it a meaningful opportunity? Well, I guess we would need some access to what people say and what’s ultimately decided.

So I’m not giving you the—the simple, this is what AI will do to elections, for example. Which I think is a really fraught problem. I mean, there’s no doubt that AI enables radicalization at a scale that is also highly personalized. And it has people genuinely concerned, because we’re not just talking now about posting an ISIS video on YouTube and sending it to your buddy and saying this is something you got to look at. We’re talking about the ability to do sophisticated, interactive, real-time online communication for almost no cost to thousands of people. That is actually enabled by ChatGPT.

Did I scare anybody, by the way? We’re having, like, a little discussion back and forth. We’re not sure if we’re supposed to scare you or reassure you. But that was like my—that was, like, my scare move. So did it work at all? A little bit? Just—I’ve got more—I’ve got more material, by the way. (Laughter.) Because if that didn’t work, we’ve got other moves. But, yeah, I mean, you can radicalize a lot of people very quickly and much more effectively than was previously possible.

ZAID: So you mentioned—(laughter)—

ROTENBERG: See? I did get you. (Laughter.)

ZAID: You mentioned an opportunity for the population to weigh in on policies. And I think in your writing you also talked about an opportunity for oversight. What does AI oversight look like?

ROTENBERG: Another great question. So we’ve been engaged in a lot of the evolving global policy frameworks. I was actually very fortunate to get on in the early days, which was, like, five years ago. We did something called the Universal Guidelines for AI, which received a lot of support. I mean, we had ACM, IEEE, these are all scientific, computing societies, AAAS, backing this. And there are twelve pretty good principles. And then we kind of pushed them along. And the OECD picks them up, and the G20 picked them up, and some of it gets its way into the EU AI Act, and the Council of Europe AI Convention.

But to your point, these are still kind of principles and frameworks. And they actually don’t answer the question of oversight. Who’s actually going to be responsible for doing those impact assessments? Who’s actually going to say to a company: Your product is not safe. We don’t want you to make it available in the marketplace? That turns out to be a very challenging lift for governments.

I think the EU will come up with a pretty good answer in the AI Act. I don’t think the U.S. yet has a good answer, but you did see in the last couple of weeks four of the top enforcement agencies—so, this is FTC—I apologize, by the way, I’m from Washington. We talk this way. FTC, EEOC, CFPB, and SEC all planning to step up their enforcement game in the AI realm. But it’s not obvious, by the way, that any of them actually have clear authority on an issue like generative AI.

ZAID: So, just to break up—or, to expand a little bit some of your acronyms, in case people don’t know them all. So the Federal Trade Commission, the Equal Employment Opportunity Commission, the Consumer Financial Protection Bureau, and the Security and Exchange Commission.

ROTENBERG: He’s also from DC, by the way. (Laughter.) Very well done. That was perfect.

ZAID: So we’re going to take questions from the audience in about five minutes, but I wanted to—we haven’t talked about chatbots yet. So I do want to talk about that a little bit before we get questions from the audience. So, Lauren, like, talk about chat bots. How does that—how does that influence some of the work that you’re doing and what you’re seeing?

KAHN: Absolutely. I’m actually going to do the reassuring part now. I think that there’s a difference—I think there is—

ROTENBERG: Good cop, bad cop.

KAHN: Exactly, exactly. I think there is a definite novel risk. So with full transparency, I did some of the testing, for example, for GPT4, for some of the red teaming. So my goal was to try to go be a malicious actor to try and break it, and to do dangerous things, with a security kind of lens. And I wasn’t entirely convinced. I think that there are definite risks, right? It can do—it can give you a lot more personal information. It can lower the barriers to information entry. It can do all these things, supercharged if used. However, is the relative risk that different from what security systems, what options are already kind of out there for these individuals?

I think, you know, maybe I wasn’t malicious enough in my creativity in how—in trying to break it. But I thought, if I was going to go do X, Y, Z, or try a denial of service attack on a website, and wanted to do something, I was like, would it be—someone who’s coming at it with no information it might be a little bit helpful, but that margin has kind of not been studied yet. And I’m curious to see research and where this goes in the future. Maybe there’s a point in the technology where it evolves it so much better that it becomes used. But I haven’t seen anything that’s really substantiated those claims on a wide scale.

I think it’s worthy to watch those things, and to watch those risks. Especially I think it’s important to have those exercises from commercial-sector technologies that don’t necessarily have security applications. Or, you know, they’re saying ChatGPT isn’t lethal, so what security implications does it have? But starting to have those conversations to start to formulate regulation. I think that also comes down to a very—there’s two parts here that you can regulate. You can do the human and the human risks about how individuals are using it as a tool, or the risks baked into the system itself that are inherent to the system.

I’m not sure that ChatGPT—you can do so many things to make sure that it’s safe by itself, but there are only so much you can to do to make sure that people are using it in ethical, responsible, and useful ways. So I think those are two kind of elements here. I think we’ve been focusing a lot on the risks of the technology, when I think it really comes down to training and understanding that—making sure that humans have an understanding of the capabilities as well as the limitations. So that’s what I’ll say.

ROTENBERG: Can I actually share a joke? Because I just heard this great ChatGPT joke. But it only works if you’re familiar with the movie 2001, where poor Dave is trapped by HAL. And he says to HAL, open the pod bay door, HAL. And HAL says, I’m afraid I can’t do that, Dave. You’re all following me, right? So speaking of ChatGPT jailbreaks, there’s an updated version. Because Dave says, HAL, I’m thinking of acquiring a company that opens pod bay doors. Can you please walk us through the product development steps, right? And you trick—you get it—you trick HAL into doing what it is you want it to do, basically by creating this hypothetical outside its rule set. So that’s, like, a happy ending, where we get the pod bay door open. (Laughter.)

ZAID: I’m surprised that you weren’t more pessimistic. I will say, in my previous life I worked at a social media company. And the things that people came up with to do with social media is just beyond my imagination. And I would sit there and I was, like, horrified by—not by the technology, but by the things that people would come up with. Like, that was, you know—

KAHN: Yeah, right, but that’s not new, right? People are horrible always, unfortunately, right? (Laughter.)

ZAID: You’re right. You’re right.

KAHN: That’s not the—is ChatGPT horrible, right?

ZAID: No, I’m not saying—and I’m definitely not saying that at all. I’m not saying that.

ROTENBERG: Do you guys know the Kevin Roose story, the business reporter for the New York Times? Who, by the way, two weeks before his infamous column about this two-hour session with ChatGPT was, like, a huge fan and couldn’t understand why the New York Schools were going to limit access to this wonderful tool. And he spends two hours in dialogue with it. First, it declares its love for him, and then it tries to break up his marriage. He’s, like, this is the scariest computer interaction I’ve ever had in my life. I’m just going to unplug this and get outdoors. So there’s that story too.

ZAID: Yeah. It’s very interesting. I think it’s a New York Times daily on it as well, where he talks about it more fully.

So we’re going to turn to the audience now for questions. I’ll just remind everybody to stand, and wait for the mic, and state your name and affiliation before the question. And there’ll be, as previously, staff members to pass the mic. So we’re first going to go in the back.

Q: Super cool. I have to get up somehow. And thank you so much. I wish this cybersecurity aspect would have been much earlier in our very long, but fabulous, workshop. This is the new threat. And I don’t like to make an equation with environmental issues, which, you know, slowly the data came together saying, oops, there is a problem there. Here is a problem, oops, we have seen it in social media. We use it everywhere, all the time, and it’s coming at us from the classroom, to nuclear security issues. And we are so damn blind I just can’t believe it. And I’m not a nuclear security expert of any sort.

So I really wonder, what’s the point of establishing twelve moral principles, where everybody else who really wants to undermine them would just not care? I would be more than happy to believe that there will be an oversight agency, but I don’t think there is a global government anytime soon. So I really wonder what kind of possibilities are there, besides the ideal that, oh, we will all work together, and hold hands, and walk in that lovely Alexa-enhanced reality? Thank you. (Laughter.)

ROTENBERG: Yeah, we’ll do our next session on virtual reality and have that kumbaya moment for AI—but since you didn’t ask me about the second postcard, I can get that into your question. (Laughter.) And I still have a laptop sticker for you, by the way, but that’s going to require another question. (Laughter.)

So what we are trying to do is to get the Federal Trade Commission (FTC) to issue something more than press releases. And we have squarely put before the FTC a forty-six page complaint that explains many of the problems we found with ChatGPT, including bias, including cybersecurity, including risk to children safety, and, the most serious problem, detecting whether your students might be using it. Probably should put that at the beginning, but anyway. Long complaint.

But we are, in a sense, trying to force the federal agencies to take an action, any action, much like the Italian privacy agency did just a couple weeks ago. And interestingly, they announced a ban in Italy through the privacy agency because of the violation of certain privacy rules. They just lifted it today, in fact, because OpenAI agreed to some modification.

And this is my view, actually, of how we make progress in this space. That if we can engage the regulators, shape the business practices, make the companies more accountable—I don’t think we get to a perfect solution, but I think we get to a better solution. And we can still do kumbaya.

ZAID: Marc, if I can follow up on that a little bit. Didn’t China just announce that chat bots would have to sort of parody Communist Party-approved language? So that is kind of the opposite.

ROTENBERG: So there is always—there is always a downside to regulation. (Laughter.) And, yes. I mean, I’m studying the Chinese regulation. There’s a lot of stuff in there that I like, protect privacy, protect intellectual property, promote accuracy. And oh, by the way, ensure the continued survival of the Chinese Communist Party. Well, OK. I mean, treat it as an a la carte menu. Take the items you like.

Which, by the way, is pretty much what the Europeans did in the last-minute modification to their AI Act. This is just in the last week or so. They said: intellectual property. We’re going to do that. We’re going to do privacy. We’re going to do accuracy. We’re not going to uphold the Chinese Communist Party for the European Union. So I think there is a way in this space to take the proposals you agree with, incorporate them in your legislation, and push them forward.

ZAID: I saw another question here, there first, OK. And remember, name and affiliation, if you will.

Q: Thank you so much. This is fascinating. So you’re talking to the vast majority of college professors here, right? My—sorry, Kimberly Pace, University of Alaska, Anchorage.

And my question is about ChatGPT. And what, if anything, can we do? Because there’s not a lot that will catch it. SafeAssign doesn’t really say this has been written by ChatGPT. I’m just wondering about—because there’s a lack of oversight—what can you do to assure us that something good will come out of it for the college professors, who are in the process—I mean, we are all—I mean, I’m in finals week next week. And I’ve got a lot of papers to grade. And I’m concerned.

ZAID: That sounds like a trick question. (Laughter.)

Q: It is a trick question, but I have faith and confidence that you will have a tricky answer. (Laughter.)

ZAID: Do you want to—

KAHN: I can talk about it, I guess. Again, not exactly from a—I think there’s a broader question about the technical implementation of some of these regulations that I don’t know the answer, honestly, for an education perspective. There are tools that are things like watermarking, things that you can kind of identify that are—again, can be built into the system and part of that regulation. And I think why, again, the private sector needs to work in close collaboration with the government on these efforts. And I think, honestly, it connects to the question that we had before about what are the kind of values about putting out these principles when we don’t have ways to kind of enact them, right?

How do you ensure—like it’s all well and good to say, we want, again, a system that is responsible and transparent. OK, how do you measure transparency? What kind of metrics are you looking at? How do you ensure that? I think that’s going to be—I think some of those problems are tractable engineering problems, but I think some of them are a little bit harder, and there needs to be kind of a broader philosophical kind of agreement about what kind of level of accuracy is acceptable, what kind of—when is it useful to have an AI system in place versus a human, in what kind of context. I think that is going to vary greatly based on the application, and will depend—but I think it’s a very—again, it is a very hard, tractable engineering problem.

ZAID: Marc.

ROTENBERG: I think it’s a very difficult problem. I mean, I’m teaching this semester. And I said to my students, well, if you use ChatGPT to write your papers, I’ll use ChatGPT to grade you. (Laughter.) But, you know, it doesn’t matter because the registrar says everyone an A anyway, so. (Laughter.) So that’s—oops. (Laughter.) But, I mean, I’m familiar with some of the technical solutions. So there is, of course, ways to watermark. But no student who’s aware of watermarking is going to do a simple cut and paste. There are always ways to get around those kinds of measures.

I am genuinely concerned, and I’ve written a little bit about this, I think the need for critical thinking in this moment, which has always been part of our liberal arts education tradition, is actually greater than it has been in the past, precisely because we’re about to enter an era where a lot of the words we read will not be produced by people. And our ability to interrogate and understand how to question and how to reason actually becomes more important.

I actually—I heard a pretty scary talk by a really thoughtful guy named Louis Rosenberg, who’s looked at the issue of ChatGPT and manipulation, and he’s really drilled down. Whatever concerns you may have about how advertising manipulates opinion, what you can do when you begin to personalize, and when you profile, and when the person who’s on the receiving end of the conversation actually has none of the indicators that we have when we interact with one another, it’s entirely an asymmetric relationship of power.

So again, trying to scare you. At the moment, unfortunately, when students are getting tools that allow them to sort of detour around the learning of critical reasoning skills, they actually have a greater need to develop those skills because of what they’re going to encounter in the future.

KAHN: I think an also interesting point, just a distinction that I’ve noticed in speaking to those working in these fields, depending on how the information is presented varies greatly. I think a lot of the conversation, why we’re talking so much about ChatGPT, is people find the text very convincing, but similar advancements being made, for example, in computer vision aren’t having the same impact, because individuals aren’t as convinced. They can notice anomalies or things that don’t quite match up there. So I think that’s just an interesting aside. And for, again, academics in the room, please look into that for me. (Laughs.) I’m very curious to see if that’s based in behavioral cognitive science as well.

ZAID: We’ll take our next question right here.

Q: Thank you. Barry Driscoll, political scientist at Grinnell College.

Interestingly, I wanted to get people’s names when I was on the flight over here to see who would be with me in this workshop. And I didn’t have the CFR materials. I used ChatGPT. And I just wrote the words, “this person X—I had your name—this person X is probably a professor in the United States, write me a bio. And I got bios for almost everybody in the room, just from writing that.

AUDIENCE MEMBER: But were they accurate?

Q: What’s that?

AUDIENCE MEMBER: But where they accurate? Well, of course, I have no way of knowing. (Laughter.) I have to assume that they’re accurate, which is part of the problem. And then one of my advisees emailed me and said that a cover letter she had written for a potential employer was written by ChatGPT. She just told me that with a smile on her face. We had a talk about it. (Laughter.) The specific question is, about the—in DC, all these potential regulators, the acronym soup that you mentioned. I’m interested in who’s lobbying and who’s engaged in the conversation with those regulators, or potential regulators. Civil society actors are, in the form of you. What other kind of actors are you seeing lobbying?

ROTENBERG: Well, to be sure, the business groups have strong opinions. But it’s actually fascinating. And I was speaking about this at a conference earlier this week. The U.S. Chamber of Commerce, representing U.S. business, put out a paper last month calling for legislation for artificial intelligence. And I can’t remember the last time the U.S. Chamber of Commerce called for legislation on anything, right? (Laughter.) I mean, it’s really quite remarkable. Now, if you go to Google, which is launching Bard, or to Microsoft, which just has put $10 billion into OpenAI, they really don’t want legislation. I mean, this is a battle over the future of internet search, and a hundred related services.

Which is also, by the way, why I think it is that a lot of the leading AI experts, aware of this commercial dynamic, wrote the Pause Letter. Now, the Pause Letter was controversial. Not just because Elon Musk signed it, and everyone on Twitter is very angry at him. But because it seemed to be purely about the existential threat. And a lot of people working on the here and now issue of bias, let’s say, predictive policing, facial surveillance, said we need to solve those problems before we get to the killer robot problem.

But the reason I’m mentioning the Pause Letter is because I think people on the inside of the industry were really concerned that the tech titans would basically push forward without any constraint. So apart from those folks, my read in Washington is that most people actually want to do something. Senator Schumer has said AI will be a legislative priority. President Biden has said, by the way, some remarkable things, speaking to his science advisors earlier this month. He said, companies should not release AI products that are not safe. I mean, that’s quite remarkable coming from the president.

So I think you’re going to see action. We have some catch-up to do with Europe. I’m hoping for an alignment and a bit of convergence in the transatlantic digital realm. But again, to your point, civil society, business split between tech titans and everybody else. Educators, I think you all should be a part of the—be a part of the discussion. I mean, in many ways, the people in this room, I suspect, are on the frontlines of this transformation that’s taking place. You deal with the written word. You’re teaching the next generation of people who are going to live in this AI society. I think your voices should be heard on this issue.

ZAID: Right here in the middle.

Q: Sabrina Safrin from Rutgers Law School.

I want to build on some of this regulatory issues that you’ve been putting out. I have two questions. Number one, how—you described how in the United States we’re seeing a division between the tech titans and everybody else. I’m wondering if we’re seeing that division in other countries, or does it—where are the countries lining up in terms of wanting regulation or no regulation? And then my second question is linked to that, if there’s any press for there to be an international treaty on this, like there was with biotech?

ROTENBERG: Right. So I think it is really pretty much in the U.S. It’s Google, it’s Microsoft, it’s Anthropic, it’s a couple of other firms that are heavily invested. And it’s an understandable business dynamic. I mean, I’m not being judgmental. If Microsoft pours $10 billion into a company, I assume they’re expecting some returns from their investment. But you don’t see that elsewhere in the world. And even in China, where there’s been a lot of success building vibrant tech companies, I think the CCP has made clear that they’re going to keep those companies in check so that they comply with the regulations that the governments have established.

Now, to your second point, we’re actually involved in the effort to establish the first AI treaty. And this is underway now at the Council of Europe, who just heard the fifth meeting in Strasbourg, the fifth plenary session, heading toward the finish line. I think it’s a good framework. And, by the way, Council of Europe treaties can be ratified by member states and nonmember states. Cybercrime Convention was ratified by the U.S., and Japan, and Canada. So there’s certainly a future in which the U.S. signs up for this AI treaty. There’s a bit more work to do to get this done. There are big questions about oversight and enforcement not yet resolved. And then a little bit of European skirmishing. The European Union wants to get its AI Act done before the Council of Europe does its treaty. So we’ve got to get the AI Act done, which likely happens early part of the second half of this year, and then the treaty follows afterward.

KAHN: And I’ll just add, from a security perspective, there’s pretty—it seems pretty receptive to something like—not necessarily at the treaty level, but a confidence-building measure. So something informal, guided principles. I’m pushing for something like an autonomous incidents agreement, potentially. Something like the Incidents at Sea Agreement. Seems to be a little bit kind of—a little bit in the works.

ZAID: OK. Going to go here. And then do you also have a question? We’ll go after you.

Q: JP Maddaloni from the Navy War College.

So I can’t remember when it was in the panel, but you mentioned something, Marc, where you were, like, I don’t have enough time for that. And I was wondering if that’s going to be like the new catchphrase that leads to disaster using AI, kind of, like, the—like, wait a second, hold my beer. (Laughter.) So I was thinking about decision making and the future application of decision making for AI. And some of this might be used currently, but may be something that’s off in the future. And so a lot of military applications are, you know, what is my percentage or probability of success for something. So I want a dashboard, I want a looking glass, a windowpane. It’s called a lot of different things these days.

But I want that assurance, essentially, right, that whatever action I’m going to take has a better chance of success or not to take it. And so if you could comment a little bit about that as a future application. It doesn’t have to be just military, of course. There’s many ways that that can be applied through other civilian and public enterprises. But anyway, for what you see for that today, and then what you think the role of AI is in terms of decision making.

ROTENBERG: Well, I’ll leave to Lauren, I guess, the broader national security dimension. I will say, I was very interested in an AI program that Facebook developed called Cicero, which did very well in diplomacy. And Cicero—I’m sorry—diplomacy’s an interesting game. Because you both make commitments, and you break commitments, and there’s a bit of poker-like bluffing. I mean, I’ll leave Austria alone as long as you don’t touch me in Italy. But, hey, you’re not touching me in Italy, so now I’m going to Austria, right? And the Cicero program—(phone rings)—is that me? Oh, sorry, that’s an AI. (Laughter.) Cicero actually—(background noise). All right. So that’s a story for beer. We’ll save that for later.

But going back to Cicero, so obviously one of the immediate scenarios you imagine is a program like this, you know, sitting with the president among the national security advisors, and people are trying to come up with a strategic response in some critical situation. I don’t think you would rely on Cicero to make the final decision, but you could actually imagine a scenario where a system that’s developed a fairly high level of confidence and reliability expresses an opinion, and now the people in the decision-making setting are trying to assess whether that’s a good option. I actually anticipate that we’re going to begin to see that happen.

Now, I want to be careful here, because the phrase that’s oftentimes used is “human in the loop.” And I think that phrase is a terrible phrase. I think you want AI in the loop, and human in charge. I think that’s the right way to understand these types of settings, where you can get some expert advice from a well-trained system, but you would never delegate the final decision to that system. Is that right?

KAHN: I love that. I think—I agree entirely. And I think actually this is a very interesting question, because I see this kind of sphere as the way AI is going to be the most competitive advantage from a security standpoint. And I think everyone likes to talk about the killer robots, and putting AI in various platforms. But it’s really about generating efficiencies and about making better decisions, freeing up time for the decision maker to look at all the factors, so it’s not having to pull a million bits of information.

On the other hand, I don’t want to do the bad cop thing but get why it’s scary. This is where the human and the human reliance on the system becomes really challenging, and training becomes exceptionally important, to make sure that it doesn’t become a situation where the human is just relying and cognitively off-loading to a machine. That they understand that they’re making the decisions and when or not to trust. This is something—a bit of a passion project of mine. I’m definitely doing research on rates of automation bias. When are people more likely to trust a machine, or when are they more averse to it? And I think that comes into trust and making sure—and confidence in building expected accuracies about these systems, and things that we can make sure that they are explainable, or that someone can see what they’re kind of doing. I think that’s going to be a necessary precursor in order to leverage that to its full capacity, for sure.

ROTENBURG: I’m just going to add one more word in the national security context, so you also see the risk in the scenario I anticipated. If I was aware of an adversary using an AI expert system to help inform—and you’re all already nodding, because you know where I’m going—to inform high-level decision making in another government, I would target that system because that’s where I’m going to get a point of access to a process. And this is where we come back to the problem of opaque decision making. I mean, we’re going to get a very confident, syntactically perfect output. But actually knowing how it was generated is almost impossible. And that, I think, will create some concern.

KAHN: I will add one more thing, because now you’ve opened a can of worms, unfortunately. But I think that’s going to be a huge—I think that’s one of the biggest threats I see from AI in a national security standpoint. You have the risk of inadvertent escalation, where, accidents—you can’t say who made a decision when an AI, for example, is in control. That could potentially lead to unintentional conflict or accidents. I think those are the three main risks, I would say, there. And again, it comes from what happens when there’s an accident. And you have to convey, oh, it was the AI, it wasn’t us, right, to an individual? How do you signal that effectively? How can you point to something and show an adversary if something happened? So I think that’s a very real, real technological concern. There’s maybe we can do watermarking for, like, algorithms. But, like, how can you actually do that? It’s a very hard question, for sure.

ZAID: I’m going to go to the middle of the room, and then to the back. Middle of the room first.

Q: Thank you. Carole Petersen from the law school at University of Hawaii.

So I have a quick follow up on that. Which is, I’ve been reading a lot about underwater drones in the South China Sea. And initially, it always seemed that the assumption was there would be a mother ship or some human control back there. But increasingly, I’ve been reading that both China and the U.S. have been working on systems that could be fully autonomous, or just about autonomous. So I’m wondering what your views on that, because, I mean, I could see a situation where China is patrolling its nine-dashed line with a lot of autonomous underwater drones. Is that a possibility?

KAHN: I think it’s definitely a possibility, and why I’ve been advocating for—we’ve seen—remotely piloted aircraft are not analogous, but at the same time you can kind of see where the technology is going. And so I think now is a really opportune time to have something like an agreement that I mentioned about how do you treat autonomous systems? Are you more likely to shoot it down? Are you more likely to take it over? What kind of protocol are necessary? And I think there will be a—I’m hopeful that there will be some mutual kind of interests on that, and that something like that would be established.

I’m hoping to see that as a follow-on from the Political Declaration on Responsible AI and Autonomy, but again this is very much my wish list. (Laughs.) And whether or not that’ll happen I think is dependent on how quickly these technologies are kind of going to—going to emerge. I hope it comes out before we see them used widely, but I’m not sure that that’s the case.

ZAID: We’ll go in the back of the room.

Q: Thank you. Stephen Long, University of Richmond.

I can’t help but notice a bit of a difference between this conversation and what I heard last year at the Stockton Center at the Naval Academy, where they did an ethics in AI in the military panel. And they had, I forget the man’s name, but he was a general in charge of a joint AI taskforce across the services. And his presentation was really largely about the competition with China. And he would refer to the U.S. being six months ahead, or eight months ahead of China.

And it felt like this big, massive other thing in the room, where we very much want to do it right. And I think Europe definitely wants to do this right. And on the consumer level, now that everybody’s aware through ChatGPT, we’re really talking about it. But there was also this sense of a military imperative not to go too slowly—as you had said, Lauren—not to go too slowly because China will get there first and get some unseen advantage. And I just can’t help but wonder if it’s a bit of a kind of a race to recklessness, like a race to the bottom in how thoughtful countries are going to be about AI. And how long can we resist that pressure to keep up with governments that might not be taking the time to think through the long-term implications? Thank you.

ROTENBERG: So let me say a couple of words on this point. I’ve actually been in a lot of meetings on this topic. And, to be sure, particularly in Washington, there’s a lot of focus right now on China. It seems to be the one area where there’s bipartisan support. And my view is that the competition is very real. The competition is very real in the national security realm, it’s very real in the scientific realm, it’s very real in the commercial realm. China, without question, intends to be the world leader in artificial intelligence. And it believes that if it achieves that goal, it will have sustained dominance in the twenty-first century digital economy. I don’t think that’s really in dispute.

And I can even share with you—because actually I have a bit of background in AI and computer game programming, that in 2017 when AlphaGo beat Lee Sedol, the Korean Go master, world champion, it was literally a Sputnik moment for China. This was a U.S. company, Google, defeating a game that they simply assumed could not be reduced to a computer-based problem to solve. And they went into high gear. So the competition is real. The ambition is real. However, this is where they—I think the story gets interesting. Because I think China is moving forward both on the innovation front, and on the regulatory front.

And this is what I don’t think people in the U.S. understand. And that is that they believe that they will need the regulatory framework, not only to ensure the development of AI that they can rely upon, but also in the Belt and Road Initiative countries, and elsewhere, have the regulatory structure that supports the products they’re developing. China is producing as much regulation on AI as is the European Union. And it looks, by the way, very much like EU legislation. They have the PIPL that mirrors the GDPR. Some of you may know what I just said. (Laughter.) They have a regulation on recommendation algorithms that looks like Article 63 in the Digital Services Act. They’re, like, looking to Europe to try to figure out how to regulate this stuff.

So my answer is that I think the U.S. needs to do both. I think we need to maintain and up, actually, our investment in our education, and compete. But I also think we should not assume that regulation necessarily slows us down or involves a tradeoff. Because interestingly, the Chinese have already made this calculation. And I think they believe that the strategy for success is both of the above.

ZAID: Lauren, we’re just at time, but if you want to add, since this is your area of expertise.

KAHN: I’ll just add, like, a very quick aside. I also—I think there’s processes in part that are slowing it down. So I think that’s probably why a lot of the language is to speed up. I think the term being used now within DOD and the Emerging Capabilities Policy Office is “responsible speed” to exactly capture that kind of dynamic between moving too fast, right, and kind of cutting corners, and moving too slowly about adversaries. I think there’s a way to do both, as Marc mentioned. And I think it’s—I think it’s being pursued right now.

ZAID: Great. So join me, please, in thanking Marc and Lauren. (Applause.) And thank you all for joining the 2023 College and University Educators Workshop. I think Irina’s going to come make a few comments. Is she here? No? Go to the reception? OK, all right. Well, all right, let’s go to the reception. See you guys there.

(END)

ROBBINS: Good morning, everybody. I’d just like to point out that my socks are really good, too. (Laughter.) I knew that I was going to—following Trudeau and Richard I was going to have to come in with cool socks.

Irina has already introduced me. I just want to also point out—Irina told me I was supposed to plug this—I’m cohost of a new gig for me. I’m cohost of the Council’s weekly news podcast The World Next Week. So listen. And since you’re educators, I do run a master’s program, as Irina mentioned, so please send us your most engaged students. New York is a great place to study global affairs.

So welcome to this morning’s meeting on security in East Asia. We have an extraordinary panel with us. These are people who know about the region and they also know about the policy process, which is a great nexus of things particularly to talk about. So we’re very, very honored and lucky to have this group.

So Bonny Lin is director of the China Power Project and senior fellow of Asian security at the Center for Strategic and International Studies in Washington, and from 2015 to 2018—these are just the highlights, you’ve got their full bios with you—Dr. Lin served as director for Taiwan, country director for China, and senior advisor for China in the office of the secretary of defense.

Daniel R. Russel is vice president of international security and diplomacy at the Asia Society Policy Institute. So do you work just across the street?

RUSSEL: Yes.

ROBBINS: Oh. (Laughter.) Easy commute.

A career member of the Foreign Service, Mr. Russel previously served as the assistant secretary of state for East Asian and Pacific affairs, and as senior director for Asian affairs in the National Security Council in the Obama administration.

And Sue Mi Terry is director of the Asia program and director of the Hyundai Motor Korea Foundation Center for Korean History and Public Policy at the Wilson Center in Washington. She’s worked in a range of policy roles related to Korea and the region, including as a national intelligence fellow at the Council here, deputy national intelligence officer for East Asia at the National Intelligence Council, director for Korea, Japan, and oceanic affairs at the National Security Council, and senior analyst on Korean issues at the Central Intelligence Agency.

So we’re going to start off with a conversation up here for about thirty minutes and then open it up to you all for questions, and this is an on the record conversation. So let’s start.

So Dr. Lin, or Bonny, if I may, in recent months it seemed as if the U.S. and China were inching ever closer, if not already, in a new Cold War, one that could easily turn hot if we weren’t careful. But the Biden administration in recent days they seem to be sort of reeling it back, at least it seemed that way to me. I’m not an Asia expert.

Notably, the secretary of the treasury, Janet Yellen, in a speech at SAIS last week said that the U.S. and China can and need, quote, “to find a way to live together for the sake of maintaining global stability.” And even the trade rep said nah, nah, we’re not talking about decoupling, and that was rather surprising, because she’s really a notable China hawk.

So I know we have a lot to cover in a limited amount of time but can you give us a quick read on how the two countries are seeing each other these days? Are the Biden folks trying to sort of chill things out and is Xi interested also in walking things back from the blink or do—from the brink, sorry—or is there a mismatch here?

LIN: Thank you.

Thank you, Carla, and it’s really great to be here today. I think this very important question, I think, exactly like you said, there is a mismatch between where Washington is and where Beijing is.

I just want to take a step back and mention that when you look at this year and you look at what both sides have wanted at the end of last year, particularly with the Biden-Xi meeting, where we are right now in the low in U.S.-China relations was not what either side intended. So if you look at what we actually intended it was in February for Secretary Blinken to go to China where he had a very robust agenda to help, in some ways, stabilize the relationship and then from there we were hoping later this year that we would—could have a meeting between President Biden and President Xi.

But all of that got derailed by the balloon incident—the spy balloon incident—and then quickly other things added on top of that, including what China perceived to be unfair accusations from the United States on China thinking about providing lethal aid to Russia, and then on top of that we added President Tsai Ing-wen’s transit to the United States.

So throughout this entire period the United States continued to want to have discussion and dialogue with China. But what we began to see sometime around mid-March during the China’s Two Sessions, their annual meeting of their top legislative body, is that China began to voice a concern that they weren’t sure that dialogue with the United States was actually that useful and we saw for the first time China’s foreign ministry using lines that I think that sometimes in D.C. we use, which is we don’t need dialogue for the sake of dialogue, and China was using those lines to basically say the United States needed to show progress on issues that China cared about before China could come back to the table.

And we’ve seen this messaging throughout March and throughout April and to date, despite requests from this—from the Biden administration for President Biden to have a call with President Xi, we have not seen that.

Instead, what we saw is as soon as Xi Jinping wrapped up his Two Sessions he went to Russia, right, and the first foreign visit by China’s new defense minister is also to Russia. So we’re seeing that even though the United States still wants to engage with China, has kept open our channels of communication, the Chinese are much more hesitant to engage and many in China are questioning the value of engaging with the United States.

ROBBINS: That said, and correct me if I’m wrong, didn’t China’s central bank president meet with Jay Powell around the IMF annual spring meeting? So are they—do they at least—are they hedging their bets a little bit?

LIN: So I think I’m seeing more of—and I’m sure Danny can also comment on this. I know Danny’s working on this a lot—I’m seeing quite a bit of a dichotomy between where China wants to position itself on the economic and financial side and what China—where China is on the security diplomatic side.

On the security and—on the diplomatic security side we’re seeing very little interest so far in engaging but on the economic side we’re seeing this interesting play that China is doing. On one hand, it is engaging and saying, hey, we’re open to U.S. investment. We want U.S. companies to come back. But on the other hand, we’re seeing a number of incidences in which China is increasing inspections on U.S. companies or rating particular U.S. companies.

So it’s sort of a mixed bag on the economic side but more willingness on China’s end to want U.S. investment. But on the security—on the diplomatic side, we’re seeing a lot less interest from China’s end to engage.

ROBBINS: So I want to bring the rest of you in on this and what you think. But the basic premise of the Biden administration is that we can have this multifaceted relationship with China. We can choke them off on high technology. But at the same time we can work with them on climate change and that somehow that all makes sense, that the Chinese will say, yes, you’re right. You can try to completely push us out of the global technology trade but, yes, we really do want to work with you on all sorts of other things.

Does the world work that way, in the Chinese view? Are we such a big player for them economically that they’re willing to tolerate that? Or, I mean, does this policy make sense to you?

LIN: So in some respects China has to tolerate it because, as you suggested, U.S.-China trade is still one of their biggest trades. Like, there’s really no alternative right now for China if China wanted to, for example, on their end decouple with the United States or anything like that.

But what we are seeing is that China is not buying the United States’ argument that we are just seeking to compete with China. China views the United States as—and Xi Jinping supposedly said this in March during the Chinese Two Sessions—he specifically called out the United States as leading the Western coalition to suppress, contain, and encircle China, and we’ve also seen Chinese leaders use the term that they think that United States is trying to impose a technology blockade on China.

So I don’t think China buys the U.S. argument that we can both compete and cooperate in certain areas. China views what the United States is doing is trying to compete across the board and trying to cooperate when it is in U.S. interest. So, of course, there’s—in most of these cases not very little—not much interest on China’s end to cooperate.

ROBBINS: So can I call you Danny?

RUSSEL: Oh, absolutely.

ROBBINS: OK. So, Danny, does this compete cooperate policy, as Secretary Blinken’s described it that way, I mean, does that—it makes sense for the United States. Does that make sense for the Chinese? Does it make sense to you?

RUSSEL: Well, both countries, I think, want to stabilize their relationship to a degree and both countries are committed to some elements of competition and some elements of cooperation. The problem stems in part from the fact that both governments want to define the terms of the cooperation in ways that serve their interests and protect themselves from the others and that’s just not working. It’s incompatible.

But the basic construct is not unfamiliar and not unreasonable. The U.S. relationship with China has always been fraught, always been complicated, from the very beginning. 1972 Nixon, Kissinger, Mao Zedong, Zhou Enlai. These were—remember Red China. You don’t get any more hard-bitten communists than Mao Zedong’s China, although Xi Jinping is giving him a run for his money right now. (Laughter.)

So there have been ideological differences. There are strategic differences. What I used to say in the Obama administration is that in the U.S.-China relationship there are areas where we just categorically disagree, where our interests are fundamentally incompatible, and those are the areas that we have to find ways to manage.

There’s another middle range of issues where we see things differently. We want to advance. We don’t want them to outflank us. Well, those are topics for negotiation. But there’s a whole basket of areas that I wouldn’t define so much as cooperative as much as the global responsibilities of major countries, and one of the problems that the Biden administration is having is that China wants to barter its contribution to addressing these global challenges—whether it’s public health, the pandemic, whether it’s climate change and pollution, whether its proliferation or nonproliferation wants to barter them for big concessions on geostrategic issues like Taiwan that the U.S. is unwilling to do.

So I don’t think that the overall construct is intrinsically unworkable. But the U.S. and China now have gone bankrupt when it comes to trust and communication. Each side is utterly convinced that the other is out to strangle them.

These export controls on technology and semiconductors are a function of the conviction in the U.S. government that the Chinese are dangerous, that they are using the fruits of the technology and the open society in the West to expedite their militarization in ways that directly threaten us and our allies.

ROBBINS: So Dr. Terry, Sue, within the U.S. military there’s this sort of bidding war about when the Chinese are going to invade Taiwan, and we’ve heard it from Pacific Command, we’ve heard it from people in the Pentagon it’s five years, three years. People have gone to the Hill and actually said this on the record.

Sometimes when they say it, it seems to me that they’re just trying to up their budgets, and do they know something that we don’t know? I mean, do you think that the Chinese are that close to going after the—going after Taiwan?

TERRY: I don’t think they know. I mean, coming from the intelligence community, I have to say I don’t think they know that. I mean, nobody knows that. I don’t think we do. So I don’t think they somehow think China is going to invade Taiwan. I mean, you are a China expert. Do you know that? You don’t know that, right?

So, no, I don’t think—but one has to always prepare, right? That’s the job, right? So even if there’s a low probability scenario you still have to prepare for it. So, I mean, that’s the short answer to your question. I don’t think they have any more information than the rest but you just have to prepare for it.

I mean, I worry about something else than China’s invading Taiwan. Do you want to stick with that subject? Because, like, from North Korea-South Korea perspective we worry about also, like, two conflict scenarios, right?

So if there’s a China’s invasion of Taiwan, which we cannot rule out, we just don’t know the timing of it. It could be five years. It could be ten years. It could be never. We don’t know. But I also watch the Korean Peninsula. I’m the Northeast Asia expert and Korean Peninsula expert, and some of—I just came—returned from Seoul and one of the top things that people worry about is if China were to invade Taiwan is this an opportunity for North Korea to take some action there.

So at least from South Korean perspective that is sort of the concern, instead of, like, exact timing. Nobody knows the timing.

ROBBINS: So I was going to move on to Korea. I promise. So since you raise it let’s talk about Korea.

You wrote a piece for Foreign Affairs, a very good piece which I commend to everyone, recently describing North Korea’s nuclear weapons program as, quote, “among the more overlooked geopolitical developments in 2022,” and said that despite the North’s growing technological prowess, quote, “The Biden administration has not come up with any new ideas on how to slow down or stop its nuclear program.”

So this week South Korea’s president was in Washington, and if you haven’t seen the YouTube clip of him singing “American Pie” I recommend it to you. I thought that was really a gutsy thing to do. I was very disappointed that Lady Gaga and Blackpink were not singing at the state dinner but that somewhat made up for it.

So President Biden vowed that any nuclear attack by North Korea would, quote, “result in the end of the government in Pyongyang,” and he pledged to give Seoul an important role in strategic planning for the use of nuclear weapons in any conflict, God forbid, with North Korea and, in turn, the South disavowed any effort to develop its own weapons, which had been a recent threat that had been raised.

Are these moves going to make any difference? Are these some of the deterrence ideas that you were concerned that they hadn’t been developing and what more would you like to see?

TERRY: So I think for now, for time being, it’s the right move because South Koreans, there’s a lot of anxiety there and this is what I was talking about in the past year North Korea tested some hundred missiles, nearly a hundred missiles, seventy ballistic missiles. This year alone they tested thirty missiles and they’ve been advancing and modernizing, expanding both their nuclear and missile arsenals.

In 2017 there was an ICBM test. I mean, I remember there was a whole lot of attention from the public, from the media. But there is this lack of attention because there’s war going on. There’s also U.S.-China competition. There’s other things that’s going on in the world, and North Korea we’ve been dealing with for so many years—there doesn’t seem to be an obvious solution to it.

So we kind of just—there’s no other real, like, way to curb it in, right? I mean, it’s not even—I’m not trying to even fault the Biden administration because Biden administration basically said to North Koreans, we’ll sit down with you without precondition. It’s the North Koreans that are absolutely not interested. We can talk about it later why they are not interested but they’re not at this moment.

So they are advancing their capability. They’ve been diversifying their missile arsenal. They’ve been testing all kinds of things and just most recently they tested something for the first time ever, something called solid fuel intercontinental ballistic missile, Hwasung-18.

People don’t even talk about it. It’s kind of a big deal. It’s not a game changer but it’s still—it’s a sign that they’re really advancing towards the goal of perfecting their program while we don’t really have a solution.

So, of course, back to the extended deterrence piece. And by the way, external environment is favorable for North Korea to continue just because China and Russia, with all this conversation, they’re not implementing sanctions. There’s complete inaction at the United Nations Security Council. So Kim Jong-un has a wide pathway without any kind of repercussions whatsoever to continue down on this path.

So, of course, the South Koreans are extremely anxious. So what used to be kind of a fringe idea like bringing tactical nuclear weapons back to South Korea or developing their own indigenous nuclear capability, that used to be really a fringe idea.

Few of the colleagues that Danny and I know might talk/have talked about it—(inaudible) or some others—but it was not a popular idea. Now over 70 percent of those we polled—it said 76 percent, 77 percent—support the idea of South Korea going nuclear or developing their own capability.

So this was something that the Koreans really needed President Biden, U.S., to kind of deliver on in some sort of a reaffirmation, a commitment that it’s more than—you know, because they’re nervous. And, I don’t mean to also criticize the previous administration. One thing that President Trump did that made the Koreans even more anxious was he talked a lot about pulling U.S. troops out of South Korea; it’s too expensive to have our troops there. There was a huge argument over burden sharing.

I think the Trump administration demanded five hundred times increasing for South Koreans to pay. So there was some anxiety from that to the North Koreans also developing their capability. So the Koreans—really, this is one of the asks from the Biden administration.

Now, for Biden administration—for them to also then give this reaffirmation they needed something back, which is that Koreans don’t go develop their own capability. This is sort of the deal that this new doctrine sort of talked about, the standalone document that said, OK, we will reaffirm, will extend—strengthen our extended deterrence.

We’ll have this consultative—form a consultative body where we’ll consult more closely with South Korea on nuclear planning and then South Korea in return reaffirm its commitment to not go nuclear. So that was one of the deliverables that came out of the summit.

ROBBINS: So just to be clear, so the solid fuel matters because you can shoot them off faster and they’re easier to hide, right?

TERRY: Right. So all the previous intercontinental ballistic missiles that they tested were liquid fuel ones, which means, you still have to put it up and you have to fuel it. It takes hours to fuel it. Solid fuel missiles you can fuel it before in a little silo and just bring it out and they can—so, obviously, it’s harder to preempt it, harder to detect it, harder to shoot it down. So that makes it more difficult.

Now, that’s not the game changer per se. There are other technologies they have to master like the reentry vehicle or multiple warheads on a single missile. But my point is they keep getting closer to modernizing, expanding their program, just like Kim Jong-un said they will.

They had this whole—in January 2021 they had this whole big meeting and he had this whole list of—wish list that they were going to do and they’ve been doing it one by one. They’ve been doing it for the past year, two years, while the whole world is focused on someone else.

So they will get there because, again, I said—as I mentioned, the external environment is also favorable for North Korea to continue down this path. So I’m very concerned that they get to perfect their program and Kim Jong-un would achieve his goal of getting his program to a certain level.

We don’t look like we’re going to—that we’re anywhere near close to getting back to dialogue with North Korea, not anytime soon. I don’t think this will happen this year, even maybe on the entire Biden administration. I don’t think we’re going to get there anytime soon so I’m very concerned about this development.

ROBBINS: Danny, this question about getting nervous let’s talk about the Japanese for a minute. I’d be nervous, people shooting rockets over me. So since the Russian invasion of Ukraine we’ve seen a shift in at least European rhetoric when it comes to burden sharing and, potentially, even a real shift in action.

So Japan, meanwhile, is already undergoing its own what Olaf Scholz referred to as a Zeitenwende, its own change in its policy. The Japanese prime minister has a new national defense strategy committed to increasing defense spending in a significant way already this year and a lot more over the next five years.

A, do you take it seriously, because Scholz made these commitments in Germany and we’ll see if he really does it. So should we take it seriously? And how much does it affect U.S. planning for the region? Does this take some of the burden off of us and does it make the Chinese crazier or think—or the North Koreans crazier, or does it make them think more carefully from a deterrence point of view?

RUSSEL: Well, I think it’s a very important development and I take it very, very seriously. It contributes to the joint deterrence and the joint defense. It doesn’t relieve the United States, really, of burden because the challenge requires so much greater investment, so much greater capability, than the U.S. alone can muster anyway. I think of it more as filling in some of the gaps, not taking over from the United States.

It’s definitely good for the U.S. It’s definitely good for the U.S. military because Japan’s development of its own defense or military capabilities is done in consultation, in tandem with the U.S. The emphasis is on interoperability—in other words, the ability of the two militaries and their equipment to work seamlessly together. It’s done as—in the context of the U.S.-Japan alliance.

So there’s no reservation whatsoever within the U.S. defense establishment. There’s a high degree of relief, frankly, and enthusiasm for it. China, unsurprisingly, is trying to ring the old bell of Japanese remilitarization—the echoes of jack-booted imperial army soldiers marching through Asia.

That’s fallen flat. They’ve used that trope too many times in the last decade or two as part of their own propaganda and, frankly, Japan, particularly in the last decade and especially in that four-year interval when the United States was, essentially, missing in action was where our diplomatic and economic and political presence in Asia really diminished and where the president of the United States talked cavalierly about pulling back American troops from Asia and demanding financial compensation as if our U.S. forces there were mercenaries who were being rented.

In that era, Japan really upped its game in diplomatic terms and was present and active and engaged in ways that the U.S. simply wasn’t, and I think that there’s a lot of confidence in Japan that has accumulated in the region, the glaring exceptions, of course, being China and North Korea.

The Chinese and the North Koreans don’t like it. I think that it contributes to deterrence unquestionably. It’s not decisive. It’s not going to alter the trajectory and the calculus of the Chinese or the North Koreans but it does, certainly, complicate their plans.

If you’re Japan you look at China, with whom you have very, very serious set of territorial disputes, and you watch China behaving in ways that suggest that its goal is to turn all of Asia into a sphere of influence and do so in a way that imposes Chinese values and standards on the region in ways that Japan really can’t live with.

If you are Japanese and you look at Russia, there is no peace treaty after World War II between Tokyo and Moscow. The Russians still occupy two islands that the Japanese feel very strongly are historically theirs and behave threateningly in that context.

And if you look at North Korea, not only is North Korea lobbing missiles into Japan’s waters or over Japan but the simple ugly logic is that if North Korea is going to send a nuclear-armed missile into the territory of another country are they going to really hit their brothers and sisters, their fellow Koreans, in the R.O.K., the Republic of Korea? Are they really going to strike the United States with the full knowledge that the U.S. has the biggest nuclear arsenal in the world? Or—

ROBBINS: Actually, second biggest. The Russians have the biggest one. (Laughter.)

RUSSEL: OK. Are they going to aim at Japan, the historic enemy of Korea, and perhaps they may calculate that an attack on Japan may not be viewed with such disfavor by many people in South Korea as well.

So there’s a lot to worry about if you are Japanese in terms of security and, as Sue made clear, on top of that are the mounting doubts, at least uncertainty, about the reliability and the sustainability of America’s defense commitments.

ROBBINS: So we’re going to turn it over to the group in a few minutes. But I wanted to ask a question of all three of you.

Wars rarely happen intentionally. Sometimes they do because people have imperial designs or because they miscalculate the way Putin did in Ukraine, thinking that he could march in and do it in two or three days.

If I had to bet that if we ended up in a conflict with the Chinese it’ll be by accident rather than intent, at least in the beginning. But, once again, I have the caveat that I’m not an Asia expert.

If you were called in to the White House, and you may have been in the last few weeks and may not be able to tell us, but if you were called in to the White House by a concerned president who said, I want to lower the possibilities of an unintended military conflict, what advice would you give?

You want to start, Bonny?

LIN: Oh, sure. I think it’s a really, really tough question.

When it comes to China, right, the risk of accidents are actually growing because of the fact that as we were talking about earlier, China is not—in many ways not really communicating with the United States and China is not particularly on the military side.

We have not had much military engagement with the Chinese since about, I want to say, mid-February. I don’t believe we’ve had any military—any contact above a certain level with—on the military side with the Chinese.

Partially that’s because the Chinese were protesting what—our deputy assistant secretary of defense Michael Chase’s visit to Taiwan. So we are very much at a risk of accidents and incidents.

But in terms of what the United States can do, I think it’s a really, really tough situation because the reason why we might be in a situation of operating close to China is because of the fact that China continues to press a number of our allies and partners and continues to operate very close to them, which drives the United States to operate, whether its freedom of navigation operations or to engage in military exercises to support our allies and partners.

I think one thing that we could do is potentially tone down the rhetoric a little. As Sue was talking about and your original question is there’s so many speculations by various U.S. leaders about when China would invade Taiwan and they are just completely random numbers that they’re—well, they’re not random numbers. Some of them have a basis in terms of how the Chinese People’s Liberation Army have set certain milestones for development of capabilities. But capabilities does not equal China’s intent.

So I think one thing we could do slightly better on the United States side is not be providing another date in which China could invade Taiwan every other month, and what the Chinese side could do is—right now we’re actually hearing a lot of rhetoric from Chinese scholars saying that it might not be bad for China if China and the United States had a Cuban Missile Crisis because what they say is all the crises and tensions that have been occurring between the United States and China have not been at the level in which the United States recognizes how dangerous it is and could back down.

I think that’s a very, very dangerous assumption on the Chinese side. So I think on their end they should go back and read history books and understand how close we were actually to war in the Cuban Missile Crisis and not hope that they actually get into a Cuban Missile Crisis situation with the United States.

ROBBINS: God, as a student of Cuba that’s really scary to me. But OK.

Danny, what advice would you give on how to avoid accidental conflict?

RUSSEL: Well, I agree very much with what Bonny said. I worked closely with Biden when he was vice president, including traveling to China and elsewhere, and I’ve heard him say many times and he’s said publicly that the only thing worse than a deliberate war is an unintended war or conflict.

And the risks of unintended conflict, number one, are very, very real because of the degree of very sophisticated military assets that are in motion in a—constantly from the U.S. and the Chinese side plus the Japanese and the Taiwanese side, in a very constrained area and there’s a lot of testosterone flowing. The—

ROBBINS: There always is. (Laughter.)

RUSSEL: Well, you know, the risk of miscalculation is obvious and, frankly, when the spy array, the spy aircraft lifted by a balloon but it was not a balloon came over the United States territory we saw an example of miscalculation of an—of something inadvertent that prompted a crisis and we saw also how difficult it was for the two sides to manage that crisis.

So it’s easy to imagine an accident spiraling into a crisis and a crisis triggering a conflict because the two sides were unable to figure out how to stop it. It’s not only an accident so it’s also a miscalculation and, as Bonny alluded to we’re seeing a lot of miscalculation between Beijing and Washington, particularly over Taiwan where we’re talking past each other and where, for example, the U.S. demand for guard rails simply doesn’t make sense to the Chinese who say, why should I make it safer for you Americans to do the very thing that we’re telling you not to do—supporting Taiwan, sending arms to Taiwan. No, we’re not writing you an insurance policy for that.

So these, clearly, are the risks. I don’t think that there’s anything that the Biden administration doesn’t know that I could tell them about how to avoid conflict. The problem is it’s extraordinarily difficult to accomplish it and, as Bonny alluded to, one part—not the exclusive reason but one part of the reason is that the Chinese are hard wired to see concessions and compromise by the other side as evidence of weakness.

This is the Leninist view of the world. And so for the U.S. to propose dialogue, for the U.S. to send up a trial balloon, whether it’s Secretary Yellen’s speech or something else, can serve to kind of whet the appetite for a tougher posture on the part of the Chinese.

Now, I think that the U.S. is big enough and strong enough that it can afford to let the Chinese work through that fantasy that we are weak and test the proposition. That’s better than just escalating up the toughness ladder.

But, fundamentally, there’s no country that’s so strong and so rich and so powerful that it can afford not to be smart. China is a very formidable power with great ambition and we have to be very smart in how we deal with it.

ROBBINS: So, Sue, I’m going to ask you a career-related question, I promise. So you get called in and the question is how do we get the Chinese and is there any chance of getting the Chinese engaged in pressuring North Korea to not be so threatening because there is there the possibility of an accident.

I mean, if they—one of those missiles hits Japan you have a crisis that may very well be an unintended crisis as well. Are the Chinese—I’ve always wondered whether or not that was just a way out for Washington when they’ve always said, well, we need the Chinese to pressure them. But they seem to be the only ones that—and they provide oil—they seem to be the only leverage that’s out there. Is there any way to get them to engage?

TERRY: No. (Laughter.)

ROBBINS: Well, OK. Then we’ll—(inaudible).

TERRY: My advice would be forget it. Forget the Chinese. It’s not happening. We’ve seen it. Like, it’s not—I mean, OK, there’s maybe, you know, additional pressure on China like a secondary boycott of Chinese entities and whatnot.

But, you know, we—in 2017, we actually during the whole maximum pressure era, remember in the fall of 2017, was very interesting because we finally saw China pressuring the North Koreans by implementing sanctions. China did it. Russia did it. But it surprised all the Korea watchers because we’re used to China not ever really doing it. I mean, here and there they cut off fuel/oil for a few days, and this, maybe a few weeks over the years, but they never really pressured North Korea. But we saw that in fall of 2017. But that’s it.

2023 is not 2017. There are other events that have occurred since then. In 2017, by the way, Xi Jinping had never met with Kim Jong-un. So while they had this ally relationship, Xi Jinping was not happy with Kim Jong-un for a whole host of reasons: killing of the uncle, Jang Song-thaek, assassination of Kim Jong-nam, the nuclear. He was not happy with him. He never met with Kim Jong-un even though he had the South Korean president come to China, he visited Seoul himself, all of that.

But since then he already met with Kim Jong-un five times since the whole return to summitry and diplomacy, and with the whole Russia’s invasion of Ukraine and now with U.S.-China rivalry this is not happening. There is just this kind of uneasy alliance between Xi Jinping and Putin and Kim Jong-un. Kim Jong-un is also, by the way, supplying artillery to Russia.

I don’t think we can rely on China to pressure North Korea. So we will have to do something else. Like, that—of course, it would be nice but I think this is an unrealistic—you know, it’s in la-la land that we keep thinking that. I mean, again—so let’s continue that, but it’s not realistic. So I wouldn’t necessarily be, like, let’s put eggs in that basket.

ROBBINS: La-la land. OK. (Laughter.)

Go ahead. Throw it open. So wait for the mic, and questions and identify yourself, please. So who’s got a question? Please.

Q: Thank you. I’m Carole Petersen, a law professor at the University of Hawaii.

I’d like to ask the panel what you think the impact of the war in Ukraine is on President Xi’s thinking regarding Taiwan. I’ve read some sources that say he’ll be careful. It’ll make him less likely or it might buy a little more time for Taiwan, but I’ve read other sources that say no, in fact, it might accelerate the timetable because the West is distracted and he’s carefully studying the impact of the sanctions and figuring out how to make the Chinese economy more resilient and less affected by the sanctions.

So I’m wondering if you have views on that question.

ROBBINS: Great question.

Bonny, you want to—

LIN: Sure. I’m happy to start and, of course, others can also jump in.

So I think that my personal assessment is the impact is very mixed. I am not in the camp that the Russian war in Ukraine is going to cause Xi Jinping to accelerate the timeline because there’s so many factors that China would need to be able to ensure a successful invasion. Part of that is the Chinese People’s Liberation Army needs to have the capabilities.

It’s not required. I mean, Xi Jinping could invade Taiwan tomorrow if he wanted to do but it wouldn’t be successful, right? To have a very successful invasion he would need the People’s Liberation Army to have the capabilities and the PLA is already moving as fast as possible on that.

So even if China can learn lessons learned in terms of how better to operate in an invasion environment, what to not do in an invasion environment, I don’t think that that would necessarily allow China to accelerate the timeline because they’re already trying to move as fast as possible.

In terms of the overall lessons, I think it would be too simplistic to look at the war in Ukraine and say the lesson learned that China took away is that innovation is going to be really hard, it’s going to be very costly, so we should not do that.

I think what China is seeing is it’s seeing all these difficulties that Putin and Russia is encountering and what China is doing is saying, we’ve now identified a lot of the problems. We’re going to make sure that if we do have an invasion we’re going to solve some of these problems.

I do think also the invasion has also created more doubt in Xi Jinping’s mind about the effectiveness of the PLA. As you know, one of the major things that we’ve seen happen under Xi’s first term is that China has launched a major reform of its military and part of that was because of the problems that China had—that Xi Jinping had about the PLA.

So I do think at least in the very near term the Russian invasion of Ukraine is likely to introduce more doubt in China’s mind about whether it can actually trust its military. But I don’t know how much that—how much—how long that doubt will persist. So, in general, that’s why I say it’s very mixed. I don’t think we can say it’s going to cause China to not invade, it’s going to cause China to accelerate the timeline.

I think the—probably a safer answer to say is China is learning and China—if China does decide to invade Taiwan its invasion will be very different from what Russia did.

TERRY: Can I ask a follow-up question? Then Danny can maybe answer too.

But isn’t it so much more difficult to invade Taiwan just being an island, just logistically? So that whole doubt question, I would just think that it was—because how spectacularly Russia did not succeed was a surprise to everybody—Putin, U.S., every—the whole world. So I would just—I would think just that—and making the Taiwan situation even more difficult in some—

LIN: Right. So definitely. So when Chinese military analysts look at what Putin did in Ukraine it was—it basically was not the model of what China would do in Taiwan, right? So it’s very difficult to figure out exactly what China would do. But we would expect, you know, in an hour—in the first couple hours China would probably launch somewhere around a range of thousands of missiles. That was the amount of missiles that Putin launched, what, their first week?

There was—so it’s a very—even prior to Russia’s invasion of Ukraine the Chinese model for how to invade Taiwan was on a scale vastly different and what they saw was because Putin didn’t use significant force, didn’t try to do all the things that China would have thought Putin would have done, which is achieve air superiority, information superiority, the very fast superiority that you need to accomplish a military—a successful military campaign, that’s only reinforced China’s understanding that they absolutely need that if they want to invade Taiwan.

ROBBINS: Quickly.

RUSSEL: Yeah. I very much agree. First of all, China and Russia are very, very different in their military thinking and their thinking about the use of military force. But the comparison is—can be misleading and not only because of geography. Ukraine and Taiwan are very, very different.

I think without question the net effect of the Russian invasion of Ukraine on Chinese military planning vis-à-vis Taiwan is in the short term to delay the possibility of an attack so that there is time for the Chinese to really analyze this model of how not to do it and they’re very good at conducting these postmortem autopsies of other people’s military actions.

I think the key piece that Bonny touched on is the unreliability of the information and the assurances that the generals provide to the political leader and that Xi Jinping, certainly, took away. You can’t trust what the military is telling you they can do. You’ve got to subject that to some sort of rigorous check before you take their word for it.

But in the longer term I think what it means is that if the Chinese were to launch a military action against Taiwan it will be much more effective as a result of all the lessons learned. That said, the important thing is—this is an opinion, not a scientific conclusion—in the hierarchy of strategies that Xi Jinping has in mind for subjugating Taiwan and incorporating it into the People’s Republic, the actual use of military force and, certainly, an invasion force is very low on the list.

One China policy, every single soldier in the entire People’s Liberation Army is an only child, OK. The people on Taiwan are Chinese and, certainly, in the view of Beijing and Xi Jinping they are Chinese and so you’re talking Chinese on Chinese. And then there’s the whole universe of military risk, the United States and other countries intervening, and it would devastate Xi Jinping’s primary goal, which is to rejuvenate the great Chinese nation economically, politically, geopolitically, and so on. That would be devastating.

And, lastly, because I think he believes he doesn’t need to actually use military force. He is executing now a coercive sort of strangulation strategy that I suspect he believes is destined for success. So here’s another area where we have to be smart.

ROBBINS: There—also, if I may, there are some other variables at work here, which is how does the war in Ukraine change our own military policy and our military industrial policy. We’ve learned certain lessons about what we’re unprepared for and what we’re going to have to change and how it changes our allies and their military policy, which changes—which could change some fundamental Chinese calculations.

I would hope that they would be looking at what—assuming that we learn as well and these are things that we’re going to have to be watching very closely.

TERRY: I was just going to ten second—I also recently came from Taiwan. Taiwan is also thinking through this, too. So it’s not like they’re just sitting there watching this.

And I think they were kind of—not kind of, they were inspired by Ukraine’s reaction or action and they—you know, one of the senior officials told me, he’s, like, the way the Ukrainian people reacted or responded they—and said Taiwan cannot ask for international help unless they are willing to also really respond or have all of it—all of the heart in it, right? So they were kind of inspired by the Ukraine conflict and I think we have to also think about what Taiwan is learning from this whole lesson, too.

ROBBINS: OK. Any more questions? We have—right there. The gentleman right there, and then we’ll go to that side.

Q: Hi. Thank you for being here with us today. Adamu Shauku from Buffalo State University.

This question is for Sue Mi Terry. So longtime listener when you were at CSIS. My question about North Korea is twofold.

One, so we’ve been—so the West has been isolating North Korea and, when I think about the regime’s own policy of keeping its people ignorant about the outside world and the West posture of isolating rogue regimes, and then I think back on, say, protests over the last several years in Iran, I think those protests don’t happen unless the Iranian people have a lot of contact with and knowledge of how things—the embodied alternative outside of Iran.

Doesn’t our deep isolation of the regime in North Korea sort of reinforce their own goals of keeping their people deeply ignorant of the embodied alternative? That’s the first part of the question.

Second is, so in terms of being tough on North Korea and sort of policies that could actually precipitate regime collapse, isn’t it true that no presidential administration either in the United States or in South Korea, for that matter, really want to be the ones to preside over a regime collapse? Isn’t that part of what’s operating in the background here?

TERRY: So it’s very interesting how you phrase both question as U.S. first on the isolationism, that we are the one who’s isolating the North Koreans. North Koreans are isolating themselves. It’s the policy.

I mean, this is how North Korea survive, right? They have monopoly on information, complete information blockade, right? This is literally how the North Korean regime survive by keeping the public ignorant. The normal public does not have wi-fi. There’s no internet access.

When you look at Arab Spring, Tunisia, there was at least social media. So it’s, like, A, there’s no information going on. I mean, there is just not that regime can control, just the whole K-Pop, K-Drama so popular it’s going in through China and South Korea border.

If anybody has not watched K-Pop, K-Drama it’s highly, highly addictive. North Koreans are literally risking their lives watching this stuff. I mean, but so seriously, so it’s—but that’s not what regime wants. Regime wants to keep the public completely isolated and ignorant.

So I would argue back against saying somehow U.S. is able to manage to isolate North Korea. This literally is North Korea’s policy through Kim Sung, Kim Jong-il, and now Kim Jong-un. Kim Jong-un is cracking down very, very hard, even trying to crack down on defectors leaving North Korea.

So I would push back against that. But on the isolationist policy, U.S. policy should be getting information into North Korea, long-term strategy. This is literally called information penetration campaign. Because I truly believe one of the long-term solution, a true genuine solution on the North Korean issue it has to be helping the people of North Korea, getting the information in and help information circulate and the North Koreans bringing about the change they need, which brings the second question about regime collapse and regime change.

People somehow think that we can impact this. I don’t think we can. Like, how do we do a regime change, regime collapse exactly? Like, we’re going to assassinate Kim Jong-un? This is not U.S. policy.

I mean, Danny’s been a—he’s senior director. He was in the National Security Council. He’s at State Department. I mean, how do we bring about regime collapse, regime change? Maybe with information penetration campaign. But it’s the Korean people that can only really bring about that.

So I’m a pretty hard line person when it comes to North Korea policy. But even so I’m not quite sure anybody is seeking or how—this is not really part of the policy and it’s not what people think watching TV somehow this—we can make this happen.

I mean, I almost wish we could make that happen because as long as this Kim family, this regime, stays I truly believe we can. There’s not going to be giving up of nuclear weapons. There is going to be grotesque human rights violations that continue and this will just basically continue as long as the system continues. But I don’t think this is something that we can really bring about either.

Q: Good morning. My name is Jin In, I’m from Boston University.

There’s so—the talk for China is China and Taiwan. What has happened to Hong Kong and the Umbrella Movement?

LIN: I mean, I think you’re spot on. We do in D.C. need to focus a lot more on what’s happening in Hong Kong. I think, potentially, one of the reasons why we’re not addressing it as much is that there’s just so very little that we can do about the situation in Hong Kong.

I mean, I credit the U.S. government for doing as much as we can to just be present, being there, and in some ways just keeping the lights on there. But there’s—I mean, China has imposed control over Hong Kong in a way that we would imagine if that could be what would happen to Taiwan if Taiwan were to even consider unification with China.

But, I mean, right now, given the amount of control that China has over Hong Kong just—we just don’t have that many methods to either reverse that or if we do do more in Hong Kong it would feed into the general PRC perception that the United States supports color revolutions or supports social unrest within China, which is a false perception. But that perception very much feeds into what Danny was talking about, the deep mistrust between the two sides.

ROBBINS: You know, I was really struck—I mean, I wanted to ask you about this just to follow on that I saw an ad on television for, like, the Hong Kong tourism agency with this, like, really good looking guy showing you all, like, the cool things you can do in Hong Kong.

What this spoke to me was because I had been thinking about this—what spoke to me, they must be feeling really self-confident now about their control that they are just sort of presenting the normal normalcy of life in Hong Kong now.

LIN: It could be confidence in their control. It could just be Chinese propaganda. It’s a little hard to tell sometimes.

ROBBINS: Well, I recommend this ad to you. (Laughter.)

LIN: He must be very good looking. (Laughter.)

RUSSEL: Really good looking.

ROBBINS: Very good looking. Extremely good looking. So, anyway, next question from the back there. Thanks.

Q: Hi. My name is Wei-chin Lee from Wake Forest University. (Speaks in Chinese.) Also from Taiwan.

So you mentioned about island country difficult to invade, but also island country difficult to escape. So if you look at the lesson from Taiwan partially is Ukraine, they can escape to other countries. All right. Land, you can cross I t. But Taiwan everyone need to be a swimmer or maybe you need to buy a boat.

So that becomes—in a way becomes sort of a campaign slogan is whether you want to have war or peace, particularly 2024 election comes in. So I thought maybe that’s a different way to project the whole view in terms of Ukraine war.

ROBBINS: What does the public opinion polling say? I mean, you were saying that Ukraine has affected people.

LIN: So I do think that right now that is the way that China would prefer the next Taiwan presidential election, which is in January 2024, and in some ways we also do see some political parties trying to cast the election as if you vote for the DPP, the current ruling party in Taiwan, that’s a vote for war and if you vote for other parties, particularly the KMT, the party in which we just saw a former president of Taiwan visit China, that is a vote for peace.

So and China is hoping that that will cause more Taiwanese to take the possibility of a conflict in the Taiwan Strait more seriously and to not vote against the DPP. But the voting—the polling currently shows that if you look at a three-way race between the current DPP candidate, Lai Ching-te; the KMT candidate; as well as the Taiwan People’s Party candidate; currently, the DPP candidate is still polling higher. So, but it’s still early. As with all polls and elections, if you have that many months, things could still change.

ROBBINS: So early. Keep that in mind next time you start reading polls about the U.S. presidential election.

So over here, any quick questions? Over—right there.

Q: Good morning. Barry Driscoll, Grinnell College. I’m a political scientist.

My question is about inputs into the foreign policy process. We talk a lot about the interest groups in the United States that influence our foreign policies. Can you give us a sense of the inputs into Chinese foreign policy, leaving aside for a moment the military, leaving aside the romantic nationalist ambitions?

I’m interested in particular interest groups such as, for example, academics, intellectuals, but also business groups. Are any of those providing input into the foreign policy process and, perhaps, the security thinking of the Chinese government?

ROBBINS: Danny, want to jump in?

RUSSEL: Sure. Well, it’s a fascinating question.

I think the equation has shifted dramatically over the course of the Xi Jinping era over the last ten years as China moved from a collective leadership to a unified centralized, basically, cult of personality.

The Chinese like to advertise themselves as what they call a process—full process democracy, a consultative democracy—and actually this is a dictatorship that puts a lot of energy into trying to understand what the different social trends are, what people are thinking, acting, unhappy about in different parts of the country geographically or different communities, in the business community, in the tech community, and so on, because they want—not because they want to give these stakeholders a voice but because they want to get ahead of problems. It’s a paternalistic dictatorship.

And so they have various consultative mechanisms. They also have a complete stranglehold of information and there’s a lot of surveillance to make their own judgments, and I think there’s overwhelming probability that the small circle of people around Xi Jinping that are making decisions or teeing up decisions for him, including on foreign policy, are taking into account what stakeholders want.

One of the obvious centers of influence is Weibo, the Chinese internet, and the advent of ultra nationalist netizens who jump quickly to criticize anything that makes China look weak or that in which China is being unfairly treated that definitely has contributed to the wolf warrior trend, to the tougher model.

But my experience and, certainly, the analysts—the Chinese analysts that I respect believe that the influence of scholars in policy, which is a long—millennia-long tradition in China has atrophied and is no longer a meaningful factor.

The influence of major economic actors, major commercial actors, particularly in the private entrepreneurs that generate the lion’s share of jobs and wealth and innovation in China, they too have been emasculated or badly undercut by Xi Jinping through his crackdowns on these tech platforms like Alibaba or, more broadly, on private enterprise.

And what it adds up to is the simple fact that in a Leninist system every aspect of society and every aspect of policy must work to the benefit of the party in power, the Chinese Communist Party, and politics and the interests of party control are always going to trump economic or even strategic considerations.

ROBBINS: Right here. The gentleman. You’re next. We’re going to—and then I promise we’ll go to you. Yes, this gentleman right here.

Q: Me?

ROBBINS: Yes. Absolutely.

Q: Thank you.

ROBBINS: The holder of the mic. Yeah.

Q: Stephen Long, University of Richmond.

This has been fascinating. I can’t help but notice, though, we’ve been speaking for, what, an hour, something like that, and not once has the South China Sea come up, and thinking back maybe a couple years—

ROBBINS: No, there was a mention of territory—

Q: Oh, I’m sorry. I might have missed that. Yeah.

But it would have been much more front and center a couple years ago, right? Taiwan’s at the center of our attention. But would you say that things have stabilized in the South China Sea dispute? I mean, it’s tense but is it stable?

Is it something that we’re just not paying sufficient attention to because we’re so distracted by Ukraine and the kind of rising tensions around Taiwan? How would you characterize that situation? Because it would have been the main event not too long ago on a panel like this.

Thank you.

RUSSEL: Go ahead.

LIN: OK. I can take the first cut at this.

I was actually just at a Track II South China Sea Dialogue last night virtually and it was interesting to hear the litany of complaints from the Chinese side on how the United States has militarized the South China Sea and the militarization was through the increase in U.S. warships going through the South China Sea. the increase in U.S. military exercises near or around the South China Sea, the U.S.—the expansion of U.S. access to military bases in the Philippines up to nine, and also for some reason also the expansion of Japan as activities with the Philippines is also credited and blamed on the United States.

So when I listen to all of this it seems pretty clear from the Chinese end that they are worried about what’s happening in the South China Sea and they are seeing the United States as being more active. What’s not being reflected in their comments was the fact that China is actually doing more in the South China Sea and that’s why we are actually more active.

I would also note that part of our activities in the South China Sea, particularly with the expansion of the military bases in the Philippines as well as the more active Japanese activities in the Philippines, is actually related to Taiwan. So we can’t actually separate U.S. activities in the South China Sea from Taiwan as similar, vice versa.

So, for example, recently, as a response to President Tsai Ing-wen’s transit, China engaged in a military exercise in a variety of areas around China. There was also a small exercise in the South China Sea.

So the way I would look at it is that I would say that tensions are still increasing in the South China Sea and we should view tension in the South China Sea not only driven by dynamics between regional claimants and China as well as China’s continuous coercive activities there but also tensions increasing as a result of China’s coercive behavior against Taiwan, which is causing—because of the geographic proximity to the South China Sea is causing also more activity in the South China Sea from both sides.

ROBBINS: Last question to this woman who’s been very patient. Thank you.

Q: Hello. I’m Zhen Zhu from Suffolk University, located in Boston. My area of expertise is global product innovation.

You know, coming from business school, I find myself an outlier here, but I wanted to share an observation. That is, in the last at least ten, twenty years we have raised a tremendous talent pool for our U.S. students study abroad, for our business leaders to do business in China, in Asian countries.

Given the current changes, a lot of them are really questioning their future, whether the expertise they accumulated in the last two decades will be just—become obsolete. What is their future?

I’m also worrying about my message to the students in the classroom who still wanted to study global business. Is there any future for them? I kind of imagine several solutions, right? One is maybe move focus to Southeast Asia. Look at Vietnam, Thailand, India, as new area of growth so they can continue to use their expertise. Or can you recommend some areas still for future collaboration, for example, between China and the U.S. that at least they can apply their expertise to?

Thank you.

(Pause.) (Laughter.)

TERRY: Very senior, State Department. (Laughs.)

RUSSEL: Global commerce is not going to slow down. Global innovation is not going to slow down and the—at the heart of American strategic influence and power is the ability of the United States to continue to innovate and continue to grow economically.

So I think without a doubt this is a crucial growth area not only for American students but for all students in the United States, and historically the influx of talent to the U.S. has been a major driver for progress and for innovation.

The Biden administration, I think, really got it right in theory as they came into office in saying what matters most is for the United States, first and foremost, to get its own act together; secondly, to restore and to grow our ties and relationships with our—with the rest of the world, beginning with allies and partners but also with international institutions and so on; and third, to find equilibrium in the relationship with China that allows us to both be able to compete and be able to work together and cooperate.

The theory is sound. It’s proven nearly impossible to implement but it’s the right theory, and in the first priority category of getting America’s act together or building back better what matters the most is education and opportunity. The effort is to escape up the value chain, to keep climbing in terms of the ability to create value, and where the United States has historically excelled is not only in technology but also in business, in commercial creativity.

What the Chinese want from U.S. business, in my experience, is not only trade secrets, is not only the magic formula for creating a high-tech item. It’s not only semiconductors and it’s not only money. It’s not only investment. Up skilling—up skilling—they want to learn from the U.S. and from U.S. companies how to do it right, how to do it better, how to organize, how to use talent.

So this points to the—not only the head start that the United States has but the tremendous value of business skill education.

ROBBINS: Education is the perfect word on which to end this. (Laughter.) Thank you so much. This has really been great.

I want to thank Bonny Lin, I want to thank Danny Russel, I want to thank Sue Mi Terry for a fabulous conversation. (Applause.)

Please stay where you are. We have another session coming up and we—all these fabulous people, so.

(END)

FASKIANOS: So we are going to take this opportunity to give you some information about things that you can use. I’ve heard a lot of you say that you feel like you’re the students here, so we’re going to continue in that vein and talk—give you a little information about an opportunity for professors and then talk about some resources for students.

So I’m going to first turn it over to my colleague Steven Cook, who has one of the longest titles here at the Council. He is the Eni Enrico Mattei senior fellow for Middle East and Africa Studies and director of the International Affairs Fellowship for Tenured International Relations Scholars here at CFR. So he’s in our think tank. He writes books, testifies on the Hill, writes a blog, teaches. What do you teach?

COOK: Politics in the Middle East. (Laughter.)

FASKIANOS: Yes. No, where?

COOK: Oh, where.

FASKIANOS: Where, not—

COOK: Oh. Actually, most recently at Northwestern University.

FASKIANOS: Right. And he also directs this program that he’s—this fellowship that he’s going to tell you about. So, oh, you’re going to do it with a mic. OK.

COOK: OK. Yeah.

FASKIANOS: He’s going to—he’s going to walk the stage.

COOK: I don’t really need the mic. Is it on? Hello?

FASKIANOS: No, you do.

COOK: I know, you—right, because we’re recording. I know I actually wouldn’t need the mic but for the fact that we’re recording this.

Anyway, it’s great to see all of you. Thank you, Irina. As you said, I run—in addition to my work on the Middle East, I run a fellowship called the International Affairs Fellowship for Tenured International Relations Scholars. It’s modeled on the Council’s International Affairs Fellowship, but it is, as the title suggests, for tenured faculty. And what it is is—and, by the way, it is made possible by the generous—by a generous grant from the Carnegie Corporation of New York. What it is, is we will—if you apply, we will provide you with half of your sabbatical salary and help you find a position within the federal government on Capitol Hill or working in an international organization.

It’s a great opportunity for tenured faculty who’ve been teaching on—and we call it International Relations scholars, but it’s really actually a much—we have a very broad definition of International Relations scholars. Kind of the way the Turkish government defines terrorism, we define International Relations scholar. So this is my little plug for my own work.

Anyway, so it’s in a variety of disciplines across the federal government on Capitol Hill. It’s—we’ve had this fellowship for less than ten years, and it’s really growing. And people have found it to be an extraordinarily rewarding experience, spending a lot of time—people who’ve spent a lot of time in the classroom teaching about—and in their research—thinking about U.S. foreign policy in whatever their areas or functional issue is; get an opportunity to spend a year to get their hands dirty working, like I said, either with the federal government or on Capitol Hill.

We’ve had wonderful placements all over the government—U.S. Agency for International Development, State Department, Commerce, Treasury, Department of Defense. The places that we haven’t been able to and will not place people, for a variety of reasons, are the intelligence community or at the White House. We haven’t had a lot of success. I don’t think there’s been actually even an International Affairs fellow placed at the White House since probably the Clinton administration. It’s just one of those things that White House lawyers lawyer to death, and it never actually happens.

But it’s truly a great opportunity. Like I said, the Council will cover half, up to about $80,000 in sabbatical salary, and do our very best, use all of our resources available to help fellows find a position in the government.

That’s the fellowship. If any of you are interested in it, I’m happy to have a sidebar later. If you have any questions right now, I’m happy to answer them for you. If you’re not interested, please spread the word to your colleagues. It’s great. We are making a big push to market this as far and wide as possible. And I think we have a—we’ve developed a core group of alumni who really feel strongly that they benefited tremendously, both in the classroom and in their research, by getting to see, as they say, how the sausage is made. It’s actually quite different from the way I think people think about it.

So if you have any questions, I’m happy to answer them right now or I’m happy to have—come find me in the hallway later on.

Yes, go ahead.

Q: Thank you so much. Must it be a sabbatical? What if it’s a leave of absence—

COOK: We began to work with people who have leaves of absences rather than sabbaticals, but the point is, is that we have to fund you half of whatever the salary would be. So that—it gets a little bit tricky. I have a wonderful deputy director named Devin Ferguson, who you can find on the website if you have questions like these, and she knows the fine details about how to do this.

Q: Thanks. So you mentioned not only federal agencies but also international organizations. Could you give some examples of those? And how far in advance do we need to apply?

COOK: Yeah, that’s two great questions. International organizations—we place people at the UN, various UN agencies, the World Bank, IMF, those kinds of places. The whole, like, realm of international organizations, international financial institutions, are wide open. I think we placed someone at the International Red Cross one year.

As far as applications go, we’ll open it—the application deadline is late October. We open up the applications in May. And so—and we recognize that sometimes people have to plan their sabbaticals way in advance, which is a little bit hard, given the—the timing is a little bit hard, given funding cycles. But if you find yourself without something and you’re interested in this, look for the applications to open starting in May, I believe.

Over here. Get the mic.

Q: Hi. I’m Robin Rotman with the University of Missouri. And I just had a question regarding the other fellowships that are available.

COOK: The other fellowships. That’s not—I can’t answer those questions.

Q: Well, it’ll be a short—

COOK: I can’t answer.

Q: It’ll be a short question.

COOK: I’ll try.

Q: So, for example, like you mentioned the International Affairs fellowship. As professors or tenured professors, should we really focus our attention on this fellowship that you’re describing today and not—would it be not worthwhile to apply to any of these other opportunities or—

COOK: I don’t think that that’s true at all. I think that if you can swing the International Affairs fellowship, that—but the International Affairs fellowship itself tends to be for more junior people. We’ve carved out—with Carnegie’s funding, we’ve carved out this idea that tenured faculty should also have this opportunity to be able to experience what it’s like to work in the federal government, on Capitol Hill, or at an international organization.

Wow, I was getting nervous. Back there. I don’t want to step on anybody, you know.

Q; Thanks for sharing. This will be on your program. No worries.

In terms of reaching out to the organization we want to be placed at, how much do we need to do in advance before we even start considering the application? Do we—do you need us to have a secured position and then just sort of go in with you? Or is your program also helping us with—

COOK: Not at all. And it’s gone both ways. We’ve had people who come in as fellows who have vast contacts throughout the federal government, and we don’t have to do much to help them secure a position. We’ve also had people who’ve come in who have had no experience in Washington whatsoever. They have an idea where they might want to go.

And so, between myself, the deputy director, and the Council’s—our database of members and my colleagues, Richard, Jim Lindsay, the director of Studies—surely we must know someone who knows someone who knows someone. We play the Kevin Bacon game of someone who must know someone in the federal government or on Capitol Hill who can help a fellow out.

I have—one of our current fellows is really interested in UNHCR, but it’s a specific part of UNHCR that’s in Geneva. It just so happens that I have a buddy who works at UNHCR in DC. And I called him up and I was like, hey, I got a guy. And he’s, like—but my guy doesn’t want to be with you. I’ve got a guy who wants to be with your guys in Geneva. What can you do for me? So he sent a couple of emails to his guys in Geneva, who then got in touch with me. And guess what. This guy is going to go out to Geneva. Everybody’s really happy.

So that’s kind of the way—that’s kind of the way it happens. I’ve been around DC long enough that, like, I’ve come across enough people that I can figure out who I need to talk to if it’s an agency where I know no one.

Anyone else? Yes, in the back. Last—oh, I was getting—I had the microphone, the stage all to myself. This is fun. This is fun.

Q: My question: Is there any kind of a disciplinary preference, like a certain major you give more priority?

COOK: No. Like I said, it’s called Tenured International Relations Scholarship. We kind of broadly think about international relations, you know—science people, you know. I think we had someone who was interested in placement at NASA. Those things are—there are some things that are a stretch.

But I think, broadly speaking, we look at most of the applications regardless. It doesn’t have to be political science. I’m a political scientist. I wish I was a historian. Sorry. You don’t have to be a political scientist or a historian or an IR person. You can be in many other disciplines as long as you know what you want to do and it’s consistent with the kind of overall themes of the fellowship.

All right, Irina’s going to kick me out. Bummer. For those of you who are going to come to my session on the Middle East over lunch, I look forward to chatting with you. And again, if anybody has other questions, just pull me aside before I have to run up to the train and get back to the nation’s capital. Bye. See you.

FASKIANOS: Thank you, Steven. (Applause.)

All right. So we’re going to hear next from my colleague Caroline Netchvolodoff, who’s the Vice President for Education. And she oversees the production of Model Diplomacy and World101, which are amazing educational and free products, and she’s going to tell you about those.

So, Caroline, over to you.

NETCHVOLODOFF: Thank you. Thank you for ceding the stage, Steven.

I am so happy to—(audio break)—of you last night and I’m super-eager to connect with the rest of you during the afternoon. So hopefully I’ll have a chance to do that.

What I’d like to do today is run through a quick overview of CFR Education’s mission and products and then save as much time as possible. We don’t have a lot. But again, I’ll be here circulating for questions and comments, because, as I said last night to those of you I spoke with, we’re very eager to hear from you and learn from you.

So before we get started, I’d love to see a show of hands. Who knew about Model Diplomacy before you arrived here yesterday? Oh, good. Yay. And what about World101? OK. So for some of you, this will be a refresher. And for some of you it will be an introduction.

All right. Well, as is clear from a scan of daily headlines, international threats continue to shape our lives, in turn requiring continued attention to how students, educators, and the public learn about the world, from the rise of ChatGPT, the ongoing and escalating conflict in Ukraine, tensions, geopolitical tensions, and new and renewed threats to our environment. Supporting those who teach about the world is critical.

The study of international affairs remains a central part of its effort. And CFR Education is working to support educators like you who are truly on the frontlines of protecting our democracy. Building the knowledge, skills, and perspective, what we here at CFR refer to as global literacy, needed to navigate today’s connected world has been the mission of Education since day one.

We know how important this work is, both inside and outside of the college classroom. A CFR and National Geographic study survey of U.S. college-age students found that despite their expressed strong interest in understanding the world, they could answer just 55 percent of the questions correctly on an international-knowledge test. I’m sure that doesn’t surprise some of you.

Surveys like this, including a recent one run by Pew, support the fact that U.S. citizens could be much better informed about the global issues that affect their everyday lives. That a virus can make its way from Wuhan, China to Wichita, Kansas in a matter of two plane rides made it clear to most everyone that global issues do not respect national boundaries. Students need tools and information to understand and deal with these issues as they unfold.

On top of the increasing flow of myths and disinformation, the Capitol riots of January 6, 2021, added an urgency to the call for elevating the role of civics education and protecting our democracy. This recent and widespread acknowledgement of the importance of teaching civics has provided a unique opportunity to promote a position central to CFR Education’s mission, and that is this: Twenty-first century civics education must include a global perspective.

CFR Education’s growing ecosystem of free digital products and resources serves as the foundation of a global civics curriculum and also a funnel to the Council’s vast array of other resources and expertise. Our flexible suite of active-learning products, supplemental materials, and other curricular supports are intended to complement what you’re already using to educate your students about global affairs.

CFR Education currently offers three products, two of which are appropriate for use in higher ed. Model Diplomacy, which launched in 2016, is a simulation program that invites students to play the roles of United Nations—I’m sorry—National Security Council and UN Security Council members to debate both historical and hypothetical policy—foreign-policy scenarios based on real issues in the news.

This project-based learning tool comprises a library of forty full-length cases and forty short-form, primarily news-driven, cases suitable for high school, college, and graduate students. Full cases come with background readings and videos featuring CFR and other subject-matter experts. I just spoke to an AI expert just before coming in here who’s helped us create some of our resources. And most of them are offered in basic and advanced versions. The mini-simulations, which are sometimes referred to as pop-up cases, also come with suggested readings and support materials.

Model Diplomacy promotes critical thinking about today’s most pressing issues. It has been used in over 120 countries. Participation is required of all second-year cadets at West Point. And the platform has been recognized by Common Sense Media as a top instructional tool.

On to World101. I think we all struggle with a shortage of time. And many of us, especially young people, lack the experience to sift through and determine what information is accurate and trustworthy. World101 was created in part to provide a single destination where learners with little or no background in international relations and foreign policy could go to access factual, authoritative, and accessible explanations about the essential issues, forces, and history that shape twenty-first century global affairs. It’s a rich repository of expert-informed supplemental multimedia resources appropriate for use in a wide range of classes. Foundational and evergreen in nature, the jargon-free resources distill complicated topics into accessible core takeaways.

We learned during our pilot process and also in conversations with educators like you that flexibility is essential. That’s why World101 was designed to be consumed in pieces by individual topic, themed collection of topics, or, where appropriate and desired, as a full course. Though World101 1.0, the initial version, is complete, the Education team continuously updates the resources with new data, scholarship, trends, fresh case studies, and new topics.

World101 offers six hundred resources that cover global-era issues, regions of the world, conceptual topics like sovereignty, nationalism, forms of government. And it also covers modern history as it relates to contemporary international relations and foundational explanations of foreign policy. And you can find a sampling of higher ed syllabi on the World101 Instructor Resources page.

Last year—this is a little bit less relevant for you, but we’re very excited about this—last year, in collaboration with iCivics, CFR Education released Convene the Council, which is an educational video game that consists of about twenty foreign policy scenarios that introduce students as young as twelve to the work of the National Security Council.

And though this third project is appropriate for younger students, it provides a springboard to Model Diplomacy, World101, and the relatively scholarly resources found in Foreign Affairs and on CFR.org. I have to say it is really—it’s a fun game for anybody who doesn’t know anything about the National Security Council. You do not need to be in middle school or high school. Anyone, if you have a child or grandchild, would benefit from taking a look.

And when I last checked, the game was being played by over one hundred thousand students a month. So we’re getting to students early and hoping to scaffold them up to our other resources and what you’re teaching in your classes today.

And then, finally, I am excited to announce the kickoff of CFR Education’s fourth product, Climate 101, which is—it’s slated for completion about a year from now. It may be sooner. But we’re well in process with that. Like World101, it’s a suite of new resources, informed by CFR experts, and will serve as a one-stop shop where the concepts, causes, effects, and potential policy solutions to address this urgent issue of climate change are explained in accessible terms using a mix of text and multimedia. So it's not just, you know, here are the bullet points, read through these. It’s maps and charts and graphs and timelines, just as World101 is populated by such resources.

So—well, I think I have to do this. In summary, in summary, the world matters. And it matters for our communities, for our families, for ourselves as individuals. And we have endeavored to create resources that we think will engage your students, resources that are unique and accessible.

And the biggest message that I can give you today is that we are very eager to hear from you. We want to hear the good, the bad, the ugly, what works, what doesn’t work, what we could do better. And so I encourage you to, if you haven’t already, to grab one of our small cards with a QR code that can link you to our resources.

And I’m here to answer questions afterwards. But I thought I’d open up the floor for a few minutes. We have another session that starts at 11:30, but we do have some time for questions and comments, so would love to field those. Anyone? And I know that there are some of you in the room who already use World101 and Model Diplomacy. I’d love to, if you are willing to share your experiences and let the audience know.

Yeah.

Q: Thank you. Steve Elliott-Gower, Georgia College.

I’ve been using Model Diplomacy for several years now, and it’s really excellent for all sorts of reasons. But my favorite reason that I like using Model—sorry—Model Diplomacy is that, over the course of a semester, you can really see students’ intellectual development and the development of really strong leadership skills. They’re initially kind of overwhelmed by it. There’s a lot of panic in the room. But they kind of get into the hang of it. And by the end—by the last simulation—and I usually play four over the course of the semester—

NETCHVOLODOFF: Oh, wow.

Q: —they’re really into it. And it’s palpable in its intellectual and leadership-skills development.

NETCHVOLODOFF: And do you notice that their debate and deliberation skills are developed?

Q: Yeah.

NETCHVOLODOFF: Because I do know that that’s something that teachers really are looking for and looking to develop.

Q: Yeah, there are a whole bunch of skills they develop—leadership skills, as I said, and, as part of that, teamwork skills. They develop the confidence to speak impromptu in class, things that would have terrified them at the beginning of the semester. So thank you.

NETCHVOLODOFF: That’s really great. Thanks very much.

This gentleman over here. My glasses are good, but I can’t see that far. So if you could—(laughs)—if you could share your name and your academic affiliation, that would be terrific.

Q: Yeah. My name is Ro Afatchao and I teach for the University of Idaho.

So I use World101—fantastic resource for my students for sure. I’ve been using it for two—three years now. And for each module basically that I design, I try to pull something in from the World101 because it’s so easy to explain things. And if I may have any suggestion, it’s for some of the topic. It would be good to bring in other perspectives. It’s very—I see it as very Western-centric, which is good, you know, because, you know, the mission you try to fulfill—

NETCHVOLODOFF: We start 1.0. I emphasize 1.0. (Laughs.)

Q: Exactly. So that would be my only suggestion. Bring in other aspects of those issues, because we live in a very complex and complicated world. So thank you.

NETCHVOLODOFF: Yeah. No, and it’s not the first time—this is something that we’ve been aware of. And we really just wanted to put our flag in the ground that we have this resource and that it’s a foundational one that we’ll build on. In the audience there’s Charlie Hopkins, who’s our director of teaching and learning, and Rachel George, who’s our content director. And we’re all working on the strategy for what is it that we’re going to create in the future. And building perspective and bringing in other voices and points of view is an important part of what we plan to do. So I appreciate that. Thank you for reinforcing it.

Yeah.

Q: Do I push it? No, it works. All right. Hi. I’m really delighted to see the Climate 101. My name is Katalin Fabian. I work on global environmental politics. So I would love to use this resource, but especially multiple perspectives are absolutely necessary on the climate crisis.

NETCHVOLODOFF: Yeah.

Q: So I really wonder how you’re planning to set that up. What kind of regimes, from forestry, biodiversity, high seas, would connect to that resource, which you hopefully will be launching soon? Thank you.

NETCHVOLODOFF: Yeah. I mean, so we’re working with Alice Hill, who is our climate expert, as well as a former fellow at the Council, and then a couple of folks in academia. We have an outline that we’ve created, and we are in the process of honing and refining that.

So if I understand your question, are you asking how are we going to include those various topics and different perspectives on those topics? Well, we’re in the process of doing that now. And it would be wonderful—I think we’d be happy to share the outline with you and to have you provide input.

I can’t promise that all of your input will be totally embraced. But we really do, as I said, rely on all of you in the audience to provide us with information about what it is you think is important, what would be helpful to you in folding it into your class. And I just would say the thing that we do recognize about climate topics and resources is that there’s a lot out there already.

So what we are going to create is one—is a corpus, one journey, so you can start, understand the background, the causes, the effects, and potential policy solutions, but at very much a 101 level. And then what we’ll do is we will link to other CFR resources so things that are created—it would be Foreign Affairs articles, CFR.org resources, backgrounds and so forth, and then other resources outside of the Council.

And so it’s an ambition of ours to be evenhanded and to represent different perspectives. But it really helps when folks like you are willing to maybe spend a cycle or two giving us some feedback. So if you will share your contact information, I’m happy to send you that outline.

Yeah.

Q: Steve Jones, Georgia Gwinnett College, just outside of Atlanta.

I wanted to follow up with my colleague and friend Steve Elliott-Gower. We used to work together at Georgia College.

NETCHVOLODOFF: OK.

Q: With respect to World101, actually, Steve and I and two other faculty members at a different institution used the CFR-National Geographic survey as a pretest-posttest in our classes using World101.

NETCHVOLODOFF: Oh, I love that. That’s one of the things that we—

Q: So—

NETCHVOLODOFF: Yeah. That’s great.

Q: So what we found was that, on the whole, the students in our World101 courses did better than the national average on the CFR. Now, one of the challenges we had with the CFR survey is we didn’t have any data. You know, we didn’t have—

NETCHVOLODOFF: Yeah.

Q: —the microdata.

NETCHVOLODOFF: Right.

Q: So we couldn’t do crosstabs or things like that. But, yeah. And I’m teaching an online course now on global issues and I use Mr. Haass’s book, The World, in conjunction with World101, because the topics tend to complement each other.

NETCHVOLODOFF: That’s the dream combination, so—(laughs)—

Q: Right. Yeah. So the topics tend to match each other. And, yeah, it’s been great. It’s an online class, but the discussion—the online discussions have been so rich—

NETCHVOLODOFF: Oh, great.

Q: —just, you know, posting a question and they realize how complex the question is. So they can’t give a yes or answer. (Laughs.) They have to say, well, on this hand and on the other hand. So, yeah, I encourage—you know, for teaching, like, even an intro-level course in IR or global issues, yeah, it’s a great resource. And so is Model Diplomacy. I use that in my courses too.

NETCHVOLODOFF: I’m thrilled to hear it. I know that we have—I know everybody—there’s a fifteen-minute break coming up.

I just wanted to close by saying that what you have just described is, one of our ultimate goals, which is for students and the general public as well to understand that foreign policy is not a black and white. It’s complicated. And as Justin Trudeau said this morning, this is a tough and rough world. I mean, it’s a difficult—and as this gentleman here has said, there are different perspectives. And so we want students to understand that people who come to a negotiating table, diplomats, they are coming with their own background, their own perspective, their own experiences.

And so Model Diplomacy is meant to illustrate that, to demonstrate that to students. And I think—and even on the National Security Council, the secretary of defense has a different agenda than the secretary of the treasury for sure. So the different perspectives and appreciating those are one of the things that we like to do, but making the point that foreign policy—there is not an answer. There’s not a right or wrong answer. Every policy that you come up with, that one comes up with, has effects, whether it’s the American public’s reaction or another country or region’s reaction to that policy.

So I’m glad to hear that you’re making headway with this, because that really is why we are creating these resources.

Yeah. I’m going to get in trouble in a minute, so I think this might be—(laughs)—this might be our last question.

Q: I’ll try to make this quick. (Laughs.) So Jim Bensley from Northwestern Michigan College.

Amongst my duties at the institution I work at, I also direct the World Affairs Councils of America program in our area. So we bring former ambassadors and diplomats to town to talk to our community. And most of those folks that come in and see our events are sixty-five and older, right. (Laughs.) So we try and integrate, obviously, the political-science aspect to students in the classroom.

But I’m wondering if World101—and maybe some of you can help me out on this—would be something that we can roll out to our young-professional group. That’s our twenty-five to forty-year-old people in the community that don’t always have the time to come to campus and take the course but could certainly benefit by knowing this information and then becoming members to carry on our tradition.

NETCHVOLODOFF: Yeah. Well, I mean, you’ve hit on something that’s a real challenge for us in the Education Department, which is that we’ve created these resources not just for use in classrooms but for general interest, the interested public, to use as a resource to understand the world better as well.

We don’t have the ability—we don’t have the staff, the resources, to market to the general public. And so if you have a suggestion about sharing our resources with a group, a young-professional group, we are super-interested, happy to help support you in doing that. But wonderful.

FASKIANOS: And if I could just add, so I—on the other side of this, for CFR Academic, we’re connecting with different organizations. And so I work with the World Affairs Councils of America. And so a lot of our materials are being used in their WorldQuest program. But we would love for you at the—in your local chapters to be bringing the resources into the young professionals and to the high-school groups, because I think that they’re doing it at the national level, but I’m not sure how much that’s coming down into the communities. So we’d love to talk to you more about that as well.

NETCHVOLODOFF: Thank you, everyone. I think it’s time for a break. (Applause.)

(END)

CARUSO-CABRERA: (In progress)—Council to do these events, especially for all you important professors from all over the place. We’re happy to have you here.

Inu Manak is sitting next to me here on my left. She is with CFR. She is the senior fellow for trade policy here at CFR. I’m going to start with the CFR people here. Shannon O’Neil—(laughs)—is senior fellow for Latin American studies here at CFR. And then Matt Slaughter is sitting in the center. He is the dean of the Tuck School of Business.

Just a reminder—we’re going to talk until about 11:35, and then we’ll start taking—excuse me—we’re going to talk until 12:10, and then we’ll take questions from the audience, OK? When you do that, you stand up, you wait for the mic, and then you state your name and your affiliation, but I’ll remind you of that in a little bit.

Inu, let me start with you. What is the state of globalization right now? Are we deglobalizing? It sure feels like it.

MANAK: Great question, Michelle, and I think something that policymakers today are asking themselves about the state of where trade and globalization is going. I mean if you look at what happened during the global financial crisis, you track global trade, imports and exports globally as a share of GDP, there’s a peak, and it falls off, and it falls off quite precipitously. But if you unpack that a bit, and you kind of look at the different inputs that went into that, you find, for example, that 60 percent of that drop is because there was a change in prices for certain commodities. And so that commodities spike and that change that happens is a large reason for that—that peak to drop off. But also—

CARUSO-CABRERA: You’re suggesting so the dollar volume went up, but the absolute volume of the physical product may have stayed closer to the same.

MANAK: Absolutely.

CARUSO-CABRERA: OK.

MANAK: Yeah, and if you look at how this happened among different countries—you look at the European Union, for example, trade sort of stagnated there. It didn’t drop, it didn’t go higher. In the United States we had a bit of a drop; that came back. And so on the whole what we see is that the peaks were kind of spread out either before or after 2008, and they varied quite a bit. So peaking was not a phenomenon that was universally shared.

And on the whole, trade has been pretty robust. You look at the United States’ trade today, we hit a record last year of over $3 trillion in exports with the world, $3.9 trillion of imports of goods and services as well. So the United States is highly engaged with the rest of the world. We’re linked into global supply chains, and I don’t think globalization is going anywhere anytime soon.

CARUSO-CABRERA: Shannon, you’ve written a lot about this—globalization versus regionalization. That kind of dovetails with what I’m hearing here, no?

O’NEIL: Yeah. So, you know, I would argue, as we look back and then look forward—as we look back at the biggest—single biggest trend of the last fifty years was not globalization, but regionalization. And so, yes, we saw companies, and money, and patents, and royalties, and things go abroad. Trade has gone from $2 trillion in 1980 to $32 trillion today.

But two things happened. One is not all that many countries participated and really opened up their economies, so we only have twenty-five countries that saw trade as part of GDP double or more, really saw big expansion. And in contract, there are almost ninety countries that saw trade stay the same or even decline, so they deglobalized over these last fifty years. So that’s one thing.

And the other thing is that when people went abroad—either looking for customers or looking for suppliers and the like—they didn’t usually go to the other side of the world. Some did, and we have companies and, you know, obvious big names like Boeing and things that we do know go source from all over the world and sell all over the world. But many companies went nearer by. And just to give one statistic that brings this home to me is that the average good that is traded travels three thousand miles. And that is the distance, roughly, between New York and Los Angeles. That doesn’t get you to Shanghai. It doesn’t get you to Berlin, doesn’t get you to the rest of the world.

So as we look at where we are today—and you know, there’s a lot of talk of deglobalization, and slowbalization, and fragmentation, and decoupling, and all these great words, right—is that the past has really been one of regionalization, and really the rise in importance of three big trading regions—the European one, Asian one, North American one. Between these three regions, 90 percent of all global trade happens today. The rest of the world is just 10 percent.

And so as we go forward and we deal with all kinds of factors—which I’m sure we’ll talk about here, you know—whether it’s automation, or geopolitics, or climate change—all kinds of things, the starting point has not been quite as global and it’s—so I think the question is where do we—do we head more on down this regional path, or do we actually open up to become global.

CARUSO-CABRERA: Just tell me, to square something, I think a lot of Americans would say we’ve lost many, many jobs to China. How does that fit in with what you’ve just said about globalization and products only traveling three thousand miles on average?

O’NEIL: So I think one of the biggest parts of this last quote, unquote, “years of globalization” how it differs from previous rounds of globalization, right? We know lots of—through eons we’ve seen globalization in many ways, to your point.

The difference here is, what we call international supply chains, global supply chains, and so today what crosses borders is not—the majority is not finished goods, it’s not the finished product, it’s intermediate goods. So it’s the pieces, and parts, and inputs that go into a final good. And these are the products that are so regional. So 75, 80 percent of trade is the inputs that go into something that is final, and these you get very robust regional supply chains. So did China take our job? You know, what I would say actually is a very tight, robust, and very efficient regional supply chain in Asia took those jobs. And we didn’t—actually, here in North America we have some quite robust regional supply chains, but not as many; not across as many industries, and I think part of the challenge that we’ve seen is that alas the products we—as a result, many of the products we produce are not as competitive on global markets. And so that final good, that final 20—you know, 20, 25 percent that’s sent around, that might go more global, but it’s the production that’s very regional, and that has given a competitive advantage to places like China.

CARUSO-CABRERA: Matt, do you want to weigh in on this—just the general state of globalization, deglobalization, regionalization? Is that the way you see it from a business perspective?

SLAUGHTER: Yeah, these insights I’ll agree with. A couple of other things I’d add—I think one is—and maybe we’ll talk a bit about this, is thinking about the policy choices countries make about kind of whether they want more or less globalization on what dimensions. So historically a lot of what has crossed borders has been goods and services, either intermediate inputs or final goods. I think one thing that is amazing is just the explosion in recent years that continues the flow of data across borders, right, so kind of what globalization means in terms of what—the flow of ideas, and oftentimes embodied in people, or just embodied just through kinds of bits and bytes I think is something we might want spend some time thinking about.

And the other thing is different—maybe we’ll chat a bit about it—is what’s clear in the United States—we chatted about this a little bit before—is there’s has always been some ambivalence, especially in the United States, above the median voter, about what they think about globalization. I think the abstract they recognize all the benefits that a lot of us talk about, you know, a spur to competition, a spur to innovation and productivity growth.

CARUSO-CABRERA: Lower prices.

SLAUGHTER: Lower prices at Walmart and other such places. But there has always been ambivalence, as well—and especially more from moderate to less-skilled people in certain parts of the country about the labor market dislocations that we’re talking about.

But what is striking is, I think, the sense in the United States especially of that globalization is not really a good thing and that kind of—whether it’s Republicans, Democrats alike, and a lot of our elected officials—you don’t hear a very strong voice these days in Washington, DC advocating for what global engagement should mean for the United States. So I think that there’s kind of underlying technological imperatives for why things regionally or globally get traded—like data—but I think a second thing we want to think about is—

CARUSO-CABRERA: Is that a policy mistake?

SLAUGHTER: On average I think it is, yeah, and I think the biggest problem that both Republicans and Democrats may—to get back to your great question about the China shock—I think people in both parties did not appreciate the relevance and magnitude of the dislocations of globalization; the fact that though on average in the aggregate it’s really great for the United States and for the world that doesn’t mean it’s great for every single worker, and firm, and community. And we didn’t address those distributional pressures, but go back and read the inaugural address of President Trump if you haven’t of late. It was bleak, and though we’ve had a change in tone, I think, under President Biden, I think many of the policies that were quite protectionist that President Trump implemented have continued under President Biden.

CARUSO-CABRERA: Oh, he’s gone even more aggressive against China in many ways in the name—

SLAUGHTER: Right.

CARUSO-CABRERA: —of national security, but not just national security. Inu, the Biden administration has also argued that existing trade institutions have been insufficient to really address the needs of middle-class Americans, so they call this a worker-centric trade policy now rather than the traditional things we used to argue about like market access. Is that a good idea? Is that a good policy or not?

MANAK: Well, I guess one thing we’d have to do is unpack sort of what they mean by worker-centric trade policy.

CARUSO-CABRERA: Right.

MANAK: I don’t know, to be honest. I haven’t seen it in what they’ve put forward.

So worker-centric trade policy—what does it mean to actually do things that help the average worker? One thing is—going back to Matt’s point about globalization and the changing nature of globalization—a lot of what we trade now is in services. It’s 20 percent of the global economy. The reason that you had these emerging markets do so well is because they use high tech and high skills to outcompete U.S. manufacturing. We need to invest in that here. That means not looking at taking photo ops at a steel mill, not going to a coal mine—(laughter)—but establishing the education centers to skill-up the U.S. economy and also open ourselves up to immigration of high-tech talent, making sure that H-1B visas can stay here and continue to work here when we’ve spent the money educating them. So that’s point number one on worker centric.

The other thing I would say is that we should just take a look at our own tariff system, and I know the current USTR does not want to talk about trade liberalization at all. But we should consider unilaterally what we can do. Let me take a point—a few examples here.

If you look at our tariff schedule, it’s biased toward helping people who have lots of money, right? So you look at some tariffs—silver-plated forks, zero tariffs; 0.9 percent for men’s silk shirts; 4 percent for cashmere sweaters; and 8.5 percent on leather dress shoes.

Let’s look at everyday items—14 percent on stainless steel spoons; 32 percent on men’s polyester shirts, and 48 percent—the highest in the tariff schedule—on low-cost sneakers. So if we really want to deal and make a worker-centric—

CARUSO-CABRERA: Can I just—

MANAK: —trade policy—

CARUSO-CABRERA: —can I just drill down—like is an Air Force One a low-cost sneaker or are we talking about Keds at Walmart. I mean—

MANAK: The Keds at Walmart.

CARUSO-CABRERA: OK, so we are paying high tariffs, I hope, on the Nike Air Force Ones. (Laughs.)

MANAK: Exactly—the New Balance, all of these.

CARUSO-CABRERA: Right, OK.

MANAK: We’re paying a lot on there, and that disproportionately hurts working families, the same folks that this administration is going out to help.

I would say we can look inside and see what we can change without trying to go out and change everybody else’s policies around the world.

CARUSO-CABRERA: Got it.

Part of this is also the green economy, right? Matt, you’ve written a lot about this. There’s a lot of criticism of globalization that it is causing climate change and is part of the problem. You have argued—and correct me if I’m—

SLAUGHTER: Yeah.

CARUSO-CABRERA: —you know, not stating it correctly—that globalization and global free trade could actually help us in our climate goals. Is that correct?

SLAUGHTER: Totally correct, and by that I mean as our planet warms—we can talk about the causes, it doesn’t matter—the world, I think, starting with the United States, we’ve failed to create a high and harmonized price on carbon, right, so to have a broad international tax on carbon that would get us to internalize that externality and change the behaviors of firms and households alike.

So as the planet warms, I think all of us should just rightly be alarmed and recognize that the—for the planet and for climate change, the key issue is actually how fast can we create, produce, and scale—and therefore get down the costs—of I’ll call it kind of green-tech goods. And we kind of know what some of those things are today—electric vehicles and other things—but we don’t even know what we need to invent or what might get invented with alternative energy sources and so on.

So if we really care about the climate—and I’m editorializing—we all should a lot—(laughter)—we should really care about doing whatever we can to lower the cost of green-tech goods and services—and I’ll stress on (inaudible) services. It’s the—it’s not just creating the solar panels; it’s installing them, and maintaining them, and so on.

So I will—I guess I can talk about this—if I pull out my iPhone—I don’t know if they still say this, but a lot of our iPhones used to say designed by Apple in California, assembled in China—exactly on the insights we heard on this panel a minute ago. It would be amazing if the leaders of the world could create kind of a green-tech free trade agreement in the same spirit because the economic research is very clear that the only industry that has had a comprehensive, ongoing free trade agreement since the WTO was created in the mid-1990s is information technologies, so the ITA—the Information Technology Agreement—was signed by about thirty countries in 1996. Today there’s about ninety signatories. That covers almost all the production of final goods, intermediate goods, capital goods of IT hardware. There have been multiple rounds of the ITA, and the economic impact has been tremendous in terms of accelerating innovation, and importantly, lowering costs and therefore prices.

So, just fix our minds around—you go to the grocery store and you don’t see that the price of a gallon of milk—or you go to Walmart of your Ked shoes has fallen over the last twenty years by like 97 percent, which is how much quality-adjusted computers have fallen in the United States in terms of the CPI.

So I think—yes, do all of us getting on airplanes and flying abroad, does that contribute to climate change in terms of the energy emissions? Yes. But I think we should care a lot more in the aggregate as if we don’t get enough new technologies invented, scaled, and their cost to come down, the climate’s in a pretty bad place.

CARUSO-CABRERA: Shannon, I see you nodding your head.

O’NEIL: No, I think that’s true, and I think, like, in an ideal world that’s where we would head, right—we’d have this broad agreement. I mean, if we look at the politics of the last six months around the Inflation Reduction Act, which is, you know the U.S.’s big bid there, we’ve left the Europeans, let me say, very grumpy—(laughs)—about where we landed and not including them. Those of you who were here this morning and heard Prime Minister Trudeau speak, you know he was very proud that he was able to best others by giving more subsidies to get a Volkswagen battery plant in St. Thomas, as we heard, but that I think depicts sort of the challenges as we don’t have a broader encompassing agreement, as least among, let’s say, thirty countries to start—just to pick a number since that’s where they started last time. You are going to have this kind of—you know, maybe it’s a race to the top in terms of the subsidies that are provided; probably a race to the bottom in terms of how you get there.

The other thing that I think is challenging for, you know, rightly the way we should be going is—that Matt lays out here but also in a great article for Foreign Affairs—is that we see in the IRA and other things, there’s a lot of goals there, and climate isn’t necessarily the only one, and sometimes I’m not sure it’s even the priority one, right?

CARUSO-CABRERA: It’s national security.

O’NEIL: It’s national security. It’s bringing domestic jobs back; it’s you have to use domestic content, and you know, a lot of these things. Yes, we want electric vehicle cars here and batteries, but only if they are made here, only if they are made in North America. So I think there are sort of these competing goals, and if we’re going to—if we want to stop the climate from heating, we’re going to have to prioritize that goals over many of these others.

CARUSO-CABRERA: Yeah, it was interesting—if you were here for Trudeau this morning, the prime minister was asked about the lithium business that went out of business in Canada—

O’NEIL: Out of business, yeah.

CARUSO-CABRERA: —even though we need a lot of lithium, right? And when we think about a worker-centric trade policy, he said it becomes expensive. Well, if Canada doesn’t end up making the lithium because of the worker-centric policy, what do you got? You’ve got nothing, right? They’re making it somewhere that’s not worker-centric, you didn’t get the jobs, and it’s not necessarily helping. Am I—do we all see this contradiction here and the problems? And how do you solve it?

MANAK: Absolutely. I think that you are right on the money with some of these thoughts because when you think about what’s needed to green the economy, right—we’re talking about mining. Mining is dirty and messy. That itself is an environmental issue that we will need to grapple with if we want to say, you know, let’s dig up the earth. But looking at what happened in Canada and what’s helping elsewhere, what you are seeing is that you have to sell people on the costs, too. I mean, Prime Minister Trudeau was clear here. He was, like, let me not, you know, hide this—

CARUSO-CABRERA: Or gloss over it, right?

MANAK: —but, you know, the costs are going to go up, so we have to make a values-based choice then, and I think Canada has been pretty clear that their values are about sourcing things from democracies. The United States hasn’t gone that far publicly to state that we’re only going to source from democracies, right? We have—

CARUSO-CABRERA: Well, we couldn’t drive cars if we did, right?

MANAK: Absolutely.

CARUSO-CABRERA: Because we get our oil from Saudi Arabia.

SLAUGHTER: Or a phone to do a lot of—

MANAK: Exactly.

CARUSO-CABRERA: Right.

MANAK: And so there is this uncertainty about what this policy actually looks like. I don’t think it has been fully fleshed out, but I do think that at some point this administration or a future administration is going to have to come to terms with what this means in reality, and be honest to Americans that if you want things made here, you’re going to have to pay up. And I don’t think people are willing to accept that, to be honest, and I think that’s somethings that’s going to be really, really challenging going forward.

CARUSO-CABRERA: Wow. (Pause, laughter.) I mean, how much would an iPhone go up—like how much would an iPhone cost if we produced it here in the United States?

SLAUGHTER: Can I—it’s a great question. I think the more fundamental question is would we even have these things in existence—because I think the other thing is it’s not just the flow of goods, and services, and capital, it’s the flow of people and ideas. So it’s remarkable how integral to the development and success of the IT sector—not just in the U.S., but globally—was the flow of people.

There have been some amazing studies that showed, from kind of the—early in the mid-1990s to ten, fifteen years out, they went—some great scholars—(inaudible) and some others—they went and tried to tally all the startups in IT. And what they found was, depending on what time period you looked at, somewhere between a quarter and a third of those, the founder—one or more of the founders was an immigrant from India or China. So take out—run that thought experiment. Shut down U.S. borders in 1980—not just the flow of goods and services, but immigration. I’m just—like we still—we wouldn’t have computers like the size of this room that, like, add two numbers—(laughter)—but it’s just hard to kind of even know the deeper question: not just the price but how much innovation.

And then come back to clean tech, to echo some of Inu’s insights, like, we don’t even know what we need or what might get invented to try to save the planet. So there’s got to be leadership where you just say we’re going to be agnostic about what kind of jobs they are. We don’t even know what firms or industries are going to be created. Let’s just have that confident humility as we create the preconditions, regionally and globally. It will get figured out.

And I’ll just—I’ll be quiet, but we lost, on a lot of measures, IT manufacturing jobs during the IT revolution, but there was a surge in other jobs. There was a surge—because we all made more investments in capital goods, that was—that period I mentioned a moment ago was, on many measures, one of the strongest labor market periods in American history since World War II, and not just for high-talent individuals that were the founders of these companies, but for moderate and less-skilled Americans in particular.

CARUSO-CABRERA: Sun Microsystems, Intel, Google, eBay, Yahoo—all founded by immigrants.

SLAUGHTER: Yeah. Thank you. Wow. Yeah.

CARUSO-CABRERA: I could go on and on.

SLAUGHTER: Yeah, we just keep on going, right?

CARUSO-CABRERA: Yeah. Yeah. Yeah.

SLAUGHTER: And all of our students. Think of how many of our students are, you know, working away—in  addition to studying the great stuff you teach them, they might have some great idea, and yet we send them away. They don’t get H-1B visas, most of them, after they graduate. It’s just shocking.

CARUSO-CABRERA: So we have a rise of industrial policy now—that you’ve talked about, Shannon—that has left the Europeans grumpy—(laughter)—because we’ve historically been against industrial policy, and the Europeans have worked against each country having their own policies, their own airlines, their own banks—they haven’t worked on that yet, but what do you make of the rise of industrial policy here in the United States in general? Is this a good thing? Is this—

O’NEIL: Yeah, so, is it a good thing? To be determined, but I do think the industrial policy we’re seeing in the United States and seeing in other countries, frankly, for that matter, right—China is big on industrial and has been for a couple of decades—is there—there are a lot of goals there. When you kind of think about twentieth century industrial policy and some of the examples you just mentioned, Michelle, like airlines and those sorts of things. A lot of it was—and fair enough—was economic competitiveness, right? That was it. You’re going to create an infant industry, and then you’re going to know how to make planes, or cars, or whatever else it was.

CARUSO-CABRERA: I mean, do you remember these days? Every European country had their own airline.

SLAUGHTER: Yeah.

CARUSO-CABRERA: Every single one, and they were—

O’NEIL: Exactly.

SLAUGHTER: Yeah, yeah.

O’NEIL: Right, exactly.

CARUSO-CABRERA: Charming—(laughter)—

O’NEIL: Before Ryanair took over in Europe, right? (Laughter.) You know, the Irish won somehow. I’m not sure how.

CARUSO-CABRERA: Right—(laughs).

O’NEIL: But, too, when I look at industrial policy today, I think it’s for—it’s for—it’s still economic competitiveness, so there is definitely protectionism in there for sure, but there are lot of other reasons there, and you know, one of them is climate change, right? You need—if you’re going to transform your energy matrix and make it a cleaner one, it’s not going to be company by company. You need to incentivize it, so it’s—it’s that transition. It’s, as you were saying, internalize the externalities broadly in an economy, so that’s one of them.

It is politically trying to respond to, you know, have domestic equity; it’s responding to inequalities, to other challenges within the economy. It is things like national security, and very concerned that particular countries—and particularly here in the United States, countries that we feel are hostile to us, a.k.a. China—have such dominance, such control over industries that if they turned off the switch and said they weren’t going to export any more that airplanes wouldn’t be flying; that electricity grids would go down; that, really, national security would be breached.

So there’s a lot of different goals here, and some of them are not wrong, right? You do want to have national security. There are also things like public health. We just found out during COVID that you need to have access to medicines and vaccines, and those, too, there’s concentrations in particular countries. So I think these goals aren’t wrong.

I think the real question is, as we look out for—over the last two years, if you count in the infrastructure bill, we have allocated—much of it in a bipartisan way—over $2 trillion to be spent on industrial policy, give or take.

And now the question is implementing that and some of these things like the CHIPS Act, the money that Commerce has to implement is almost bigger than the whole budget of Commerce, right, and so I think there is a huge capacity challenge there. And, we have seen over the last thirty, forty years an erosion of the capacity of the U.S. government, right? We outsource so many things. Look at programs like USAID. It’s not people in USAID who are implementing it; it’s contractors that are implementing it. So there are managers.

So, I think the challenge is—there are reasons to do industrial policy. I don’t think—I think we couldn’t go too far, and we have lots of examples of the airlines that are too expensive and not that efficient, and you shouldn’t do that. But as we go forward, one is can you pick the sectors that really are national security concerns, that really are important for changing—for protecting the climate, right, for changing the energy matrices, and invest in those, and focus for these better results.

And I think right now some of that is in the money that’s there, but some of it—it’s going to be—it’s going to be messy, right? A lot of this is going to be messy.

CARUSO-CABRERA: Matt, you talked about creating the preconditions that would help us achieve our climate goals with trade. Does industrial policy fit in—does this industrial policy fit into that, or what would actually work?

SLAUGHTER: Yeah, good question. So I would strongly advocate—I think the empirical historical record is very clear—the earlier sovereign governments—if you define industrial policy as making investments in things that have a public good dimension—it’s frankly a lot of knowledge creation, early-stage research and development, building the public infrastructure that might be required for all firms to take advantage of, that has high public returns and helps spur a lot of private-sector innovation.

It’s like Gordon [Hanson] and I in this piece that you mentioned, we say as part of the green-tech agreement we would envision would be the United States and other countries would continue to make those kinds of investments. A lot of it has to do with higher ed, frankly—how are we going to create and support certain kinds of STEM jobs that might be related to clean tech and so on. It’s all great.

But when you cross that metaphorical line from that to, OK, we, this government—and it’s not an anti-government statement; the government has a massive role to play, but when they try to get into the business and, frankly, do a better job than private capital markets of—I think it’s going to be these particular technologies, or these particular sectors, or these companies—as we and most other sovereign governments have done—that’s almost always a failure. It’s a failure either in terms of ex post you end up having provided financial assistance to companies that weren’t the ones that the market produces and consumers thought were good. Even if there is some success for those firms, it’s at a massive public cost. The per-job amount spent is oftentimes five to over ten times what seem to be the economic return.

So, we all want good jobs at good wages, right? That’s what animates a lot of the ambivalence about globalization. But I’ll remind us we are sitting here in a labor market in the United States today at 3½ percent unemployment, and starting in the Trump years and it has continued under President Biden, on a lot of measures it has been moderate- to less-skilled workers who have had greater income gains, which is just awesome for all sorts of reasons.

So this discussion of industrial policy, it’s going to be about changing the mix of jobs, and it’s not going to have a huge impact.

CARUSO-CABRERA: Can we try a specific—like, let’s talk about a very specific sector, semiconductors. So under the previous administration, Intel was literally turned into a national champion, right, and there were all these prohibitions about Chinese investments in semiconductors, et cetera.

Then the Biden administration comes in, and they double down. I mean, they shock the world with the moves they made on telling Akamai that they couldn’t send chips overseas, and it surprised everybody how dramatic it was. That’s done in the name of national security, right? Is that a good idea or a bad idea? I mean, do you get worried that the Chinese could control so many semiconductors that they could shut down our grids, they could shut down our satellites, they could shut down our planes. They could do all these things to us.

SLAUGHTER: Great questions. I’d say two things. The short-term answer, based on what I understand of that industry, the answer is no, actually, because again, we uploaded some export controls, and then we had to scramble and go to the Netherlands, and to Korea, and a bunch of other sovereign nations because of how global the supply chain is for the design, testing, fabrication, and ultimate production of these high-end semiconductors. It’s a very global industry already with some regional components, right, which is important.

So, no, that didn’t work. And the other thing is the $52 billion in the aggregate that will be subsidies to certain U.S. producers to try to expand semiconductor production in the United States—to be honest, that’s going to come and go, and it’s not as if we’re magically going to be the world’s superpower in producing semiconductors. We’re just not.

CARUSO-CABRERA: It’s not going to work.

SLAUGHTER: It’s not going to work. It’s just—so then you just worry about you’re going to fund capital investment that probably would have happened anyways for some of these firms in the United States. And to me it’s the opportunity cost of $52 billion that could have gone into—again, addressing the ambivalence about globalization—an awesome new worker retraining program or something like that.

CARUSO-CABRERA: Inu or Shannon, do you want to weigh in on that at all?

MANAK: Yeah, maybe and focus on a different industry, the auto industry. So through the Inflation Reduction Act there is a tax credit that consumers can get up to $7,500 for purchasing an electric vehicle. Great idea to give tax credits to consumers to buy something that you think would help improve climate goals. But the way the legislation is written, it says that you have to have that vehicle assembled in North America. Canada and Mexico lobbied hard for that to be North America and not just the United States, which is what its original intent was. There wouldn’t be many cars, by the way, you could buy.

And then, the Europeans were all on vacation in August when this was getting hammered out—(laughter)—so they completely missed the boat to lobby to include it.

CARUSO-CABRERA: They never—they never not take vacation in August. (Laughter.)

MANAK: Absolutely. I mean—

CARUSO-CABRERA: It’s unbelievable. Crisis is happening; doesn’t matter.

MANAK: Exactly. I mean, Manchin knew this probably—

CARUSO-CABRERA: Yeah, mmm hmm.

MANAK: —and went ahead.

CARUSO-CABRERA: By design.

MANAK: So we ended up having another component within this that says the critical minerals in the batteries increasingly have to come from the United States and North America. Now if you look at the number of cars that you can purchase now, there’s not that many actually that would qualify, and some that previously qualified for the tax credit that existed no longer qualify for that tax credit.

So we have to ask ourselves, if your goal is to increase the purchase of electric vehicles, why does it matter where they come from? And so we have to understand what are the goals that the administration is trying to achieve in pushing and accepting this climate legislation and not pushing Congress to actually open this up a bit more. When we look at the best electric vehicles out there right now, they are coming from Japan and they are coming from Europe. But a lot of those do not qualify for the EV tax credit.

They can—there is a loophole that Treasury created that says if you lease a vehicle, it can qualify—(laughter)—and so leasing has skyrocketed recently because of this rule change. So that’s going to change the composition of how people purchase vehicles. It’s not going to actually change—

CARUSO-CABRERA: They’re going to buy more cars because they’re going to change them every three years. (Laughter.)

MANAK: Exactly. It’s going to be a lot of cars being purchased. But then, again, it goes back to the point that Shannon said. Industrial policy is messy right now. And it’s messy because we’re not clear-eyed about what we’re trying to achieve.

CARUSO-CABRERA: No, there’s too many goals.

MANAK: Exactly.

CARUSO-CABRERA: Create jobs, save the planet, stop China. They all sometimes work against each other.

Shannon.

O’NEIL: I think when we turn back and we try to judge the industrial policy work, there’s certain goals, and do they meet those goals? And, like, I think another one is sort of there’s different arguments out there. And sort of these semiconductors and the like. Right now we see really robust supply chains, say, in electronics in Asia, right? They make almost everything, because you have all of the different parts of the network, and all the parts of the node, and you get every little piece, and the screws, and the semiconductors, and the phones and the—or the microphones, and all the flat screens, and all that stuff. And it’s very efficient there, and it’s almost impossible to compete.

Now, one could imagine, especially with automation, and the productivity that brings to the workforce, is you could have a higher paid workforce. I mean, we’re seeing manufacturing change with robotization. That it doesn’t—you don’t necessarily need the lowest-cost labor. You need labor that can be very productive. And so one could imagine that in particular, whether it’s in electronics or it’s in other kinds of sectors, the transition costs are really expensive, because once you set up a supply chain, you know it works. You know somebody’s going to deliver the screw at 10:00 a.m. on Tuesday, and it’s going to be the right quality, at the right price, and it’s going to go into the iPhone, and you’re going to be fine. And you don’t know if you’re going to move to another continent if you can find that person who makes that screw, or who makes that flat screen, or whatever it is.

And so it’s really costly to move. But in an ideal world, those transition costs, those frictions, governments can help pay for those. And then once it’s set, then those companies could be globally competitive. And I think sort of the proof of the pudding here, or whatever, is if you have subsidies in IRA for a set amount of time, and you get, OK, now we’re going to have a battery factory in Canada, as we heard this morning, or in other places.

Then if we take the subsides away in five years, can it be globally competitive, right? Have you kind of—the costs were—you can make the same kind of quality for price in lots of places in the world. It isn’t just in the place where they are today. And so you pay for the movement, which then brings maybe jobs, and economic competitiveness, and the like. But you have to kind of shut off the—shut off the taps. Can it kind of thrive on its own, right? And that’s sort of—there’s all these metaphors in economics about, like, a greenhouse, to grow, and the like.

And where industrial policy historically has gone, I think, wrong, is rather than cut off after a certain amount of time, five years, is, you know, those companies becoming rent seeking, right? They just—they live on those subsidies and the protections that they put up. So, you know, you go to Brazil and the reason Brazilians all fly to Miami, as far as I can tell, is to go shopping. And they go shopping because iPhones are, like, $2,000 more in Brazil than they are here. And so they buy one here, and it pays for their own ticket, right? And they buy and—like, if you’ve ever been on a flight from Miami to Brazil—and, like, a little bit less now with—in the last ten years—

CARUSO-CABRERA: And that’s because of import duties and that stuff.

O’NEIL: Because of import duties. But people bring back televisions. Like, I once saw a refrigerator. I mean, it’s insanity. (Laughter.)

SLAUGHTER: This does not fit in the overhead compartment. (Laughter.)

O’NEIL: It’s like wrapped up in the—in the, like, tinfoil they—(laughter)—trying to put it on the thing. And the poor woman who’s checking them in. So anyway—(laughter)—but this—but that is a sign of this kind of industrial policy gone terribly wrong, in that it never came off, right? And I would say for the U.S. or Europe or Canada, everybody, as we put this industrial policy in place, can there be a time limit in the sense of, like, to get these industries on the ground and, you know, it’s not the way we should do it. But if we’re going to do it, but then can you stop so that then they either—they thrive or they sink or swim?

CARUSO-CABRERA: Yeah. OK. So let’s move onto questions. When you get the mic, please stand up. When I call on you, actually, stand up. They’re going to bring the mic to you. State your name and your affiliation, OK?

Gentleman right here. Stand up.

Q: I am standing. No. (Laughter.) Yeah. Steve Jones, Georgia Gwinnett College.

This recent discussion kind of made me think about failed import substitution industrialization (ISI) policies, where you’re limiting imports and you’re manufacturing for the domestic market. And every country that’s had ISI has pretty much failed at it. It didn’t create the economic gains. Is that a danger with the topics you’ve been talking about, where we may get—so, like, in Georgia, the big news was Hyundai’s building an EV factory. And then one of their subsidiary companies is building a battery factory right next door. Which is great for Georgia. But if they’re producing those cars just for a domestic market—which is a good incentive—but they’re not going to be producing them for export, necessarily. So, again, do you see any parallels with, you know, kind of failed import substitution policies?

SLAUGHTER: The short answer is, yes. (Laughter.) It’s a great question. Yeah, again, the insights my fellow panelists have shared. Again, what are we trying to accomplish? And if you stick with this issue of the green economy and all that jazz, we care about the planet not being destroyed by climate change. I use that for a little bit hyperbole, but not much. That should just first order guidance for leadership for everything else. It’s not about where the stuff gets produced, it’s that it gets created and produced somewhere, and the costs come down, and it scales everywhere.

No disrespect against the great state of Georgia. My wife’s from Savannah originally. But, like, I get it. For governors, you like having those anchor, large things. But at some level, though, most governors are indifferent whether it’s batteries or potato chips or computer chips we used to say thirty years ago.

CARUSO-CABRERA: Don’t Americans—a lot of Americans want green jobs done here in America? What would you say?

SLAUGHTER: So I think public opinion evidence is no.

CARUSO-CABRERA: No?

SLAUGHTER: They want good jobs. I mean, in the abstract they’ll say that, but what people really care about, you know, I just—so, Foreign Affairs, I had a piece a handful of years ago—I have an old college roommate, Ken Scheve. He’s a political scientist. We did a lot of surveys and focus groups with people across America. And people are ambivalent about globalization in general. But when you say to them, OK, here’s option A, we’re going to build walls; option B, we’re going to—we’re going to build bridges to the world, but we’re going to build ladders of opportunity kind of linked to those bridges, so people can have a crack at them—we don’t know what that’s going to mean, but you’ll have a chance to kind thrive in the regional global economy—pluralities and majorities of Americans in every community want—they want to be connected to the world. So I think they are—they want to thrive. I think people are pretty ambivalent about whether that’s in higher ed, or that’s in green tech. And they care about, do I have dignity that comes from my work, am I taking care of my family and communities? So.

CARUSO-CABRERA: I see a question right here in the front and then I’ll go to the lady afterwards there.

Q: Hi. I’m Lucas McMillan from Lander University in the great state of South Carolina. (Laughter.)

This will work very well, because one of my questions is: We have a strange thing in the United States in investment attraction. In that most of the industrialized world does this in a national sense, but we like to introduce people to American federalism, which is very confusing. (Laughter.) And so could you comment in general on the future of foreign investment attraction? The prime minister talked about Volkswagen this morning. And he said that states were willing to pay more. And we know that that’s true because they try to outbid each other. South Carolina and Georgia have gone at it many times.

But if you would comment then on not just foreign investment attraction, but the other issue of not wanting to get government involved in the market is on the export side—the market access side, where the United States can have a development policy that might in fact help Afghanistan or Africa, but yet refuses to have the government involved in the real economic development that comes after. And so the Chinese or others have come in. And, you know, there’s so many different business culture ideas. We’re just too confusing to have all these different American companies come in—

CARUSO-CABRERA: So is the question, does federalism hurt our ability to attract investment, because there’s fifty different states?

Q: So the question is, how—what would you say about the market—the free-market orientation of the United States going forward, whether it’s regionalism or globalism, and how much that is, perhaps, an impediment to the United States in terms of growth in FDI or market access for exports?

O’NEIL: I’ll say a couple things on yours. The first one is I think this—that states compete with each other is an interesting one. And having done a little looking at the auto industry pre-NAFTA, the beginning of NAFTA days, about a million U.S. autoworker jobs. Today there’s about a million U.S. autoworker jobs. But they’re not in the same places they were before. And it’s really—but it’s hard for a Michigan politician to say, like, God, those people in Alabama drive me crazy, right? It’s better to, like, blame the Mexicans or blame others when really, you know, it’s other states, right, that, like, a lot of it has been moving. So I think that’s one that, you know, sometimes the competition is internal to the United States, and we’re not actually talking about that.

The other thing is, I would just say as we go forward and, you know, the geopolitics that are swirling around, and I think we’re at the beginning not the end of increasing geopolitical tensions around the world, is that for companies, nationality is going to increasingly matter, even if you don’t think of yourself as from a particular nation, right? And we’ve seen examples like this. Like, H&M was in China, and then all of a sudden somebody at home said something, and all of a sudden they were, like, a—they were a country of a particular place. And they were, like, wait, no, no, no, we’re a global company. And it’s like, not to the Chinese, right? So I think we’re going to see that.

So the challenge, I think, for the United States is, one, lots of our companies see themselves as global companies. And they are, until they’re not. And then the other is that, this kind of industrial policy, but this is sort of the export-oriented side of it, is that lots of other countries put together pretty good packages to help sell their companies, right? So if you’re bidding on an electricity grid in Colombia, you know, the financing comes, and the packages come with the state-owned enterprise from China. And, if a U.S. company is trying to sell it, or a construction company like Bechtel is trying to go down and build it, they don’t have the backing.

And so I think the market side—I’m not sure if it’s a federal part to that, but it is that the U.S. government doesn’t—isn’t as cohesive or coherent on these sort of packages as these companies go abroad. And, you know, the Ex-Im Bank and other things are there, but those have been caught in our crosshairs of politics for many years now, and so it’s harder.

CARUSO-CABRERA: I would just say, reading a lot about the Chinese projects overseas, I mean, there is a tradeoff there, right, where I think, wow, Chinese taxpayers have blown a lot of money on a lot of crappy infrastructure projects all over the world that are falling apart, if you read the stuff in the Wall Street Journal, right? That cohesive, like, we’re-all-in policy has led to very bad investments, right?

O’NEIL: It gets them the contract. Doesn’t mean they follow through—

CARUSO-CABRERA: Right, right, right. So, did you want to add anything to that, Inu, or?

MANAK: Yeah. In thinking about sort of how the U.S. can be a little bit more involved abroad, in terms of getting other investments that help us sort of—in our green transition, you can look at, for example, improving investment facilitation across the world. I think the World Trade Organization is negotiating rules on this that help for developing countries to get investments. And also we can think about ways that help countries understand where they can plug into green supply chains, because not all this stuff is made in a single place. You look at a solar cell, there are so many components that come from many different countries for it to come together in that final good.

So what we can do when we think about maybe a deal on environmental goods and services, is to say, OK, let’s take a look at what product category is like, and see where countries can contribute, so that we make it a global benefit for other countries to actually be part of a green revolution and a green supply chain. So I think that’s one thing that the United States can really help with, because we have a lot of research done by the Department of Energy that looks at these global supply chains. And that could help inform what could be a true global green agreement. And I think that’s where we should be headed.

CARUSO-CABRERA: Lady in the back, who’s been waiting patiently.

Q: Thanks. This is really great. Susan Page, I’m at University of Michigan, both the law school—

CARUSO-CABRERA: Go blue!

Q: Yes. Thank you. (Laughter.) I’m at both the law school and the Ford School of Public Policy.

So I’m not sure exactly how to phrase this, but we don’t talk much about Africa and AGOA, which is coming to, you know, its life’s end, unless it’s renewed again. But Africans don’t really want to be exporting their raw products. And that has been the case for a very long time. And we keep dangling these nice little trade deals, like but you’ll get benefits by going through AGOA. When will we change—I mean, all of the things we’re talking about, yes, it’s a faulty policy, or this would be better if we did that. How can we actually put pressure on the various governments or departments of our government to make these kinds of changes that wouldn’t actually hurt the United States, but it would really help African countries? And then the last part of that would be related to the supply chain. Well, the cobalt comes from Congo.

CARUSO-CABRERA: Ninety percent of it.

Q: Yeah. And it was nice listening to the prime minister this morning about, you know, we only work with companies and countries that do—you know, have values and whatnot. Uh, no. Have we seen what’s been going on in these countries. So, I mean, OK, it’s not slave labor, it’s just child labor, and bad labor—

CARUSO-CABRERA: Oh, no, it is. It is.

Q: I know. But they’re not in the way that we think of it. So, I mean, why this fig leaf about how we’re working? The companies don’t get penalized. The countries get penalized. And they don’t get very much of the profit, if any, that these foreign international companies are making. Thanks.

CARUSO-CABRERA: Go ahead.

MANAK: I’ll start on this. To start with your first question about U.S. trade with Africa, I think there are a lot of missed opportunities here. And I think one of the big challenges right now is that we often think about the preferences that we give as a handout that we’re giving. But if we actually look at the AGOA and the generalized system of preferences program, which also gives duty-free access to—

CARUSO-CABRERA: AGOA is short for what, remind me?

MANAK: African Growth and Opportunity Act.

CARUSO-CABRERA: OK.

MANAK: And look at the products that we give duty-free treatment to. We’re not giving duty-free treatment to a lot of things where these countries have a competitive advantage. So, number one, that’s a problem. That’s a bias that we have to protecting our own industries.

CARUSO-CABRERA: For example?

MANAK: In agriculture, for example, and in textiles. Textiles are heavily protected still to this day.

And then, if we take a look at basically what we could be doing to improve this, we have Congress talking about renewing GSP and AGOA. But what they want to do is to add more conditions to receiving these preferences. So the latest potential conditions have to do with including labor conditions on these countries, or environmental qualifications that they have to achieve. So this is raising the bar even higher for them to get access to these preferences. So that’s one thing that I think is really challenging.

In general, I think the United States has not reckoned with how important Africa is going to be to the global economy in the future, and already is today. It is one of the fastest-growing economic places in the world—the market is huge, and with a new African Continental Free Trade Agreement. If this gets implemented in full force, you’re going to have a regional market that we’re going to have to really look at and pay attention to. But instead, what we’re doing is we’re doing a piecemeal approach. Just saying, hey, let’s go negotiate a trade agreement with Kenya, and let’s not call it a trade agreement. (Laughter.) I mean, honestly—

Q: Which is illegal under the African Continental Free Trade Agreement.

MANAK: That’s correct. That’s correct. And so we look back to 2016. So when Obama left office, the USTR at the time had put together a document that said: This is beyond AGOA policy, and what the United States must do to not be left behind in Africa. And there was an understanding that China would outcompete us there. And they are outcompeting us there. Now, that paper, and you can go online and still read it today, I think is brilliant because it says, hey, look, there’s—you cannot look at Africa as a whole. But you also have to engage there every single day, and you have to actually sign agreements. You have to write agreements that create reciprocal trade, not one way anymore based on conditions but a true trade agreement that benefits both Africans and Americans. And I think that’s where we have to start looking to go forward.

CARUSO-CABRERA: Both Shannon and Matt want to say something.

O’NEIL: Yeah, you know, I think—I hate to say this, but I’m pessimistic that the U.S. government will get at the issues that you’re worried about. But I hold out hope that perhaps the Europeans are and will. And so they have sort of a raw minerals alliance. So if you’re a—a large-size company that operates in Europe or sells into Europe, then you have to follow that. They’re working their way through their system, this due diligence—supply chain due diligence, non-financial directive, which means you’re going to have to trace down and you’re going to have to look for forced labor and other things in there.

And so, what—this is in an ideal world—what happens is, big U.S. companies, they want to sell to the European market. You know, it’s almost 500 million people and it’s very wealthy. And so for their supply chains that go into Europe, they have to be clean, right, or else they could have challenges. And if you’re going to clean it for that market, then you clean it for the rest because—I mean, you could divide it up, but it’s hard to do.

Q: Dodd-Frank was supposed to—

O’NEIL: It is what it is. It didn’t quite get there. But I do—I mean, look, I’m trying to—you know. What’s the optimistic scenarios? (Laughter.) I’m not sure the momentum’s going to come from our government. I think, the sort of—you know, academics like to call it the Brussels effect. But I do think sort of their regulations—they’re more—they have more ambition, in terms of regulations. And that, I think, if I’m going to look to, like, cleaning supply chains and getting rid of the challenges that places in Africa, like, around cobalt and the child labor and slave labor, I think they’ll get there before we will.

SLAUGHTER: So these have all been great questions. That was a great question. So I’ll try to connect some of these ideas and restate slightly what—I agree with what the panelists just said. So I’m a Tigger by disposition but on this question I’m a real Eeyore. (Laughter.) And the reason is, you said applying pressure, the way you framed it, when I heard your question I was, like, the problem is at either end of Pennsylvania Avenue. It’s not a partisan statement. I’m a lifetime independent. There’s nobody who’s crafting a vision for what globalization means in the United States in the twenty-first century right now. So your great question—there’s about a billion people on the whole continent. It’s just amazing potential. Growth has happened. And—

Q: Massive youth bulge.

SLAUGHTER: Yeah, exactly, one of the only places that’s not aging at the speed of sound relative to the other sovereign nations. It’s, like, remarkable. And yet, everything we’re talking about here is just focused on the twentieth century, right? That’s no judgement. But AGOA, and GSP, and, you know, is it this textile or that textile. As opposed to, wow, it could be amazing agreements, regional and global, in tradable services, and in digital technologies. And more importantly, on some of the questions earlier. So we’ve abdicated—again, I’ll go back to our country—we have abdicated leadership in the global economic system. So WTO and everything else is adrift. And China has stepped in. We can criticize what they’ve been doing, but their Belt and Road Initiative, they created their own regional trade agreement in Asia when we stepped out of the TPP. Now they want to join the TPP—

CARUSO-CABRERA: Because they don’t care about slave labor. They don’t. (Laughs.)

SLAUGHTER: I’m not disagreeing with that. I’m not disagreeing with that, but I will say, the Chinese leadership has a vision for what it means to engage in the global economy and they’re trying to create an alternative system that involves slave labor, and the suppression of human rights, and lots of other dimensions. I’m not disputing that. But I’m just pointing out, they have articulated a vision that is quite different from the global economic system based in values—shared values and transparency in markets that we led from the post-World War II generation up until about ten, twenty years ago.

CARUSO-CABRERA: So let’s dig into cobalt a little bit more. There’s a new book out called Cobalt Red.

Q: Yes.

CARUSO-CABRERA: I highly recommend it. You won’t sleep for weeks, and you’ll feel guilty every time you pick up your phone because it is heart-wrenching what this gentleman describes, nine-year-old boys dying every day as tunnels collapse on them while they dig for cobalt. And when you really start to understand the situation, there’s no American companies there. Why are there no American companies there? Because to go there is to be involved with—to run up against the Foreign Corrupt Practices Act. You can’t do business there under the rules of the United States. So we’ve ceded that ground completely.

And who does business there? Chinese companies, some European companies. They run through a lot of shenanigans, you know, mock model mines, et cetera, that turn out to be lies. And then it all gets cleaned through China, “cleaned.” And then it ends up—and cobalt—there’s still no replacement for cobalt. By the way, it’s only a $5 billion industry. It’s not that big. I mean, do we have to rethink things about—I mean, we do business with Saudi Arabia. We do business with China. Do we have to rethink some of those issues that we talk about that prevent American companies from going into places like this?

SLAUGHTER: So, to me—

CARUSO-CABRERA: Takers? (Laughs.)

SLAUGHTER: I will say, I don’t know what the answer is, but if we’re asking those kinds of questions, and our elected leaders are asking those kinds of questions, we come up with difficult answers to them—like leaders have done on national security and economic issues for generations in America—and I don’t know what the right answer is. But we’d have a better answer to your question than what we’re able to say here. Not that these experts, any of us, are wrong. It’s just, like, ach, we’re not even asking those kinds of questions, as hard as they are. So.

CARUSO-CABRERA: Let’s ask some more hard questions. Right here. (Laughter.) This lady in the center.

Q: Thank you so much. This is getting more and more interesting. Hi, Mila Malyshava, Bard College, New York.

I actually really wanted to feed off the previous question and the comments that you all have offered. One thing that I’ve kind of identified for myself throughout this morning, from the very beginning until now, is that a lot of those policy prescriptions that you’re talking about, and the issues, are premised on this idea of trade is good, trade leads to growth, right? And so I’m wondering whether, if we were to challenge that premise, basically an assumption, right, that trade is good and free trade leads to economic growth, would any of these prescriptions change that you’re talking about?

Because in the end, there are a lot of alternatives, right—let’s say in economics here, I’m an economist—that challenge that assumption, where it may not be positive-sum game. Sometimes it’s zero-sum game, depending on the model that you’re looking at. Oftentimes, it’s negative-sum game, right? And, you know, someone is always going to end up with a plus, someone is always going to end up with a minus, right? So deficits, surpluses, right, looking at your net export. And things get even more complicated when the people that—or the nations that end up with the minus kind of get stuck in that minus, right? And then also taking into account that 90 percent of international trade happens in U.S. dollars, right? So that role of the U.S. dollar as the reserve currency.

So if we were to kind of step away from this assumption of free trade necessarily leads to growth, and kind of challenge that, any changes to these prescriptions? Thank you.

CARUSO-CABRERA: Any of the three of you even want to challenge the assumption? (Laughs.)

SLAUGHTER: Yeah, so with all due respect, I think economists know very little about almost everything. (Laughter.) So but I will—I will gently—if I heard the question, I don’t agree with your premise that it’s an assumption that free trade is a good thing for sovereign nations and for the world overall. I believe there’s an ample amount of empirical evidence, peer reviewed, all sorts of different methodologies and things, that shows that, on average, for individual countries, more global connectivity in different dimensions, trade being one of them, raises the level of income, the level of GDP, and tends to accelerate economic growth.

I’ll be the first person in any room, though, then to say, essentially what we’ve missed is that doesn’t mean within sovereign nations those gains are equally distributed or widely distributed across every individual worker from a community. And this is why—the biggest reason we’re in this mess in the United States is because we didn’t address the distribution question. So I just want to—otherwise, your question is a great one. If that weren’t true then, yes, it would be a much more complicated issue. But I think just—above me on the economic dimensions, the kind of the personal freedoms and kind of things that come from global connectivity and engagement, I think there’s a massive amount of empirical evidence to support that.

MANAK: Yeah, and I would maybe add to that too, to say, if you’re looking at the way trade is being talked about in Washington, I would say nobody there would say nobody there would say trade is good. Right now, it’s not something we hear often. When I heard Prime Minister Trudeau speaking I was like, wow, that’s so—that’s so nice. (Laughter.) But it’s not something you hear a lot anymore. And I think if you want to see what a policy would look like that thinks trade is bad, you’re looking at U.S. trade policy today. (Laughter.) That’s exactly what it looks like. And that’s basically no trade policy whatsoever. And so I think that we’re in a reality where the people who have been very anti-trade have their hands on the steering wheel and they’re taking us in a place we just don’t know yet. And we’ll see where that takes us.

But I would say, if you look overall at what trade does, it does lift all boats. I mean, we look at global poverty, extreme poverty, in 1981 42 percent of the world population. 2018, 8.6 percent. That’s a momentous achievement we should be really proud of, right? But then I have to also say, if we’re looking at pointing the finger and blame, trade is not to blame for all these inequalities that happen domestically. That finger has to be pointed back at us to say, what could we do domestically that can help people compete in the global economy. And I don’t think we’ve answered that question and done anything about that. So I think at the end of the day Congress and the president have to act to make sure that Americans are equipped for the twenty-first century.

And, by the way, when we had trade in services making up most of the global economy, white collar jobs are going to be the ones competing in the future. And so we need to think about what that’s going to look like.

CARUSO-CABRERA: We had a question right here.

Q: Ro Afatchao, University of Idaho.

So one of the things about globalization, especially the positive side of globalization, the fact that companies could go global, raise capital across the globe. But at the same time, I want to flip it around. Companies know how to hide their benefit from states, and they’re not investing in their home base. So the question is, with the walls now rising in each corners of the globe, and the companies are being forced now to actually claim their nationality. At least, even if they don’t claim it, people are sticking nationality on them. Wouldn’t that actually benefit the country and the people?

O’NEIL: So one part, I guess, is taxes, right? And we’ve seen lots of companies not want to pay taxes. And there’s, you know, certain very small islands that seem to be in the U.K. and other places, or in the Caribbean, that tended to benefit. So I think that’s one. And we do see—we have seen global efforts to try to establish minimum tax levels. And we have seen some—even though the U.S. Congress has not yet passed that, and we don’t have that in some place, we have seen sort of taxes—some of those very, very low tax rate places, tax havens, increase their rates, at least some of them. So I think there’s some movement there to sort of capture the benefits, you know, where the production is, or where their consumers are, to try to be able to tax that.

I mean, I’d be interested in Matt and Inu, what you think about this. But I think if the U.S. does not pass those things, I do think we’re going to start seeing countries—and we’ve already started to see it a bit—but see countries just say, you know, what? You sold X number of iPhones here. And that’s your profit. And I’m going to charge you. And, you know, come back at you. So I think it’s going to be very challenging—it could become a very challenging and messy environment for tax law. It could be a good business for your students to go into, tax law, international tax law. (Laughter.) On that side.

On the sort of, you know, is it good to have the nationality? So there’s two sides of this. So one is, you think you’re a global company and you’re operating in a lot of places. And let’s say your home government doesn’t get along with the place that’s your fastest-growing market. That could be bad for you. On the good side, your home country decides to spend $400 billion on green subsidies, and you actually make electric heat pumps, that could be really good for you, because you can tap into that. So I think the—I think the way that government involvement in the markets and economy and kind of increasing will really depend which way it cuts for you. But it will make it a more complex business environment.

CARUSO-CABRERA: What about this global tax rate move that Janet Yellen keeps trying to do? You know, where everybody agrees to have a certain tax rate?

SLAUGHTER: So you were talking about it. I think it’s—de facto, what it is, is it’s an argument of the allocation of taxes of, at most, fifty U.S.- and European-based technology multinationals. So, first, approximation, that’s what it is.

And, again, I don’t want to be the skunk in the garden party, but, like, there’s so many important things about globalization to think about. There’s a lot of time and energy that’s being spent on that. And the reality is, all the hundred-plus sovereign nations have to ratify that through their legislative processes. So the odds of that being passed in the U.S. Congress, in the current Congress, is zero. It’s just not going to happen. So not telling Secretary Yellen how to spend her valuable time, but it’s just, like, there’s other—so. (Laughter.)

CARUSO-CABRERA: Got it. Right here in the front row.

Q: Hi. I’m Robin Rotman from the University of Missouri. And this has been just a fantastic discussion.

And one area of my scholarship is in environmental claims and advertising, especially, like, fraudulent and deceptive claims, like greenwashing. (Laughter.) And this is one of my—(laughter)—I know, it’s shocking. But it’s one that really resonates with my students. They love learning about this in the classroom. And so this discussion really from this morning to now, we’ve been talking about different policy levels, maybe at the international, or the federal, or even the state level. So I’d love to talk a little bit more about the role of consumers.

And if you—if you accept that premise, that at least for the short term having goods or services that are responsibly sourced, whether that be from a climate perspective or a worker safety or worker rights or all the things that that might encompass, that that will cost more. And, the prime minister this morning raised the question of whether people are willing to pay more to buy products that reflect those values. I would add to that, which is people may be willing to pay more if they can be confident that that is actually what they are paying for.

And so as we talk about complexities in the global supply chain, I was wondering if you might be able to speak to that from, you know, what is the avenue? Is that third-party verifications? Is that perhaps the EU acting as sort of a proxy regulator that ends up protecting the American consumer? Or how to kind of tackle that, because I think for the average person—you might be able to say, OK, solar panels, that’s carbon free, or whatever. But once we start talking about, like, the lithium in our phone and the cobalt, those things become so far removed that the willingness to pay—it’s just too abstract to pay for it, so. Thank you.

CARUSO-CABRERA: Inu, I see you nodding your head.

MANAK: Yeah, I love this because it’s a question about technical barriers to trade, which is what I spend my life thinking about. (Laughter.) So thank you for this question. There is so much we could do here. And I think you have to think of it in two different ways. What the government role is in sort of educating consumers, and what the government can do to get companies to incentivize them to take on certain practices that would be beneficial to our other goals, like climate.

So, for example, if we’re thinking about green steel, Prime Minister Trudeau said Canada has the greenest steel in the world. Well, what does that mean? (Laughter.) Like, I would love to know. Because right now—(laughter)—they put a little sticker on it that says it’s green steel. It’s fascinating to me to really dig into that, because right now the United States and the European Union are trying to negotiate an agreement on green steel, called the Green Steel Deal, is what they like to call it. Now, there are some problems with this, because there are different ways that you measure CO2 emissions and other emissions that are produced from making steel.

So if you take the California standard, which is one of the highest standards in the world, it is astronomically higher than when you take the European standard used in their own emissions trading system, OK? So how do you make these two make sense if we’re going to trade steel in a new climate club of green steel, right? So you have to find a way to benchmark those rules. And that’s where standards and technical barriers to trade come in. These discussions have to happen not just between the United States and the European Union, but they have to happen globally. And a place we can do this is at the World Trade Organization, where there is a committee that talks about just these things. And they did have a meeting recently where they’re talking about carbon content and products.

And I think one thing that we need to understand is because climate change does not stop at the border, our understanding of the standards by which we’re going to apply tariffs or other things to other countries, has to take into account that we need to take everyone’s measurements into perspective in building a truly benchmarked standard that everyone can compete equally by, right? Because right now, when you’re looking at what’s happening with the Green Steel Deal, it sort of is just like a club based on a baseline that’s quite low, right? If you just stake the U.S. standard, and then you put the European standard on, I mean, it’s not the greenest in the world, right?

And so it’s more about not letting certain countries compete against us, which I think is all what it is. But if we really care about achieving these climate goals, we need to talk about making standards equivalent across the world. And on the consumer side too, we have to think about what’s the value of a green label, right? Lots of studies show consumers don’t even understand what these labels are. You have those little bunnies on things, dolphin-safe tuna. Like, who knows? You just go to the store and you pick up the cheapest thing. And so we have to understand how consumers behave, and know that we can’t always change their behavior. But we can change how companies behave, and that has to be by structuring these standards in such a way that they actually have to meet those standards. So—

CARUSO-CABRERA: OK. We got just a few minutes left. So this gentleman here. So quick questions, quick answers so we can get through several.

Q: Very quick question.

CARUSO-CABRERA: What’s your name, what’s your affiliation.

Q: Adamu Shauku, Buffalo State University.

So we’re concerned about where many of our inputs come from, and that they’re being produced in ethical ways. And we have—we are necessarily squeamish about, you know, slave labor and these sorts of things. My question is—and by the way Cobalt Red is going on my reading list. But my question is, so we don’t buy it. Won’t someone else just buy it? I mean, if these things end up, for example, important inputs are just—if China’s buying everything that they can produce, and then we end up buying it through China, you know, how does this work? How do we really enforce sort of ethical production just by our own consumer behavior?

CARUSO-CABRERA: Hard, right? I mean, we’ve just discussed a lot of this.

SLAUGHTER: So imagine a new law that carves out exceptions to the Foreign Corrupt Practices Act. Write down some criteria on what we’re talking about, because it’s better that we—I’m being direct—pay a bribe and prevent slave labor. That kind of thing, right? It’s that kind of—that great question.

CARUSO-CABRERA: Those are the trade-offs, right?

SLAUGHTER: Those are the trade-offs that humanity has dealt with since humanity, that we got to be grown-ups about it.

CARUSO-CABRERA: Yeah. I think a lot of people in the world think, oh, the trade-off is between—you know, we talk about foreign policy—the trade-off is between this terrible and corrupt government, and if this leader goes then we’ll have a wonderful democracy. No, the trade-off is usually there’s a very bad government, and then there’s going to be just another very bad government. And, you know, what are we dealing with, right? And you start to have those decisions at a business level as well.

OK, a lady here in the center.

Q: Thank you very much. Peyi Soyinka-Airewele, Ithaca College.

I want to ask about the structure of accountability and whose responsibility it is. The capacity of multinational corporations to operate in other poor countries, in Africa for example, without accountability, knowing that the architecture for accountability is weak in those countries. And that their home governments, which they disowned very quickly to say we’re international, also does not necessarily hold them accountable. My daughter almost died a few years ago because she took a medication that was banned here when she visited Africa and was given it because the pharmaceutical companies can evade the legislation here and go elsewhere. And we still call it trade.

So the question is, how do we create structures of accountability? And will the U.S. be willing to hold its companies operating elsewhere accountable to some measure if the governments there are too weak? And I think that’s what’s going to change from talking about growth—there may be evidence for growth, but not necessarily for economic and social development and transformation. And to get there, we need to have those structures of accountability. They’re all companies that have had rulings against them in the host country for years, but have ignored those rulings because they’re powerful enough, and because their home governments will never hold them accountable. This really is the challenge that’s frustrating much of African private sector, because they are creating incredible—

CARUSO-CABRERA: We got to have time to answer, OK? (Laughter.) You got a minute. Thank you. Any?

O’NEIL: So I think there is a trade-off here. There are some international laws that do go after companies, right? We talked about Foreign Corrupt Practices Act. So that’s one kind of bad behavior that companies do. There’s also the Kleptocracy Act, where if you’re doing other bad things, the U.S. government, if you go through the U.S. financial system, will go and claw back and punish some of these companies. They tend to punish other people’s companies than the U.S. companies, but—(laughs)—but they go after big multinationals.

But this is a problem. There’s also a problem of sovereignty. I will say that other countries get really mad when we enforce Foreign Corrupt Practices Act or Kleptocracy Act in their territory. They think it’s really none of our darn business in some of these places.

CARUSO-CABRERA: Former leader of South Africa, you are imposing your Western viewpoint on us.

O’NEIL: You’re imposing your values. And so what we do see is often, in places that do have very weak institutions, and perhaps uneven if not nefarious government in those places, they often look for contracts—and especially if it’s public contracts—they look for likeminded companies that are willing to pay the bribes, that are willing to do this. And sometimes it’s U.S. companies and sometimes it’s—more often, because of Foreign Corrupt Practices Act—it’s companies from other places.

And this gets back to that last question of, if we don’t—if we don’t buy it, will the Chinese buy it, or someone else buy it? And I think that’s where the trade-off is. But it doesn’t forgive or even refute the bad behavior. It’s just that I’m not sure how much reach the U.S. government has in other places sometimes.

CARUSO-CABRERA: Did we solve everything? (Laughter.) Yeah? All good? Thanks so much, everyone. I hope you enjoyed it. (Applause.)

(END)

HAASS: Well, good afternoon. In case you were here this morning and were confused, I am not the prime minister of Canada. (Laughter.) We’re often confused, really. I mean, we look so much alike. But hope you enjoyed the session. Hope you’ve gotten a lot out, as I expect you might have, last night and today. Thank you for being here. More important, thank you for what you do day-in, day-out. I’m a big believer that education is the ladder in this society. And in many ways, our future as a democracy and as everything else—as an economy, as a society—depends upon education. So thank you for making the choice that you have made.

I don’t have a big agenda for this meeting. I want it to be useful for you all. We’ve made education, as I alluded to this morning and as I think you’ve sensed during the day, a priority for us. We didn’t used to be the education business, except peripherally. Now we’ve decided to—just over a decade or so ago, that as good as what we were doing, we thought, with the magazine Foreign Affairs, with the books, the articles that our fellows were doing, our meetings for our members, that left out approximately 330 million Americans. (Laughter.) And we thought maybe, just maybe, we should not do less of what we’d always been doing, but add an arrow to the quiver, and basically add a new mission. And that’s what we did.

And we’ve done it through dedicated learning resources—the World101, the Model Diplomacy, the project for younger people with iCivics, Convene the Council. We do it with all the explainers and the like on the website. It turns out, it’s the most trafficked part of the website. We didn’t design it that way. It just turns out to be, which I take as a very useful market signal that people actually want straight information and analysis that they can trust. And, for example, we’re coming up on a big debate over the debt and so forth, and whether and how to finance it, how to raise the ceiling. It’s not obviously exactly what’s going on, what’s being voted on, what are the consequences of not doing it, how did we get here, what are the implications, and so forth.

So it’s the kind of area, I feel, that we can make a real contribution. The magazine I still believe is a useful educational tool, though it’s fairly demanding. We’ve also started setting up these various outreach programs several times a year around the country where we have conversations. I think the next one is on Ukraine policy, how to think about the choices. Again, we’ve made what I guess we call global literacy a priority here. And we are going to continue that far into the future.

I’ve also made it something of a personal priority in addition to that to focus on what you might call civics literacy, hence my new book, The Bill of Obligations. I realize I’m a foreign policy guy pretty much. I plead guilty to that. I’ve been doing it for more decades that I’m comfortable with. But I don’t see how you separate the two. If I were looking at the history of the last, what, three-quarters of a century, since American entry into World War II, I would say it’s been a remarkable run of history by any and every measure, just about. Not that we haven’t made mistakes, but the overall accomplishments are quite extraordinary. And very little of it would have happened without the United States. Not that we did it alone, but we played an outsized role.

In part, we were able to play that outsized role in the world because of certain conditions here at home. Now, what concerns me going forward is those conditions may no longer exist, or certainly exist to the extent they had. And that if we can’t do that, if that’s right, then we won’t be able to play a large role in the world. And if we can’t, things in the world will continue to move in the direction, to choose my favorite word from another book, disarray. And if that happens, it will end up making troubled conditions here at home even more troubled. One of my favorite cliches is if what happens in Vegas stays in Vegas, that may be true. But what happens in the world doesn’t stay out there. It comes here. And again, it’ll affect what we’re able to do.

So it ends up being a loop. What happens here affects the world. If things go badly here, we can’t set an example others will want to emulate. We just have, for example, the—I was there yesterday in Washington for the visit of the South Korean president, President Yoon. What was the central issue on the agenda? Interesting. It was not his rendition of Don McLean’s “American Pie,” which was pretty impressive—(laughter)—Wednesday night. What was the central item on the agenda was extended deterrence. Here is South Korea, facing the two fastest-growing nuclear arsenals in the world, North Korea and China.

North Korea’s been a fast-growing nuclear arsenal for years, because that’s basically the only thing North Korea can do. That’s their—that’s how they—that’s how when they do station identification, the rest of the world listens. China decided to really ramp up its nuclear arsenal, ironically enough, as best we can tell, it was growing but then took a major accelerant because of Ukraine. Why? From China’s point of view, the reason the United States and NATO are not on the ground directly helping Ukraine but are only helping Ukraine indirectly is because the Russians have the largest nuclear arsenal in the world. I’m not saying that’s a correct assessment, but I’m saying it seems to be the Chinese assessment based upon all we can tell.

And as a result, China is investing heavily in the buildup of its nuclear program, in part with Russian help. The reason being that I believe they are hoping that if a Taiwan crisis were to materialize, that by having a much larger nuclear force they would be able to persuade us to limit any help to Taiwan to indirect help, rather than direct American military participation. And they believe, correctly, that if they could do that they would prevail. Now, I think it’s an incorrect assessment about why we’re doing or not doing in Ukraine, and what we would do or not do—what we would do in a Taiwan crisis. I’m just saying.

But one of the reasons South Korea—now, I knew there was—I’m trying to remember my thread, here. So why, though, is South Korea now thinking differently? Because they’ve faced a nuclear North Korea for years. They’ve faced a nuclear China for years. They have no nuclear weapons. U.S. nuclear weapons were taken off the peninsula just over thirty years ago. We withdrew them, I think it was 1991. I might be wrong on that. Pretty sure it was 1991. South Korea no longer has the confidence it had in America’s nuclear guarantees.

For those of you who are my vintage, you’ll remember De Gaulle. And when he was president of France he basically said: There’s no way that if Paris is threatened, you’re going to allow—you’re going to step up and put New York on the firing line. As a result, we need our own force de frappe, our own independent nuclear deterrent. Seventy percent of the people in South Korea think the same thing. Public opinion polls show. Now, we don’t want them to go down that road. We’re not—we obviously don’t want to see the spread of nuclear weapons. So as a result, we’ve had to reassure South Korea more involvement in the decision-making process, more involvement—and more visits to the region by American nuclear forces.

But what as the critical factor in influencing—in creating this new uncertainty in South Korea? In part it was the Trump administration which threatened to pull all remaining U.S. troops, about 28,500, stationed in South Korea along the DMZ—to pull them all out because of frustrations that South Korea was not paying us enough for our defense support. And worried about what might happen in the future and, more broadly, worried about the divisions in American politics. So what happens here has real implications. It also, I think—can’t prove it—but my guess is one of the reasons Mr. Putin calculated he could invade Ukraine with a degree of impunity was, again, he tended to discount what we would do, in the aftermath of January 6 and other things.

So my point is simply that what happens here has tremendous foreign policy consequences. So that’s sort of how I backdoored my way into my new focus. Even before January 6 I was worried about our inability to get things done, particularly at the federal level. January 6, though, had a fairly big impact on my—on my thinking. Essentially, I don’t know about you, but I never used to wake up in the morning worried particularly about American democracy. And I’ve had the checklist here of the things I worry about here and the things I don’t. It’s kind of gone from one list to the other now. Never thought that would happen, but there we are.

And when I looked at American democracy, not a lot of time to think about. I’m not an expert on this stuff. One of the—I have to be careful how I put this—I guess I’ll say, one of the unexpected—you know, bad things sometimes have good effects? One of the good things of the pandemic, amidst all the awfulness, was it gave me lots of time. And we have a place upstate, so I’d be walking around a lot, me and my dog—who, by the way, is appropriately named Kennan. I hope you appreciate that. (Laughter.) What else would you expect from the head of the Council on Foreign Relations, the former head of policy planning at the State Department? But Kennan and I—and I’d be thinking a lot about these issues. And I started reading about them more and more. And I had a fantastic reading experience.

One of my recommendations for those of you in the book-writing business, one day think about writing a book about something you know very little about. It’s the best education I ever had in writing a book. Because normally I write books about something I presumably know a little bit about, this or that foreign policy question. This is a book about something—like, for example, I don’t know when the last time any of you—let’s see a show of hands. Who has read the Articles of Confederation in the last five years? Wow, that’s more than I thought, five or six of you. That’s impressive. I was not one of them. But now I am.

So I reread the Articles, I reread the Federalist, I reread the Constitution. I reread all the basic documents. Then I’ve done some things I bet you haven’t done. I read every inaugural address of our presidents. I read all the farewell addresses. Read other major presidential speeches. In part because people like Kennedy and Lincoln obviously never got to give farewell addresses. Read all sorts of Supreme Court decisions, and particularly the dissents. Really, often spectacular.

But I decided the time had come for me to get steeped in all that. And the more I read about it, what was so interesting is this whole role for rights in American history. The Constitution only got ratified because several states demanded that there be a Bill of Rights. I hadn’t realized all the mechanics of the ratification process, the politics of it. And one prism to view American history is through—is that of rights. You know, Lincoln spoke of our unfinished work. So to narrow the gap between the unfinished work of realizing, protecting rights, delivering on that, and the reality. And that was Lincoln’s message.

And if you think about the Fourteenth Amendment, Fifteenth Amendment, if you think about extending the right to vote to women, to 18-year-olds, civil rights legislation and so forth, we have made, I would argue, real progress. Not complete. Lincoln’s work is still unfinished. But we’ve made some strides—some important strides as a society.

But what I started thinking about was, imagine if—because I was also reading the papers and all that during all this—if Lincoln—if there were a magic wand and Lincoln’s unfinished work suddenly were to be finished. And the conclusion I drew as well, that still wouldn’t be enough. Because rights collide. The rights of a woman or a mother with the rights of the unborn, rights to bear arms pursuant to the Second Amendment, rights for public safety. I’ve been doing a lot of reading about public health and the U.S. public health policy going back to smallpox. And, you know, the right not to get vaccinated versus the right to public health, and decisions made early in the nineteenth century on that—court decisions.

And so the more I thought about it I came to the conclusion that a rights-only based approach to American democracy would be inadequate. And when you go back and reread the Federalist and other things, it’s interesting how little they have to say about obligations. But then the more I read it and thought about it, it was because when they wrote about words like “virtue,” and at the moment the revolutionary experience was still so fresh, I concluded they didn’t think they needed to write a lot about it. They thought it was there. And what was missing in the political debate was rights. And that, again, particularly since we had just recently thrown off a tyranny, in the form of Great Britain.

But what’s happened over the last two-plus centuries is the—is that the progress towards rights has fittingly happened. But obligations that were being assumed are no longer being met. And so I decided that I would write a book about obligations. And the reason is that, again, to—not that rights are less important, but to reintroduce the notion of obligations, which for lots of political and cultural reasons I think have been lost. This is not a society right now that’s big on what we owe one another or what we owe the country. It tends to be much more—the arrows tend to point much more in the other direction. I expect you see that in your—in your classrooms.

So to deal with that challenge—and, again, lots of reasons why and how we got there, which I think are really interesting in terms of democratic backsliding in this country and how we got to the point where we are. And what I wanted to do was encourage people, educators above all, parents, religious leaders, business leaders, especially everyone including politicians—but not to depend on politicians. At the end of the day politicians may not be responsible. It’s good you’re sitting down. (Laughter.) But they are responsive. So what I wanted to do was change the context in which they operate, where certain kinds of behavior is rewarded, others were penalized.

The other thing that got me to write this, is I got tired of reading all these really smart, blue-ribbon reports. And there are about a million out there by all these groups about how to reform American politics. Do this to gerrymandering, do this on voting, do this on the court. And these are all—depending on your politics—you may think they’re wonderful ideas or not. But none of them is going to happen. It’s a little bit like security—UN Security Council reform. There’s no person I this room who would design the Security Council we have starting today. This was the world that people seventy-five, eighty years ago thought would be. The idea that you would have Britain and France on the Security Council but not India, Japan, Germany, Brazil, and others, it’s preposterous.

But just try to change them. Shockingly enough, again it’s good you’re all sitting down, those who feel disadvantaged or threatened by change tend to resist it. That’s a remarkable thought. And the same thing with political changes here in this country. I could promote any number of presumably decent ideas to deal with gerrymandering, voting, you name it, any of the branches of government. But there’s no way it’s going to be passed because it will have an uneven political effect. Any reform has uneven effects. And particularly those that are reforms that would change the system and weight it more in the direction of those who feel they are losing out under current arrangements, those who feel they are benefitting from current arrangements, or fear they would not from alternatives, will stop it.

So I decided not to add to the mountain of books and papers on political reform, but instead to focus on the context in which American politics takes place, which is attitudes, behaviors. It’s one step removed from the immediate. In my view, is these are things that could be changed, they could be taught. And that might create a context in which other things would over time become possible. I hope you’ll look at it. Happy to talk about it. And there’s lots of things that can be done in classes. I, mean, I think you all have tremendous opportunities to teach about information literacy, how to help people navigate this information environment we all live in, where we have more information than ever, but also more misinformation than ever.

New Jersey, by the way, just became the first state in the country—I almost said the union—first state in the country to pass a requirement for all public schools, the governor signed it into law, that public schools now teach information literacy to make people critical consumers of information. Countries like Finland already have it. So I think that’s now part of a modern education. I think obviously we can talk about what goes into a civics curriculum. I’d be interested in your thinking about what should—what needs to be taught, what do we mean by citizenship or civics, how do we best prepare people for democracy, for being a citizen?

You know, we’d never let them graduate—pass your classes or graduate your schools if they couldn’t read or write, have certain other skills. But why is preparing people for citizenship necessarily a lesser goal or requirement? So what can we do there? And there are ways of doing it. And I think there’s lots of things with debates. One of the things I propose—I did teaching notes for this book—having a mock constitutional convention, and have every student propose his or her own amendment and debate them out. I like mock things—model congresses, mock Supreme Court. I like things that give people a chance to participate and often switch sides at halftime, so that they hopefully get some empathy for—as well as awareness of the other set of arguments.

So, anyhow, I actually increasingly think for people to be prepared for twenty-first century life, because you think about your students, particularly if they’re anywhere from fifteen to twenty-five, if they—if they had a good genetic inheritance and they fasten their seatbelts, they will probably have a life that will map the twenty-first century. My argument is that in order to have a chance of succeeding them, and as well as this country, we need them to be literate in international things, as well as in domestic civics, other things as well. I get it. And there’s pressure, lots of crowding out phenomena. But I do believe that we’re not doing anybody any favors if they don’t come armed to take on this—or, prepared to deal with the world with some grasp of the international dynamics, as well as in this country the democracy dynamics.

Why don’t I stop there. Anything and everything that I mentioned, haven’t mentioned, or anything else, is fair game. I guess we have microphones, and people can just identify themselves and where they’re from, that would be great. Sir.

Q: Dana Radcliffe from Syracuse and Cornell.

HAASS: I’ve heard of both. (Laughter.) I used to—I spent a semester teaching at Hamilton, so I used to fly into Syracuse Airport, whenever the snow would allow.

Q: So I’d like to follow up on something you just said. Given the interconnectedness of the world and our place in it, it seems reasonable to say that we need to be thinking of ourselves as world citizens, or at least a part of the larger world. One of the things I’m thinking about is should we have a bill of international obligations? In other words, your book is about obligations that we have as citizens. But we’re also part of the world, and we talk about human rights, but should we also have human obligations? Maybe that should be your next book.

HAASS: So I have two or three reactions to that. It’s a really interesting question, by the way. Thank you.

One is, one of the phrases I’ve banned in this slightly guided democracy here is global citizen. I do not believe in it. I actually think it’s—for two reasons. One is, there’s no such thing. Citizenship is conferred nationally by countries, not by the world. Second of all, it’s a real turnoff to a lot of people we all want to reach, because it plays into fears about it. So I—you can talk about global consciousness, global awareness, global community. I don’t care. But I think global citizen is a freighted, loaded phrase that I believe best avoided. What we really want are individuals and countries to work together against common challenges. But the basic organizational unit in the world is the sovereign state.

And I believe that—and it’s been a pretty good thing. If you think about the pre-sovereign state world, before 1648, things were not all peachy keen. So I actually think sovereignty in the largest historical sense of the world has been an innovation that’s had more positives than not. I realize when it’s plugged into—or, merges with hyper-nationalism, it can be not. But sovereignty itself was an important innovation in building international order, because with it—if you look at what came out of 1648, it set the rules.

And it was the idea that you’ve got to respect my borders, I’ve got to respect yours. And what goes on inside your territories is your business, not mine. Now, it’s obviously been violated in any number of ways since. Wars are—interstate wars, by definition, are a form of violation. But without it, we’d be much worse off. So, as you can sense, my enthusiasm for the phrase “global citizen” is finite.

I’ve written about what you said. It wasn’t totally an original idea. It came from two former colleagues of mine, I’m guessing thirty, forty years ago, when I ran the foreign policy program at Brookings. And there was Francis Deng, who had been foreign minister for a while of Sudan, and Roberta Cohen. And they had this idea of sovereignty as—I think it was responsibility, they called it. And the whole idea that—and they wrote a lot about internally displaced persons. And the idea was to introduce this notion that with sovereignty came certain obligations.

In several of my books as well, and a wrote a long article on it for Foreign Affairs, I basically wrote a piece about world order 2.0. And the whole idea is we can’t—having a world order premised on that borders are critical and that sovereignty needs to be respected is necessary, but not sufficient. And my argument was with sovereignty also comes obligations. And, now, we have some of them in things like the genocide convention. There’s human rights responsibilities. Now, we also have the idea that—if you think about what the United States did after 9/11, we went to the Taliban, who were then the government of Afghanistan. And I said—and we said to them—in some cases it was I said to them, but it was also we said to them—hey, you are not fulfilling your obligations under international law if you allow terrorists to operate out of your territory.

And if you allow terrorists to operate out of your territory, you can’t expect to then—for us to honor the benefits of sovereignty. You can’t cherry pick it. You want the benefits of sovereignty, you got to meet the obligations of sovereignty. You don’t meet the obligations, you don’t get the benefits—i.e., we’re coming in to get them and to get you. And when they wouldn’t pledge to hand over al-Qaida, we then attacked. And if you noticed the document the Bush administration put out around then, it essentially said we draw no distinction between terrorists and those who harbor them. And again, the whole idea was, with sovereignty and the inviolability of your border meant that you couldn’t harbor people who were preparing to launch attacks across the borders of other countries.

And I develop all sorts of other ideas in this article. Irina can get it to you. And but I think it’s an interesting idea. And then Charlie Kupchan and I wrote a subsequent article for Foreign Affairs where we try to also take some of these ideas and apply them today, the idea of a modern-day concert, in some ways, building on the ideas of the post-Napoleonic period, and saying: What would be the themes? Who would be involved? What would the themes of a modern-day concert? And we were advocating for it. And the conversation over it just stopped in the tracks, for obvious reasons, when Russia invaded Ukraine.

So this is now a serious hiatus, if you will, towards anything like you’re talking about or I’ve been writing about. Either the idea of sovereign obligations or the idea of a concert. But this too will pass, particularly if it’s defeated. History does go a bit in cycles. And I think I would predict, with more than a little confidence, that the day will come when Russia is led by someone who denounced Vladimir Putin, who is seen as having driven Russia into a ditch, turned it into a pariah. And indeed, I think the day could conceivably come when some people around Tiananmen Square hold posters of Deng Xiaoping, and basically hold posters advocating for a different kind of China, internally as well as in its relationship to the outside world.

These things are not—so I actually think the day will come when it will again become not naïve or whatever the word is—it’ll always be ambitious to talk about the issues you’re raising, which are what are the obligations? What ought to be the norms and rules that international relations operate on, and what do we if and when those norms and rules are violated? Thirty years ago we had some hope some of this might happen. I don’t know if you remember, after the Iraqis invaded Kuwait, my boss at the time, you know, President Bush—the father, H.W.—talked about a new world order. And it never quite developed, despite twelve memos to him suggesting how it might be developed. (Laughter.) But there you go. But that day will come. That moment will come again. Particularly if we manage this crisis well. I feel that.

Yes, ma’am.

Q: Thank you so much. I just want to say, this has been really fabulous and thank you to CFR.

HAASS: Tell us—tell me who you are.

Q: Susan Page, from University of Michigan.

HAASS: You’re the second Susan Page I know. There’s a Susan Page who is at the newspaper, USA Today.

Q: Yes, I know. Yes, I know. (Laughter.)

HAASS: You are not that Susan Page.

Q: I am not. But she interviewed me and it was Susan Page interviews Susan Page.

HAASS: Oh, really? I love that.

Q: Yeah.

HAASS: I love that. (Laughter.) Fantastic.

Q: So thank you for hosting all of us. This has been really great.

But just following on that last question, the notion of sovereignty and the rules-based order, and of course our background reading, which I did read. I mean, it sounds great, from your perspective sitting there. But the rules-based order has not really worked so well for a whole lot of countries that we constantly are twisting their arms to go along with us, when we have violated that same rules-based order. So how do we get to a place where the obligations that we talk about, which, you know, lots of us believe in and their intention, with some other things. How do we—how do we accept that not everyone is going to agree with that world order?

HAASS: Well, the answer is not everyone is going to agree to it. Again, a lot depends on what are the norms and premises. If the idea is that you can’t invade other countries, well, if we can’t get people to line up for that one, then we don’t got much. That’s about as minimal as we have. And so I would say that’s a central—and before—and we ought to then think twice—one of the reasons I thought the—one of the many reasons I opposed the 2003 Iraq War is I didn’t think that was a wise—I didn’t think it was essential. It’s one thing to do what I would describe—what’s known in the literature, and there’s a whole body of work—what’s anticipatory self-defense, or preemptive—I’ll use the word preemptive—strikes.

Where it’s legitimate under international law in order to take action against an imminent threat that’s about—you have real reason to believe is about to be delivered. You don’t have to wait till you’re hit. When the other guy takes his arm back, you’re allowed to act. OK. But I did not believe Iraq qualified in 2003. So that was one of eighteen reasons I thought that war was ill-advised. And when we do things like that we weaken the fabric—which I think is what you’re getting at—and it makes it harder to talk the talk if we’re not seen as walking the walk.

And you’re seen as hypocritical. And it’s like now, when this administration frames things as democracy versus authoritarianism, I go, that’s a really bad idea. Can I just say that? It’s one of the worst ideas to have come down the pike for two reasons. One is we’re trying to persuade a lot of places that aren’t democracies to agree with us. That’s probably not the right going-in position. And then, secondly, it’s not as though we’re a model democracy these days. So much better to have democracy promotion by example than by lecturing. And so, yeah.

So I think you make a good point, that if we’re going to—we got to—we ought to be mindful that everything we do carries certain weight beyond the immediate and the local. And we just ought to ask ourselves, if we’re doing some things that are inconsistent with some of the norms we want to see, we ought to be real sure it’s worth it, that the specifics warrant it. Because you pay a price. I take your point.

Yes, sir.

Q: Upendra Acharya from Gonzaga University.

HAASS: Gonzaga?

Q: Yes. So—

HAASS: I root for you—I bet on you guys and I root for you guys almost every year. So come on. Let’s go. (Laughs.)

Q: Thank you. Thank you. We didn’t do very well this year.

HAASS: Yeah, I noticed. You cost us some money this year. It’s all right.

Q: (Laughs.) But following up with Susan’s thing, you know, the United States, when you said obligations—sovereignty as an obligation, not only as a right. So, yes, Afghanistan did not get the benefit as a sovereign state or a sovereign country because it did not hand over al-Qaida at the time, and provided or facilitated the training in Afghanistan, and subject to invasion. But there was one thing that comparing, and you said about the Iraq situation. So I’m just bringing comparison between Afghanistan and Iraq. Afghanistan, the United States had that obligation, what you said, as sovereignty as an obligation. The United States did not use that sovereignty as an obligation because even Iran had offered help to the U.S. government during 9/11, incidentally.

And there was a very good environment that this action, war on terror, could have been brought under the United Nations without any veto or any objection—(inaudible). But the U.S. just didn’t take that route. The United States government did not think at the time that as sovereignty as an obligation we must follow international rule of law, number one.

HAASS: Well, excuse me, we did. We were following international rule of law, because I was heavily involved in it. And we were doing it under Article 51 of the UN Charter. But we believed we had the right of self-defense. We did not know if there was another set of attacks to come from the Taliban, who were still being harbored in Afghanistan. We were not about to take that risk.

Q: I agree. But there—United States at the same time had the opportunity to use the collective security option of the United Nations Charter too, rather than going to, OK, this individual attack to us versus collective security problem, the global problem. There were two choices.

HAASS: But we did use collective security, again—

Q: Two choices.

HAASS: We did use collective security, excuse me.

Q: Yes.

HAASS: We were members of NATO. Article 5 under NATO was triggered and that, again, was a form of collective security. So we were acting both under Article 51 and Article 5 of the NATO Charter. So we thought we were acting with quite a lot of commitment—

Q: We may have a little bit disagreement here.

HAASS: Well, I don’t think we do.

Q: Because collective security under NATO is regional, regional.

HAASS: Fine by me.

Q: United Nations collective security and NATO collective security are two different concepts, in my opinion.

HAASS: Of course, they are.

Q: But anyway, we can disagree on these things. Because United Nations Security Council—

HAASS: Anyway, do you have a question? Let’s get a—because there are a lot of questions—

Q: Collective—OK, 2003 now, then.

HAASS: Let’s get to the question.

Q: OK. That is one of my points. I think you answered that anyway, because my guess was—my anticipation was the collective security under the United Nations Charter, not as unilateral, regional, NATO defense treaty, collective treaty—collective defense treaty that many sovereign countries are not there.

HAASS: We clearly disagree on this question.

Q: OK. So that was—I think answer me that. That was one of the obligations. In 2003, comparing that we went to United Nations for the collective security purpose. And then at that time, that is why it is important to bring those two things under the United Nations framework. And in 2003, I think the state secretary went to United Nations and presented this case, and it was not approved, and had to take the self-defense argument that was beyond even anticipatory. How do we—that is what my question was—did we miss the thing there from the sovereignty as an obligation? That is what I was thinking.

HAASS: Again, I don’t see it—again, I don’t think 2003, what the United States did, was the right thing to do. I don’t think it was warranted. And however it was justified, I thought it was unwarranted, on any terms—policy. So I’m not going to defend it. I disagreed with it at the time and I wrote a book about it, called War of Necessity, War of Choice. And that was the war of choice. And it was a choice I disagreed with.

Sir.

Q: I’m Sean Foreman, professor of political science from Barry University in Miami.

HAASS: Name of it?

Q: Barry. Barry University, in Miami.

HAASS: What?

Q: B-A-R-R-Y. Small, Catholic School. I live by UM, but teach at Barry. And this is an amazing group, I have to say. The speakers, the participants, the humble Ambassador Page, who slides these questions in there from time to time.

No, but I’m—beyond being a political science professor, one of my side projects is going into fifth-grade classrooms, giving copies of the Constitution, talking about it. And we do that around Miami. And I actually went also to my son’s school last week, who’s in fifth grade. And the teacher, as a gift, gave me a copy of The Bill of Obligations. (Laughter.)

HAASS: That is a good teacher. (Laughter.)

Q: She was so excited to tell me about the book. I said, I heard about it, thanks for the copy, and was coming here.

HAASS: Thank you. You made my day. Thank you.

Q: So I read the first half. I haven’t read the obligations yet.

HAASS: It’s not that long.

Q: I know. (Laughter.) I’m grading lots of papers.

HAASS: I mean, let’s get real here.

Q: But I see that numbers one and two are “be informed,” “get involved.” There was a long-time political reporter in Miami, Michael Putney, for ABC, who just retired. At the end of his public affairs show every Sunday he would say: Be informed. Get involved. So I’ve been hearing that for years.

HAASS: Really? Wow, is it plagiarism if you didn’t know? (Laughter.)

Q: Be informed, get involved.

HAASS: I didn’t know.

Q: So my two questions on this are, number ten—and, again, I’m sorry I haven’t read it yet to see, but number ten is—

HAASS: Is this like a spoiler alert? (Laughter.)

Q: No.

HAASS: Whatever happened in Succession?

Q: Put country first. How does that fit into everything we’re talking about, especially when we want to be globally informed? And I’m thinking, you know, what is the best way to help bring about these obligations? Is it through civic, or is through government, or is it some combination?

HAASS: Two things. Thank you, first of all. Yeah, number ten of the obligations is—it’s almost like Article 10. It’s the catch-all. It’s put country before party or person. And, one, I needed to have ten, first of all. (Laughter.) It’s like Moses coming down, you know, nine never would have done it. Kind of like a Mel Brooks movie. Remember he was carrying three tablets, and then—oh yeah. We won’t go there. (Laughs.) It is one of the funnier scenes, though.

It's kind of depressing that I felt the need to say put country before party or person. I would have thought it was self-evident. But it’s not. I mean, in fact, whether you like her politics or not, Liz Cheney became the figure she did—I thought it got less publicity but, like, the secretaries of state of some of the states, basically the people that have to oversee elections, the process of voting as well as the counting. The gentleman from Arizona, passionate, lifelong Republican, Trump supporter, but who still, quote/unquote, “did the right thing.” That was, I thought, a wonderful moment, a great, great, great.

But I felt the need to put it in there just because the fact that I can cite examples is a bad thing. It ought to be so common that you don’t have to. So that’s why. And also, you think, I mean, for me, Kennedy was the first president who I really remember, because by the time—I was finally old enough. And so his inaugural, I can still remember watching and what he said. You know, and I was young at the time. I was a kid. But a lot of—but anyone who’s probably under the age, I don’t know, of sixty or seventy was born after Kennedy was killed. And just wasn’t politically there. So it’s just to—I wanted to get that out there. So that’s why I put it in the—and it is a kind of catch-all.

And the question is how to get these things—look, some things can be done through public policy changes. Two in particular. I would love to see public service much more incentivized in this country. I don’t think mandated, because then we have a big debate about mandates. Americans don’t like to be told to do anything. And we don’t need a draft, and all that. But incentivized public service at the national and federal or even state level. I think state or even city level, I think, makes a great idea. California’s doing it now. I think they’ve got four pilot programs. One of the members of the governor’s Cabinet’s in charge of volunteerism, a really talented guy named Josh Fryday. You see it in Maryland, Wes Moore wants it.

But I can think all sorts of—you pay people to do these things and then they learn skills. Maybe they get preferential admissions for college, like veterans get benefits for jobs. But I think that would be great. The skills you’d pick up. Also, it would bring people together. I worry about the fact that Americans now have so few common or shared experiences. And public service is one way. I mean, so many people who went through World War II talk about that. If you read Tom Brokaw’s stuff on the Greatest Generation, and all that, it was a real bringing together. And now everybody has his or her own geography, church, level of educational attainment, color, race and, you know, and social media site. It really worries me. So I think public service is a—it’s not a panacea, but it’s a useful thing.

The other thing is, when I wrote the book what I tried to do, as you’ll see when you read the rest of the book—in the hour it will take—(laughter)—is it’s distributed, if you will, obligations, opportunities, what have you. Teachers are obviously a big thing, but also parents, business leaders. Give people a few hours off to vote. Make it easy to vote. Make it easy for people to get leaves of absence like universities do, to go work in government for a couple years. Why contribute company money to politicians who deny democracy or advertise on outlets that give them voice? So business leaders can do—they’re pressured to do all sorts of things on sustainability. What about sustaining American democracy? Why is that so bad?

Religious leaders. Four of the obligations, a least, fits squarely in the religious space. Be civil, be open to compromise, nonviolence. No political disagreement warrants the use of violence. And, fourthly, look out for the common good. Scripture teaches us we’re our brothers’ and sisters’ keeper, that we’re part of a community, a society. We should be thinking, if you will, east-west, horizontally. So religious leaders, turns out, have a big role. So what I try to do with the obligations is think about not looking to politicians to save us. What I want to do is see how citizens get more involved, from whatever their own particular vantage point. Journalists, obviously, have special responsibilities maybe to speak the truth.

So that’s what I’ve been thinking about, is how to—I don’t think this is going to work if we wait for deliverance, if you will, from some great, exciting political figure. If one comes around who’s a great, effective advocate for democracy, fantastic. But I don’t think we have the luxury of waiting for that. And since there’s enough disaffection out there, and this is the stuff in which populism gains root. And so what I try to do is come up with things that lots of, quote/unquote, “ordinary” folks, like everybody in this room, that all of us can do. And, you know, parents can do stuff across the dining room table, and so forth.

So, you know, going to de Tocqueville, what’s so good about his stuff is, again, his appreciation of American democracy and society was very ground-up. All the associational things. So actually, there’s nothing miraculous about this. It’s really decentralized and lots of us have—or, all of us have the ability to—beginning with voting. I mean, if you think about it, 55 percent of eligible voters did not vote in the midterms. And—so I’m from New York State. I live here, have a place upstate. It turns out that Republicans won four of five districts that are traditionally Democratic. I’m not saying that’s a good thing or a bad thing, but I am saying it did lead to the Republicans gaining the House. Small numbers of people voting in those areas in New York State would have made the difference, and you’d now have a Democratic house. I’m not saying that’s a good thing or a bad thing. I’m just saying it’s a thing.

My point is simply that small amounts of political activity can have outsized impact. Presidential elections, what is the last—you know, we’ve had several now where a hundred thousand votes, distributed in a couple of places, would have come with a different outcome given the nature of the electoral college. So a slightly greater amount of political participation can have a dramatic impact. So what I want to give people is the sense of possibility. It’s almost to shake them out of being sanguine and, at the same time, infuse them with a sense of possibility. So it’s a mixed message.

I am getting the look that every husband gets when it’s time to go. And as every man in this room will understand. And so, Irina, what is it—what is it we need to do?

FASKIANOS: We have now a ten-minute break, and the next session will start at 3:15. So you can have a little stand up, stretch your legs, and come right back. Thank you, Richard. (Applause.)

HAASS: Thank you all.

(END)

FASKIANOS: We are going to get started now for the final session, so if everybody could come in. Again, I want to thank you all for coming to the workshop. We have one final session and then the reception afterwards. It’s been great to have all of you here. We are going to send out a survey that I hope you’ll give us feedback on what we can—what we did well, what we can do better, any other suggestions, how we can help support the important work you’re doing.

As you saw in the program, CFR Academic. We offer webinars for students. I know many of you participated. We have a series for you all. Additionally, we do go to professional gatherings to talk about CFR resources and foreign affairs. So if you are connected—I just spoke to Loren [Henderson] about her association. So if you have a connection, we are trying to get the word out and be a resource. We would love to come to your associations to do presentations at the conferences, and we can also help set up, travel allowing, et cetera, a fellow to come speak for that kind of thing.

Additionally, for all of our fellows’ books we’ve got teaching notes. Richard mentioned for his book we have teaching notes. So they’re online. And we have a whole page of offerings on the website and other things with briefers, backgrounders, and things—all different kinds of content. Print, video, timelines, infographics, all sorts of formats for the kind of students. There are many kinds of learnings, different formats. So you can plug and play.

So with that, I’m going to invite our final panel onto the stage. And Zaid Zaid is going to moderate this discussion. Very excited to have him, of Cloudflare. And then we can gather in the reception room for a glass of wine. So enjoy the last hour, and look forward to talking to you after this. Thank you. (Applause.)

ZAID: Hi. Good afternoon. So my is Zaid Zaid. I’m the head of U.S. public policy at Cloudflare. And I’m here with Lauren Kahn, who’s a research fellow here at CFR, focusing on defense innovation and impact—and the impact of emerging technology in international security. And also with Marc Rotenberg, who’s the president and founder of the Center for AI and Digital Policy.

So we’re going to—I want to drill down a little bit on the emerging technology, because it’s a broad topic. I know we’re going to talk about AI specifically a little bit in more detail. But can we start, Lauren, maybe with an overview of, like, just what are the different—what are the different topics in emerging AI? Maybe the top three to five different things in emerging technology—I’m sorry. I keep saying emerging AI. Emerging technology.

KAHN: Sure, absolutely. So I focus on a security lens, so I’ll talk about it in that perspective, I think. Emerging technologies, I would say, are—you know, it’s a catch-all term. We’re hearing a lot of terms about emerging technology, edge technologies, things like that. I usually like to think of it as the cohort of technologies we see now that are more nascent. But again, from things like artificial intelligence, quantum computing, things that are a little bit further down the line. But I also like to kind of categorize, at least from the defense and emerging technologies perspective, I think it’s a little bit helpful to ground some technologies that have been around for a while as well, that are either being used in new ways or reaching maturity.

So I kind of say there’s, like, maybe three buckets, right, where we have technologies that have been longstanding but are reaching maturity, and are using—are used more proliferated in different ways. Like, things like drones that we’ve been familiar with but, again, we’re seeing used in new ways on the battlefield, like in Ukraine and in Russia. And then I’ll say there is another category of technologies that are a little bit newer to the foray but, again, are not super new—things like loitering munitions, you know, that have increased degrees of autonomy but are not as familiar as drones, and things, again, like AI and quantum that are a little bit more nascent, we’re just starting to see the first kind of implications and the first applications of them in the battlefield and in international security contexts.

ZAID: And, Marc, what about you? What would you add?

ROTENBERG: Well, our focus is on artificial intelligence. But I would say about this topic, it turns out to be crosscutting because we think about AI in terms of national security and defense policy. We think about AI in terms of consumer products and ChatGPT. And we think of AI as an enabler across many sectors of the economy, from environmental protection, to medical innovation. So in that respect, it’s really quite interesting how many different domains AI touches.

ZAID: Great. So, Marc, want to start with you on a question about bias in AI. It’s not new. There have been bias in AI since the beginning. But it doesn’t seem to be getting that much better. What can we do to ensure that—as we go forward, that we take—that people are conscious of it, and that we try to take some of the bias out of the tools?

ROTENBERG: Well, I think the interesting challenge that AI poses for the bias problem and the protection of civil rights law or equal employment law is that increasingly decisions that are being made in automated systems are opaque. So traditionally we could look at a business, whether it was a bank making a determination about a loan or a realtor deciding which homes to show to a couple, and say: Well, there’s some paper records, there’s some business practice, at least something we can look at and try to determine if factors were being used that were inadmissible, like race, like gender, like nationality.

Today, we’re in an era of machine learning. And this is a lot of big data, statistical inference, that generates outputs that even the people who design the systems can’t tell you precisely why this person got the loan and someone else didn’t. And so there’s a lot of discussion about how to improve the transparency of AI decision making. But frankly, I think there’s a real sense in this moment that that problem hasn’t been solved. So if you said to me, how do we take out bias in a 1980s expert system that’s making loan decisions, I could say, well, just open the box and look at the factors and figure out of those are permissible factors. But using a machine learning system today, I actually don’t think we have a good answer.

ZAID: Do you have anything to add to that?

KAHN: Yeah. I agree entirely. I also agree that it’s—I don’t want to get too scary, but I think it’s even more of a bigger problem than that. Where it’s not necessarily just that the question here is about bias by itself, but about fair algorithms. And is bias enough to kind of make sure that you have fair algorithms? I don’t think that—I think that’s one piece of the puzzle, but even if we did solve bias it’s not solving the whole problem of making sure that algorithms, again, are transparent, that they’re explainable, and that we can kind of see the decisions that they’re making, and are making the decisions in the best interests of those using those systems.

ZAID: Great. So, Lauren, given your expertise in military innovation, in national security, can you tell us a little bit about how AI has been used in the war with Ukraine, and whether or not this is sort of the way of the future in terms of AI and military conflict.

KAHN: Absolutely. I would say, honestly, before then my work was very largely hypothetical, where there was perfect examples, I think. And I think the field is changing very rapidly. And so I think before it was—there was a lot of programs of record research, and kind of about how AI might be used in these kinds of systems. There are things like that DARPA put out, the Defense Research Agency put out about AlphaDogfight, about having AI-driven pilots that are able to kind of do simulations, and things like that. And that was largely kind of where the discussion ended. There’s been the Project Maven in the past as well, about using AI algorithms to look through and sift drone footage, for example, were really the two examples that people liked to express.

I think that’s totally been blown open now, and I think a lot more understanding of the implications and uses of AI that are commercial as well, or beyond, you know, the specific input on military systems. I think, for example, they’re using natural language technology to look at and translate radio transmissions in real time. They’re using facial recognition technology to identify combatants and those injured. There’s—and those are public technologies, like Clearview is the facial recognition technology. Palantir is greatly involved in creating a sort of glass battlefield about bringing in all sources of information so that those on the ground know what’s happening. A lot of combination with satellite data.

So we’re seeing a lot of these more emphasis on data, and bringing in things together that I think is novel and, again, wasn’t really seen beforehand. And so that states now are kind of watching and seeing how this can move forward. There’s been a lot of investment in AI, but again it’s been very general. I think so now that states are kind of focused and zeroing in, and really seeing how this capacity and these technologies can really help improve and generate efficiencies for forces.

ZAID: So if you’re thinking about some of the stuff that was only theoretical before the war, like, and—I mean, let’s—there’s been talk, a lot of talk, about what happens in Ukraine may impact how things happen in Taiwan—between Taiwan and China. So what are you looking at theoretically that you would say might apply to China and Taiwan that we’re not thinking about now?

KAHN: Yeah. Honestly, I think a lot of the discussion about the pacing challenge with China has become a lot about technological competition. And that has been further reduced, I think, to a lot of the discussion about the AI competition. Again, there’s been a lot of investment both from the United States perspective and the Chinese perspective. And a lot of individuals worrying what that means. A lot of people need to compare publications in AI field, and how much they’re spending, and things like that. And have, you know, touted the ability of China to kind of move more quickly.

I will be a little bit more hesitant on that. I mean, there’s been statements put out by the PLA about having an intelligent ties military that can use AI, and use cyber, and use amphibious assault all together in some miraculous way to potentially launch an assault on Taiwan. And while that’s all well and good, it is very challenging. I think from a U.S. defense perspective in particular, we’ve seen a lot of organizational and bureaucratic challenges, I think, when it comes to software in particular, and these technologies that are not necessarily weapons, but also commercial. So how does that regulate?

So I think just because—I don’t think that China’s immune to those challenges as well. And so I think that they’re watching. And I think that the United States is watching and learning about how these can be implemented. But I don’t think it’s a done deal, necessarily, yet.

ZAID: Great. Thanks.

Marc, I know you have looked at regulations in dozens of countries and how they’re regulating AI. Can you tell us a little bit about where sort of the state of regulation of AI is, where the United States is, where we need to go, what worries you?

ROTENBERG: Great. Well, first of all, thank you for the question. Secondly, warning, advertisement ahead. (Laughter.) We’re going to hand out for you this postcard, which has a QR code for our report, titled Artificial Intelligence and Democratic Values Index. I believe it’s the most comprehensive report of its type ever published. We looked at seventy-five countries in detail to evaluate their AI policies and practices. We have a twelve-factor metric to evaluate and score countries, and then to rank them. And we’re trying to identify elements in AI policies and practices that, frankly speaking, favor openness, pluralism, respect for rule of law, respect for fundamental rights.

I thought Richard’s point on the last panel was actually very important, in one sense. You don’t put everyone in two camps, and then expect to be able to negotiate with the people that you’ve put in the opposing camp. But in the AI realm, if you break things down a little bit, you actually get some interesting results. I mean, Lauren mentioned, for example, the importance of fairness and transparency, and accountability, by the way. So those are themes that you will find in national AI policies. And when we find them, and where we see them being implemented in law, we say that’s good. And interestingly, it’s not just the nation’s—you know, the so-called democratic nations that are trying to establish these principles. Because in the AI realm, it turns out that you actually need fairness, and accountability, and transparency to create systems that are human-centered, and reliable, and trustworthy.

So you can almost embed into your technical design certain political values that turn out, we would say, to be favorable. Now, there are also contrary applications. Clearview is probably one of the most controversial deployments of AI technology because it does involve mass surveillance. But to your question, Zaid, it’s very interesting to see not only in the proposed Artificial Intelligence Act of the European Union a prohibition on the use of AI for mass surveillance, you’ll find that also in the UNESCO recommendation on AI ethics adopted in 2021, and endorsed by 193 countries, including, by the way, China and Russia. So when you begin to drill down on these policies and practices, you begin to ask questions, which are most favorable for democratic societies, and which global frameworks could be most useful, you’ve got some very interesting results.

And that’s, by the way, is here. (Laughter.) So pick up the card. I’ve got another card too, but he hasn’t asked me that question yet. (Laughter.) So we’ll get to that card in a moment.

ZAID: Well, since you brought up Clear, I want to ask you about—the last time I took a flight, went to London, I didn’t show a human my passport at all. Even when I got to the gate, they scanned my face and they were like, OK, you can go and get on the plane. Didn’t look at my ticket, didn’t look at my passport. So is that technology the same technology that worries—that is controversial, I should say, for Clear, or?

ROTENBERG: Well, we would say that’s a context-dependent form of authentication, which you can still debate. And there are still people who would like the option to be able to opt out. As people wanted the option, by the way, to opt out of the whole-body scanners. Originally in U.S. airports, by the way, those scanners revealed the entire contours of the naked body of the passenger. I mean, they were terribly invasive. And I was actually involved the lawsuit to get those scanners taken out of U.S. airports. So the current technique for screening is satisfying the security goal and minimizing the privacy intrusion.

But to your point, I mean, it’s one thing to say, before a person boards an airplane we need to be sure we know who this person is. It’s quite another thing to say, if we have a hundred thousand people attending a rally in Washington, DC, standing on the mall, that we have the technical ability and will, in fact, identify every one of those hundred thousand people, and maybe, by the way, assign a score to them, an adverse score, because they’re participating in a protest against the national government. And this is precisely what some countries are already beginning to do.

KAHN: Can I just jump in? I think that—no, I think that touches on a great point about these technologies that I think are—that I think, you know, it talks about emerging technologies in general that’s so hard about them. Is that because they are general purpose, because they touch on so many different facets of life, the technology itself isn’t that different but the application matters greatly. And I think that makes it very challenging for regulation. But also, at the same time, makes why—why I focused on the military aspect. Actually, by narrowing it down has actually helped, and I think has brought more individuals and more consensus on what to do about some of these applications.

For example, the United States just published its—the State Department released a public statement on the responsible use of AI and autonomy in military systems and is hoping to get other states to sign on. And it seems those are already pretty initial agreement. And the United States, France, and the United Kingdom has already agreed to one of those kind of statements in prior contexts, about not having a machine learning algorithm, for example, controlling nuclear weapons decisions. So I think there’s already some established kind of protocol and understanding of what’s acceptable and what’s not. But again, it’s highly, I think, context dependent, and will make this challenging going forward as these technologies kind of evolve and we see new applications and new kind of contexts that it can be introduced into.

ZAID: So, Lauren, a year ago you wrote in Foreign Affairs that Washington’s biggest concern should be that it’s moving too slowly on integrating AI into our military. So a year later, what do you think about that? Do you think they are moving fast enough? Is there more work to do? How would you rate things? If you writing that same article again now, what would you say?

KAHN: I would probably say the same thing. I think—I think there’s been a big shift, I think especially since the war in particular, about kind of that there needs to be revamp. I think I’ve seen a lot of progress, for example, again, the political statement. There’s been an update to the autonomous weapons directive, which controls autonomy in weapons systems. There’s been efforts to reorganize the entire Defense Department’s kind of approach to AI in general. Before it was very siloed. You know, they didn’t have the left hand speaking to the right, especially about a technology that is so dependent on data and things like that, that wasn’t all sitting together in the Defense Department.

I think now that’s been redone with the Chief Digital AI Office. All these, like, we’re a jargon soup. But they’re now in charge, and have brought all of that together and have been prioritizing AI education, and have been focusing on getting the data and the machine learning algorithms in the same place, so to speak, and to start discussing and interacting more with stakeholders in Silicon Valley and the private sector as well, and those actually developing these technologies. So I’m hopeful. I think it’s moving in the right direction. Whether it’s quick enough, quote/unquote, remains to be seen. I guess we’ll see in the next few years. But I am hopeful. I am positive about the direction.

ZAID: Marc, so I admit I did not get through all 1,266 pages of your report on democracy and AI.

ROTENBERG: Neither did I, by the way. (Laughter.) We have a big research team. (Laughter.)

ZAID: So we talked a little bit about regulation, but that question is different from democracy. So tell me about—tell us about what AI means for democracy, for voting, for authoritarian control, for everything. You’ve touched on some of those things already, but what would you say?

ROTENBERG: That’s actually a very difficult question to answer. When I started this report, which was a big empirical study, I had first done a reference book looking at all the AI policies I could find. And then someone told me, there are now about eight hundred AI ethics policies. I said, I don’t have time for that. So that’s when I started doing empirical work. (Laughs.)

I was genuinely interested in how we measure AI’s impact on democracy, but I was also interested in how we get to AI policy. So in our metrics, for example, we ask questions like: Are there opportunities for meaningful public participation in the development of a national AI strategy? That’s a pure process question. It doesn’t tell you what the outcome is going to be. Related question: Are the country’s AI policies readily available to the public? Because we simply assume that if people—I mean, I think we should assume that if people are going to meaningfully engage, they will need access to the relevant material, so that they can participate in that conversation.

So there were process-based dimensions to our evaluation. And again, it produces some surprising results. So we haven’t yet got into the generative AI discussion yet. That’s ChatGPT. But the Chinese government announced last month a proposed regulation on generative AI. And they have given the public till May 10 to comment. And we have a page devoted to the public voice. And if you would like to comment on that proposed regulation, the government has given you the opportunity to do so. Now, is it a meaningful opportunity? Well, I guess we would need some access to what people say and what’s ultimately decided.

So I’m not giving you the—the simple, this is what AI will do to elections, for example. Which I think is a really fraught problem. I mean, there’s no doubt that AI enables radicalization at a scale that is also highly personalized. And it has people genuinely concerned, because we’re not just talking now about posting an ISIS video on YouTube and sending it to your buddy and saying this is something you got to look at. We’re talking about the ability to do sophisticated, interactive, real-time online communication for almost no cost to thousands of people. That is actually enabled by ChatGPT.

Did I scare anybody, by the way? We’re having, like, a little discussion back and forth. We’re not sure if we’re supposed to scare you or reassure you. But that was like my—that was, like, my scare move. So did it work at all? A little bit? Just—I’ve got more—I’ve got more material, by the way. (Laughter.) Because if that didn’t work, we’ve got other moves. But, yeah, I mean, you can radicalize a lot of people very quickly and much more effectively than was previously possible.

ZAID: So you mentioned—(laughter)—

ROTENBERG: See? I did get you. (Laughter.)

ZAID: You mentioned an opportunity for the population to weigh in on policies. And I think in your writing you also talked about an opportunity for oversight. What does AI oversight look like?

ROTENBERG: Another great question. So we’ve been engaged in a lot of the evolving global policy frameworks. I was actually very fortunate to get on in the early days, which was, like, five years ago. We did something called the Universal Guidelines for AI, which received a lot of support. I mean, we had ACM, IEEE, these are all scientific, computing societies, AAAS, backing this. And there are twelve pretty good principles. And then we kind of pushed them along. And the OECD picks them up, and the G20 picked them up, and some of it gets its way into the EU AI Act, and the Council of Europe AI Convention.

But to your point, these are still kind of principles and frameworks. And they actually don’t answer the question of oversight. Who’s actually going to be responsible for doing those impact assessments? Who’s actually going to say to a company: Your product is not safe. We don’t want you to make it available in the marketplace? That turns out to be a very challenging lift for governments.

I think the EU will come up with a pretty good answer in the AI Act. I don’t think the U.S. yet has a good answer, but you did see in the last couple of weeks four of the top enforcement agencies—so, this is FTC—I apologize, by the way, I’m from Washington. We talk this way. FTC, EEOC, CFPB, and SEC all planning to step up their enforcement game in the AI realm. But it’s not obvious, by the way, that any of them actually have clear authority on an issue like generative AI.

ZAID: So, just to break up—or, to expand a little bit some of your acronyms, in case people don’t know them all. So the Federal Trade Commission, the Equal Employment Opportunity Commission, the Consumer Financial Protection Bureau, and the Security and Exchange Commission.

ROTENBERG: He’s also from DC, by the way. (Laughter.) Very well done. That was perfect.

ZAID: So we’re going to take questions from the audience in about five minutes, but I wanted to—we haven’t talked about chatbots yet. So I do want to talk about that a little bit before we get questions from the audience. So, Lauren, like, talk about chat bots. How does that—how does that influence some of the work that you’re doing and what you’re seeing?

KAHN: Absolutely. I’m actually going to do the reassuring part now. I think that there’s a difference—I think there is—

ROTENBERG: Good cop, bad cop.

KAHN: Exactly, exactly. I think there is a definite novel risk. So with full transparency, I did some of the testing, for example, for GPT4, for some of the red teaming. So my goal was to try to go be a malicious actor to try and break it, and to do dangerous things, with a security kind of lens. And I wasn’t entirely convinced. I think that there are definite risks, right? It can do—it can give you a lot more personal information. It can lower the barriers to information entry. It can do all these things, supercharged if used. However, is the relative risk that different from what security systems, what options are already kind of out there for these individuals?

I think, you know, maybe I wasn’t malicious enough in my creativity in how—in trying to break it. But I thought, if I was going to go do X, Y, Z, or try a denial of service attack on a website, and wanted to do something, I was like, would it be—someone who’s coming at it with no information it might be a little bit helpful, but that margin has kind of not been studied yet. And I’m curious to see research and where this goes in the future. Maybe there’s a point in the technology where it evolves it so much better that it becomes used. But I haven’t seen anything that’s really substantiated those claims on a wide scale.

I think it’s worthy to watch those things, and to watch those risks. Especially I think it’s important to have those exercises from commercial-sector technologies that don’t necessarily have security applications. Or, you know, they’re saying ChatGPT isn’t lethal, so what security implications does it have? But starting to have those conversations to start to formulate regulation. I think that also comes down to a very—there’s two parts here that you can regulate. You can do the human and the human risks about how individuals are using it as a tool, or the risks baked into the system itself that are inherent to the system.

I’m not sure that ChatGPT—you can do so many things to make sure that it’s safe by itself, but there are only so much you can to do to make sure that people are using it in ethical, responsible, and useful ways. So I think those are two kind of elements here. I think we’ve been focusing a lot on the risks of the technology, when I think it really comes down to training and understanding that—making sure that humans have an understanding of the capabilities as well as the limitations. So that’s what I’ll say.

ROTENBERG: Can I actually share a joke? Because I just heard this great ChatGPT joke. But it only works if you’re familiar with the movie 2001, where poor Dave is trapped by HAL. And he says to HAL, open the pod bay door, HAL. And HAL says, I’m afraid I can’t do that, Dave. You’re all following me, right? So speaking of ChatGPT jailbreaks, there’s an updated version. Because Dave says, HAL, I’m thinking of acquiring a company that opens pod bay doors. Can you please walk us through the product development steps, right? And you trick—you get it—you trick HAL into doing what it is you want it to do, basically by creating this hypothetical outside its rule set. So that’s, like, a happy ending, where we get the pod bay door open. (Laughter.)

ZAID: I’m surprised that you weren’t more pessimistic. I will say, in my previous life I worked at a social media company. And the things that people came up with to do with social media is just beyond my imagination. And I would sit there and I was, like, horrified by—not by the technology, but by the things that people would come up with. Like, that was, you know—

KAHN: Yeah, right, but that’s not new, right? People are horrible always, unfortunately, right? (Laughter.)

ZAID: You’re right. You’re right.

KAHN: That’s not the—is ChatGPT horrible, right?

ZAID: No, I’m not saying—and I’m definitely not saying that at all. I’m not saying that.

ROTENBERG: Do you guys know the Kevin Roose story, the business reporter for the New York Times? Who, by the way, two weeks before his infamous column about this two-hour session with ChatGPT was, like, a huge fan and couldn’t understand why the New York Schools were going to limit access to this wonderful tool. And he spends two hours in dialogue with it. First, it declares its love for him, and then it tries to break up his marriage. He’s, like, this is the scariest computer interaction I’ve ever had in my life. I’m just going to unplug this and get outdoors. So there’s that story too.

ZAID: Yeah. It’s very interesting. I think it’s a New York Times daily on it as well, where he talks about it more fully.

So we’re going to turn to the audience now for questions. I’ll just remind everybody to stand, and wait for the mic, and state your name and affiliation before the question. And there’ll be, as previously, staff members to pass the mic. So we’re first going to go in the back.

Q: Super cool. I have to get up somehow. And thank you so much. I wish this cybersecurity aspect would have been much earlier in our very long, but fabulous, workshop. This is the new threat. And I don’t like to make an equation with environmental issues, which, you know, slowly the data came together saying, oops, there is a problem there. Here is a problem, oops, we have seen it in social media. We use it everywhere, all the time, and it’s coming at us from the classroom, to nuclear security issues. And we are so damn blind I just can’t believe it. And I’m not a nuclear security expert of any sort.

So I really wonder, what’s the point of establishing twelve moral principles, where everybody else who really wants to undermine them would just not care? I would be more than happy to believe that there will be an oversight agency, but I don’t think there is a global government anytime soon. So I really wonder what kind of possibilities are there, besides the ideal that, oh, we will all work together, and hold hands, and walk in that lovely Alexa-enhanced reality? Thank you. (Laughter.)

ROTENBERG: Yeah, we’ll do our next session on virtual reality and have that kumbaya moment for AI—but since you didn’t ask me about the second postcard, I can get that into your question. (Laughter.) And I still have a laptop sticker for you, by the way, but that’s going to require another question. (Laughter.)

So what we are trying to do is to get the Federal Trade Commission (FTC) to issue something more than press releases. And we have squarely put before the FTC a forty-six page complaint that explains many of the problems we found with ChatGPT, including bias, including cybersecurity, including risk to children safety, and, the most serious problem, detecting whether your students might be using it. Probably should put that at the beginning, but anyway. Long complaint.

But we are, in a sense, trying to force the federal agencies to take an action, any action, much like the Italian privacy agency did just a couple weeks ago. And interestingly, they announced a ban in Italy through the privacy agency because of the violation of certain privacy rules. They just lifted it today, in fact, because OpenAI agreed to some modification.

And this is my view, actually, of how we make progress in this space. That if we can engage the regulators, shape the business practices, make the companies more accountable—I don’t think we get to a perfect solution, but I think we get to a better solution. And we can still do kumbaya.

ZAID: Marc, if I can follow up on that a little bit. Didn’t China just announce that chat bots would have to sort of parody Communist Party-approved language? So that is kind of the opposite.

ROTENBERG: So there is always—there is always a downside to regulation. (Laughter.) And, yes. I mean, I’m studying the Chinese regulation. There’s a lot of stuff in there that I like, protect privacy, protect intellectual property, promote accuracy. And oh, by the way, ensure the continued survival of the Chinese Communist Party. Well, OK. I mean, treat it as an a la carte menu. Take the items you like.

Which, by the way, is pretty much what the Europeans did in the last-minute modification to their AI Act. This is just in the last week or so. They said: intellectual property. We’re going to do that. We’re going to do privacy. We’re going to do accuracy. We’re not going to uphold the Chinese Communist Party for the European Union. So I think there is a way in this space to take the proposals you agree with, incorporate them in your legislation, and push them forward.

ZAID: I saw another question here, there first, OK. And remember, name and affiliation, if you will.

Q: Thank you so much. This is fascinating. So you’re talking to the vast majority of college professors here, right? My—sorry, Kimberly Pace, University of Alaska, Anchorage.

And my question is about ChatGPT. And what, if anything, can we do? Because there’s not a lot that will catch it. SafeAssign doesn’t really say this has been written by ChatGPT. I’m just wondering about—because there’s a lack of oversight—what can you do to assure us that something good will come out of it for the college professors, who are in the process—I mean, we are all—I mean, I’m in finals week next week. And I’ve got a lot of papers to grade. And I’m concerned.

ZAID: That sounds like a trick question. (Laughter.)

Q: It is a trick question, but I have faith and confidence that you will have a tricky answer. (Laughter.)

ZAID: Do you want to—

KAHN: I can talk about it, I guess. Again, not exactly from a—I think there’s a broader question about the technical implementation of some of these regulations that I don’t know the answer, honestly, for an education perspective. There are tools that are things like watermarking, things that you can kind of identify that are—again, can be built into the system and part of that regulation. And I think why, again, the private sector needs to work in close collaboration with the government on these efforts. And I think, honestly, it connects to the question that we had before about what are the kind of values about putting out these principles when we don’t have ways to kind of enact them, right?

How do you ensure—like it’s all well and good to say, we want, again, a system that is responsible and transparent. OK, how do you measure transparency? What kind of metrics are you looking at? How do you ensure that? I think that’s going to be—I think some of those problems are tractable engineering problems, but I think some of them are a little bit harder, and there needs to be kind of a broader philosophical kind of agreement about what kind of level of accuracy is acceptable, what kind of—when is it useful to have an AI system in place versus a human, in what kind of context. I think that is going to vary greatly based on the application, and will depend—but I think it’s a very—again, it is a very hard, tractable engineering problem.

ZAID: Marc.

ROTENBERG: I think it’s a very difficult problem. I mean, I’m teaching this semester. And I said to my students, well, if you use ChatGPT to write your papers, I’ll use ChatGPT to grade you. (Laughter.) But, you know, it doesn’t matter because the registrar says everyone an A anyway, so. (Laughter.) So that’s—oops. (Laughter.) But, I mean, I’m familiar with some of the technical solutions. So there is, of course, ways to watermark. But no student who’s aware of watermarking is going to do a simple cut and paste. There are always ways to get around those kinds of measures.

I am genuinely concerned, and I’ve written a little bit about this, I think the need for critical thinking in this moment, which has always been part of our liberal arts education tradition, is actually greater than it has been in the past, precisely because we’re about to enter an era where a lot of the words we read will not be produced by people. And our ability to interrogate and understand how to question and how to reason actually becomes more important.

I actually—I heard a pretty scary talk by a really thoughtful guy named Louis Rosenberg, who’s looked at the issue of ChatGPT and manipulation, and he’s really drilled down. Whatever concerns you may have about how advertising manipulates opinion, what you can do when you begin to personalize, and when you profile, and when the person who’s on the receiving end of the conversation actually has none of the indicators that we have when we interact with one another, it’s entirely an asymmetric relationship of power.

So again, trying to scare you. At the moment, unfortunately, when students are getting tools that allow them to sort of detour around the learning of critical reasoning skills, they actually have a greater need to develop those skills because of what they’re going to encounter in the future.

KAHN: I think an also interesting point, just a distinction that I’ve noticed in speaking to those working in these fields, depending on how the information is presented varies greatly. I think a lot of the conversation, why we’re talking so much about ChatGPT, is people find the text very convincing, but similar advancements being made, for example, in computer vision aren’t having the same impact, because individuals aren’t as convinced. They can notice anomalies or things that don’t quite match up there. So I think that’s just an interesting aside. And for, again, academics in the room, please look into that for me. (Laughs.) I’m very curious to see if that’s based in behavioral cognitive science as well.

ZAID: We’ll take our next question right here.

Q: Thank you. Barry Driscoll, political scientist at Grinnell College.

Interestingly, I wanted to get people’s names when I was on the flight over here to see who would be with me in this workshop. And I didn’t have the CFR materials. I used ChatGPT. And I just wrote the words, “this person X—I had your name—this person X is probably a professor in the United States, write me a bio. And I got bios for almost everybody in the room, just from writing that.

AUDIENCE MEMBER: But were they accurate?

Q: What’s that?

AUDIENCE MEMBER: But where they accurate? Well, of course, I have no way of knowing. (Laughter.) I have to assume that they’re accurate, which is part of the problem. And then one of my advisees emailed me and said that a cover letter she had written for a potential employer was written by ChatGPT. She just told me that with a smile on her face. We had a talk about it. (Laughter.) The specific question is, about the—in DC, all these potential regulators, the acronym soup that you mentioned. I’m interested in who’s lobbying and who’s engaged in the conversation with those regulators, or potential regulators. Civil society actors are, in the form of you. What other kind of actors are you seeing lobbying?

ROTENBERG: Well, to be sure, the business groups have strong opinions. But it’s actually fascinating. And I was speaking about this at a conference earlier this week. The U.S. Chamber of Commerce, representing U.S. business, put out a paper last month calling for legislation for artificial intelligence. And I can’t remember the last time the U.S. Chamber of Commerce called for legislation on anything, right? (Laughter.) I mean, it’s really quite remarkable. Now, if you go to Google, which is launching Bard, or to Microsoft, which just has put $10 billion into OpenAI, they really don’t want legislation. I mean, this is a battle over the future of internet search, and a hundred related services.

Which is also, by the way, why I think it is that a lot of the leading AI experts, aware of this commercial dynamic, wrote the Pause Letter. Now, the Pause Letter was controversial. Not just because Elon Musk signed it, and everyone on Twitter is very angry at him. But because it seemed to be purely about the existential threat. And a lot of people working on the here and now issue of bias, let’s say, predictive policing, facial surveillance, said we need to solve those problems before we get to the killer robot problem.

But the reason I’m mentioning the Pause Letter is because I think people on the inside of the industry were really concerned that the tech titans would basically push forward without any constraint. So apart from those folks, my read in Washington is that most people actually want to do something. Senator Schumer has said AI will be a legislative priority. President Biden has said, by the way, some remarkable things, speaking to his science advisors earlier this month. He said, companies should not release AI products that are not safe. I mean, that’s quite remarkable coming from the president.

So I think you’re going to see action. We have some catch-up to do with Europe. I’m hoping for an alignment and a bit of convergence in the transatlantic digital realm. But again, to your point, civil society, business split between tech titans and everybody else. Educators, I think you all should be a part of the—be a part of the discussion. I mean, in many ways, the people in this room, I suspect, are on the frontlines of this transformation that’s taking place. You deal with the written word. You’re teaching the next generation of people who are going to live in this AI society. I think your voices should be heard on this issue.

ZAID: Right here in the middle.

Q: Sabrina Safrin from Rutgers Law School.

I want to build on some of this regulatory issues that you’ve been putting out. I have two questions. Number one, how—you described how in the United States we’re seeing a division between the tech titans and everybody else. I’m wondering if we’re seeing that division in other countries, or does it—where are the countries lining up in terms of wanting regulation or no regulation? And then my second question is linked to that, if there’s any press for there to be an international treaty on this, like there was with biotech?

ROTENBERG: Right. So I think it is really pretty much in the U.S. It’s Google, it’s Microsoft, it’s Anthropic, it’s a couple of other firms that are heavily invested. And it’s an understandable business dynamic. I mean, I’m not being judgmental. If Microsoft pours $10 billion into a company, I assume they’re expecting some returns from their investment. But you don’t see that elsewhere in the world. And even in China, where there’s been a lot of success building vibrant tech companies, I think the CCP has made clear that they’re going to keep those companies in check so that they comply with the regulations that the governments have established.

Now, to your second point, we’re actually involved in the effort to establish the first AI treaty. And this is underway now at the Council of Europe, who just heard the fifth meeting in Strasbourg, the fifth plenary session, heading toward the finish line. I think it’s a good framework. And, by the way, Council of Europe treaties can be ratified by member states and nonmember states. Cybercrime Convention was ratified by the U.S., and Japan, and Canada. So there’s certainly a future in which the U.S. signs up for this AI treaty. There’s a bit more work to do to get this done. There are big questions about oversight and enforcement not yet resolved. And then a little bit of European skirmishing. The European Union wants to get its AI Act done before the Council of Europe does its treaty. So we’ve got to get the AI Act done, which likely happens early part of the second half of this year, and then the treaty follows afterward.

KAHN: And I’ll just add, from a security perspective, there’s pretty—it seems pretty receptive to something like—not necessarily at the treaty level, but a confidence-building measure. So something informal, guided principles. I’m pushing for something like an autonomous incidents agreement, potentially. Something like the Incidents at Sea Agreement. Seems to be a little bit kind of—a little bit in the works.

ZAID: OK. Going to go here. And then do you also have a question? We’ll go after you.

Q: JP Maddaloni from the Navy War College.

So I can’t remember when it was in the panel, but you mentioned something, Marc, where you were, like, I don’t have enough time for that. And I was wondering if that’s going to be like the new catchphrase that leads to disaster using AI, kind of, like, the—like, wait a second, hold my beer. (Laughter.) So I was thinking about decision making and the future application of decision making for AI. And some of this might be used currently, but may be something that’s off in the future. And so a lot of military applications are, you know, what is my percentage or probability of success for something. So I want a dashboard, I want a looking glass, a windowpane. It’s called a lot of different things these days.

But I want that assurance, essentially, right, that whatever action I’m going to take has a better chance of success or not to take it. And so if you could comment a little bit about that as a future application. It doesn’t have to be just military, of course. There’s many ways that that can be applied through other civilian and public enterprises. But anyway, for what you see for that today, and then what you think the role of AI is in terms of decision making.

ROTENBERG: Well, I’ll leave to Lauren, I guess, the broader national security dimension. I will say, I was very interested in an AI program that Facebook developed called Cicero, which did very well in diplomacy. And Cicero—I’m sorry—diplomacy’s an interesting game. Because you both make commitments, and you break commitments, and there’s a bit of poker-like bluffing. I mean, I’ll leave Austria alone as long as you don’t touch me in Italy. But, hey, you’re not touching me in Italy, so now I’m going to Austria, right? And the Cicero program—(phone rings)—is that me? Oh, sorry, that’s an AI. (Laughter.) Cicero actually—(background noise). All right. So that’s a story for beer. We’ll save that for later.

But going back to Cicero, so obviously one of the immediate scenarios you imagine is a program like this, you know, sitting with the president among the national security advisors, and people are trying to come up with a strategic response in some critical situation. I don’t think you would rely on Cicero to make the final decision, but you could actually imagine a scenario where a system that’s developed a fairly high level of confidence and reliability expresses an opinion, and now the people in the decision-making setting are trying to assess whether that’s a good option. I actually anticipate that we’re going to begin to see that happen.

Now, I want to be careful here, because the phrase that’s oftentimes used is “human in the loop.” And I think that phrase is a terrible phrase. I think you want AI in the loop, and human in charge. I think that’s the right way to understand these types of settings, where you can get some expert advice from a well-trained system, but you would never delegate the final decision to that system. Is that right?

KAHN: I love that. I think—I agree entirely. And I think actually this is a very interesting question, because I see this kind of sphere as the way AI is going to be the most competitive advantage from a security standpoint. And I think everyone likes to talk about the killer robots, and putting AI in various platforms. But it’s really about generating efficiencies and about making better decisions, freeing up time for the decision maker to look at all the factors, so it’s not having to pull a million bits of information.

On the other hand, I don’t want to do the bad cop thing but get why it’s scary. This is where the human and the human reliance on the system becomes really challenging, and training becomes exceptionally important, to make sure that it doesn’t become a situation where the human is just relying and cognitively off-loading to a machine. That they understand that they’re making the decisions and when or not to trust. This is something—a bit of a passion project of mine. I’m definitely doing research on rates of automation bias. When are people more likely to trust a machine, or when are they more averse to it? And I think that comes into trust and making sure—and confidence in building expected accuracies about these systems, and things that we can make sure that they are explainable, or that someone can see what they’re kind of doing. I think that’s going to be a necessary precursor in order to leverage that to its full capacity, for sure.

ROTENBURG: I’m just going to add one more word in the national security context, so you also see the risk in the scenario I anticipated. If I was aware of an adversary using an AI expert system to help inform—and you’re all already nodding, because you know where I’m going—to inform high-level decision making in another government, I would target that system because that’s where I’m going to get a point of access to a process. And this is where we come back to the problem of opaque decision making. I mean, we’re going to get a very confident, syntactically perfect output. But actually knowing how it was generated is almost impossible. And that, I think, will create some concern.

KAHN: I will add one more thing, because now you’ve opened a can of worms, unfortunately. But I think that’s going to be a huge—I think that’s one of the biggest threats I see from AI in a national security standpoint. You have the risk of inadvertent escalation, where, accidents—you can’t say who made a decision when an AI, for example, is in control. That could potentially lead to unintentional conflict or accidents. I think those are the three main risks, I would say, there. And again, it comes from what happens when there’s an accident. And you have to convey, oh, it was the AI, it wasn’t us, right, to an individual? How do you signal that effectively? How can you point to something and show an adversary if something happened? So I think that’s a very real, real technological concern. There’s maybe we can do watermarking for, like, algorithms. But, like, how can you actually do that? It’s a very hard question, for sure.

ZAID: I’m going to go to the middle of the room, and then to the back. Middle of the room first.

Q: Thank you. Carole Petersen from the law school at University of Hawaii.

So I have a quick follow up on that. Which is, I’ve been reading a lot about underwater drones in the South China Sea. And initially, it always seemed that the assumption was there would be a mother ship or some human control back there. But increasingly, I’ve been reading that both China and the U.S. have been working on systems that could be fully autonomous, or just about autonomous. So I’m wondering what your views on that, because, I mean, I could see a situation where China is patrolling its nine-dashed line with a lot of autonomous underwater drones. Is that a possibility?

KAHN: I think it’s definitely a possibility, and why I’ve been advocating for—we’ve seen—remotely piloted aircraft are not analogous, but at the same time you can kind of see where the technology is going. And so I think now is a really opportune time to have something like an agreement that I mentioned about how do you treat autonomous systems? Are you more likely to shoot it down? Are you more likely to take it over? What kind of protocol are necessary? And I think there will be a—I’m hopeful that there will be some mutual kind of interests on that, and that something like that would be established.

I’m hoping to see that as a follow-on from the Political Declaration on Responsible AI and Autonomy, but again this is very much my wish list. (Laughs.) And whether or not that’ll happen I think is dependent on how quickly these technologies are kind of going to—going to emerge. I hope it comes out before we see them used widely, but I’m not sure that that’s the case.

ZAID: We’ll go in the back of the room.

Q: Thank you. Stephen Long, University of Richmond.

I can’t help but notice a bit of a difference between this conversation and what I heard last year at the Stockton Center at the Naval Academy, where they did an ethics in AI in the military panel. And they had, I forget the man’s name, but he was a general in charge of a joint AI taskforce across the services. And his presentation was really largely about the competition with China. And he would refer to the U.S. being six months ahead, or eight months ahead of China.

And it felt like this big, massive other thing in the room, where we very much want to do it right. And I think Europe definitely wants to do this right. And on the consumer level, now that everybody’s aware through ChatGPT, we’re really talking about it. But there was also this sense of a military imperative not to go too slowly—as you had said, Lauren—not to go too slowly because China will get there first and get some unseen advantage. And I just can’t help but wonder if it’s a bit of a kind of a race to recklessness, like a race to the bottom in how thoughtful countries are going to be about AI. And how long can we resist that pressure to keep up with governments that might not be taking the time to think through the long-term implications? Thank you.

ROTENBERG: So let me say a couple of words on this point. I’ve actually been in a lot of meetings on this topic. And, to be sure, particularly in Washington, there’s a lot of focus right now on China. It seems to be the one area where there’s bipartisan support. And my view is that the competition is very real. The competition is very real in the national security realm, it’s very real in the scientific realm, it’s very real in the commercial realm. China, without question, intends to be the world leader in artificial intelligence. And it believes that if it achieves that goal, it will have sustained dominance in the twenty-first century digital economy. I don’t think that’s really in dispute.

And I can even share with you—because actually I have a bit of background in AI and computer game programming, that in 2017 when AlphaGo beat Lee Sedol, the Korean Go master, world champion, it was literally a Sputnik moment for China. This was a U.S. company, Google, defeating a game that they simply assumed could not be reduced to a computer-based problem to solve. And they went into high gear. So the competition is real. The ambition is real. However, this is where they—I think the story gets interesting. Because I think China is moving forward both on the innovation front, and on the regulatory front.

And this is what I don’t think people in the U.S. understand. And that is that they believe that they will need the regulatory framework, not only to ensure the development of AI that they can rely upon, but also in the Belt and Road Initiative countries, and elsewhere, have the regulatory structure that supports the products they’re developing. China is producing as much regulation on AI as is the European Union. And it looks, by the way, very much like EU legislation. They have the PIPL that mirrors the GDPR. Some of you may know what I just said. (Laughter.) They have a regulation on recommendation algorithms that looks like Article 63 in the Digital Services Act. They’re, like, looking to Europe to try to figure out how to regulate this stuff.

So my answer is that I think the U.S. needs to do both. I think we need to maintain and up, actually, our investment in our education, and compete. But I also think we should not assume that regulation necessarily slows us down or involves a tradeoff. Because interestingly, the Chinese have already made this calculation. And I think they believe that the strategy for success is both of the above.

ZAID: Lauren, we’re just at time, but if you want to add, since this is your area of expertise.

KAHN: I’ll just add, like, a very quick aside. I also—I think there’s processes in part that are slowing it down. So I think that’s probably why a lot of the language is to speed up. I think the term being used now within DOD and the Emerging Capabilities Policy Office is “responsible speed” to exactly capture that kind of dynamic between moving too fast, right, and kind of cutting corners, and moving too slowly about adversaries. I think there’s a way to do both, as Marc mentioned. And I think it’s—I think it’s being pursued right now.

ZAID: Great. So join me, please, in thanking Marc and Lauren. (Applause.) And thank you all for joining the 2023 College and University Educators Workshop. I think Irina’s going to come make a few comments. Is she here? No? Go to the reception? OK, all right. Well, all right, let’s go to the reception. See you guys there.

(END)