← Sinica Podcast
Loading summary
Transcript23 lines
- [00:01]
A
Dish has been connecting communities like yours for the last 45 years, providing the TV you love at a price you can trust. Watch live sports, news and the latest movies, plus your favorite streaming apps all in one place. Switch to Dish today and lock in the lowest price in satellite TV starting at $89.99 a month with our two year price guarantee. Call 888-add-discosh or visit dish.com today.
- [00:30]
B
If you're a therapist listening, you already know your work doesn't magically end when the session does. There's scheduling, notes, billing, insurance, follow ups, all of the admin that happens before and after the work you actually care about. That's where SimplePractice comes in. SimplePractice is an all in one EHR built specifically for therapists with HIPAA compliant tools and high trust certification. No juggling systems or cutting corners just to keep things moving. So scheduling, documentation, billing, insurance, client communication, even automated appointment reminders. It all lives in one place. And if you're starting or growing a practice, SimplePractice also offers a credentialing service that helps simplify insurance enrollment, which can be a huge lift alone. Right now, SimplePractice is celebrating mental Health Provider Day with an exclusive Offer up to 70% off for one year. Yes, up to 70% off for 1 year. But hurry. Offer ends May 15 at SimplePractice.com SimplePractice.
- [01:40]
C
Welcome to the Seneca Podcast, a weekly discussion of current affairs in China. In this program we look at books, ideas, new research, intellectual currents, and cultural trends that can help us better understand what's happening in China's politics, foreign relations, economics and society. Society. Join me each week for in depth conversations that shed more light and bring less heat to how we think and talk about China. I'm Kaiser Goa, coming to you this week from my nearly empty, soon to be on the market home in Chapel Hill, North Carolina. Sinica is supported this year by the center for East Asian Studies at the University of Wisconsin, Madison, a national resource center for the study of East Asia. Listeners, please support my work by becoming a paying subscriber@senecapodcast.com I know there are a lot of substacks out there and they start to add up, but I think this one delivers serious value and I do need your help to keep doing this work. So please do subscribe so I can continue to bring you these conversations. This week I'm sharing the fourth and final installment from the day long conference convened by the Institute for America, China and the Future of Global affairs at Johns Hopkins School of Advanced International Studies. That was on April 3rd in Washington called the China Debate. We're Not Politics, Technology, and the Road Ahead. The first episode featured Jessica Chedweis opening remarks and the panel on what China Wants. The second panel turned to the equally important question of what the United States wants from China, which is something Jessica has been particularly good at problematizing. I've often borrowed her language to express the idea that recent American administrations seem somehow unable to articulate an affirmative vision for what the US Actually wants things to look like. Then you heard another excellent panel called Tech Rivalry and Competing Visions of the Future, which was focused, unsurprisingly, on artificial intelligence. This week, still more AI the AI Race Reconsidered, featuring moderator Henry Farrell, the Stavros Niarchos Foundation Agora Institute professor of International affairs at sais, who will be in conversation with Alondra Nelson, who is Harold F. Linder professor of Social Science at the Institute for Advanced Study. This is a terrific conversation that you will not want to miss. At the end, you will hear closing remarks from the organizer herself, Jessica Chen Weiss, ACF's inaugural faculty director. Once again, my deep thanks to Jessica for organizing this terrific conference and for so generously letting me share this audio with Seneca listeners. Please enjoy.
- [04:28]
A
I'm here to interview Alondra Nelson, and I'm really, really looking forward to the conversation that we're going to have. So Alondra is somebody who has had an extraordinary career over the last number of she is now the Harold F. Linder professor at the Institute for Advanced Studies. Before that, she was the Director of the White House Office for Science and Technology Policy. But some of us have been following her work for a long time. Because if you are in this intersection where you're thinking about the relationship between society and sociology, culture, technology and political economy, this is something that you've been working in for decades, and I think your work has really made the field and established it. I think it's a really valuable perspective to have to bring together the stuff that you've been doing more recently and also your deeper intellectual roots, because I think that they suggest a set of issues, a set of questions that don't get nearly, nearly enough attention. So I hope, if that's okay, that's what I would love the conversation to be about.
- [05:34]
D
Yeah, well, I hope it will be a conversation and not an interview. And I think the theme of today is clearly sort of asking, going to topic areas that we're not quite getting to in this area in this space. So I'm looking forward to it. Thank you, Henry.
- [05:47]
A
Okay, so what I would like to do is maybe begin with something that Selena said in the previous panel and maybe add a little bit of spice, which is. So, you know, I look at the debates around AI, which frame the broader debate around US China, competition, and in the United States, there are a number of rather unusual ideas that have managed to gain a lot of weight. So I spent a lot of my misspent youth in Ireland in the 1980s and 1990s reading science fiction, some of it good, some of it bad. I never expected that some of the ideas about the Singularity, about AGI, all of these ideas would actually begin to become major cornerstones of the US debate about AI, where it is going and what it is going to do. So I'm wondering if you could just maybe talk a little bit about that, where these ideas came from, what. What they point towards and maybe what they hide from us as well.
- [06:44]
D
So I think those of you who've studied technology, sort of like in California, so I'm a Californian, if you studied Silicon Valley work, social science work, popular sort of theoretical work in the sort of 80s, 90s up to the present, talked about things like the emergence of Silicon Valley out of cultural formations, like the. Well, out of cultural formations, like formerly hippie communities, out of things like what would become called the Californian ideology. Right. So that, you know, in my home state in California and Silicon Valley in particular, there has been for good and for not a history of kind of utre ideas. And sometimes these have brought us really incredible software development, new kinds of companies. You know, you might think about sort of Steve Jobs and the fusion of both technology and design, aesthetics and design as sort of an output of that. But it has also been the place of, you know, I think, friends, ideas or ideas that are sort of on the margins of mainstream society. And I think a lot of those have had to do with new technologies. And so, you know, it's not a surprise. I think we shouldn't be surprised if we kind of think back to that history of California more broadly. You know, that there have been ideas about the Singularity. And these go way back before AI or generative AI, go back some to science fiction, that there have been ideas about wanting to sort of people, you know, not thinking that having a sort of symbiotic or cyborg life with a. With a robot is a bad thing, that some people think it's a good thing. So ideas that have been circulating. So. But I think what was. What's been challenging in the policy space and for those of us who are policy researchers, is that, you know, they really have taken hold in the conversation about policy. And I think this has been. And Selena really did flag this has been distinctive in the way that discourse about AI and AI policy has emerged and evolved in the United States over the last few years in particular. And it manifests itself in, I think Saint Selena also mentioned this in polling, high negative sentiment in the United States. Right. Whether or not you're talking about the Edelman Trust Barometer, Pew, Gallup, and other kinds of more political polling sites, typically what happens is you get adoption over several years and negative sentiment, if there's any about a technology goes down because there's more adoption, people are more used to it. What we are seeing across these different waves of polling is that the negative sentiment has gone up since 2022. You know, it sort of really went up sort of in 22, 2024, and it is really plateaued where it continues to go up or the sort of issue space about which the negative sentiment is happening is spreading. So going from algorithmic discrimination to concerns about jobs to concerns about whether or not people can get a fair shake in health care, concerns about young people's jobs coming out of the workforce more broadly, concerns about csam, sexual exploitation, and young people's mental health. And so now you've gone from, I think, in some instances, fantastical concerns about existential risk, but that, you know, I think I would also say with an asterisk, as somebody who's worked in the Biden administration and the federal government, that, you know, there are people, thank God that they exist, who get up every day and their job is to think about existential risk. So, like, it is not, you know, I don't. I don't. But there's existential risk and then there's risk. Right. And how are you kind of thinking about that? And so that. I think that has been a kind of tailwind for a whole set of other negative sentiment. And we are really, I think, in a space in which, you know, Kat Duffy said this, that we are gonna need policies, laws, some kinds of guidelines to actually begin to have just basic consumer trust on these tools and their adoptions. So on the. And what it makes it distinctly, well, not only American, but certainly quote, unquote, Western, is the extent to which it is the U.S. australia, Canada and others, and not China, not Nigeria, not Brazil, that has the high and plateauing or growing negative sentiment around AI. This is quite the opposite in places like China I think there's probably lots of different reasons why, but one, as Sam said in her comments, is in part because there is a rule book. I mean, we don't, you know, in China for AI, whether or not it's followed evenly, it's totally followed, it's working. I mean, you can have lots of, I think, probing, empirical questions about that, but we've also, in the United States distinctly decided to say, and we can come back to this, that we are not regulating AI. And so there are lots of differentiators for AI policy in the US right now.
- [11:42]
A
Well, I would love to get to the question of exactly what we mean when we say we are not regulating AI, but before I do that, I would like to ask you about something else which comes from your ideas around this. So I know that you are very interested in the way in which the United States, many people in the United States have pursued what might be described as a Cartesian vision of AI, whereas perhaps in China, it's a more embodied approach and a more embodied perspective. And I wonder if you could talk a little bit about what the details of that differentiation actually involve.
- [12:13]
D
Yeah, so we've got this kind of mind, body, I think, AI, Cartesian divide that I think the US is now having to face. And what if you. If you go back, if you go back to 2020, the introduction or 2022, the introduction of ChatGPT, and some of those early sort of conversations and interviews were early from that period or from that period with Sam Altman. One of the things that's really distinctive about those interviews and talking about ChatGPT is that, you know, he was like, you know, 2022, the OpenAI spins off their internal robotics team. So they actually had a robotics team that was working. They sort of shut that team down. Altman says, we really are focusing on AGI. We can't have a diffusion of our purpose. What the robotics team had been working on had been this hand that was a robotic hand that was aiming to solve the Rubik's Cube. And that was going to be the task that it was, the benchmark task that it was aimed to do. So they shut down that work. And it was very clear that Altman and others would say, we're working on the brains. You don't need a body, you just need the brains. And there's some versions of this in the technology space from, like, Internet stuff from the 1990s, right, where we'd just be like, we're all going into, you know, this is kind of old sort of Bruce Sterling, Cyberspace stuff, we're all going to go into the machine and you don't need the body or the wetware or the meat or any of that anymore. And that was a kind of version, I think, of that, you know, which is which to say we're going to work on the brains and other people can work on the like, you know, the, you know, material, the physical that we're now calling it physical AI. And, and it was a curious, I think, perspective in part because those of us who know people who've been working in AI and the longer, you know, through the last couple of AI winters, know that a lot of those people were robotics people. So there was actually quite a lot of expertise and experience working in the robotics space. So now, in part because, you know, this came up in both of the last conversations, you know, China's differentiator, one of them has really been advanced manufacturing robotics, actually thinking about what you might call the brain's AI and the embodied AI together and building out systems that do those dually. We're beginning to do more of that in the United States as well. But I think that we think maybe both in the private and public sector R and D space tripped over ourselves a little bit by thinking that the brains of AI was somewhat superior to or distinctive from. And that's a kind of old stereotype to them working together.
- [14:51]
A
No, I think that's really interesting and also to work through, in a sense, why it is that we think about our progress in AI as involving scaling, as involving all of these ideas of geniuses in the lab. That's the amudea version, of course. Sam Altman has his own particular take on this. And really I'm sort of trying to strive towards, as you say, this kind of. This almost a disembodied brain in a VAT vision of what it is that AI is going to do. But the other side of that, as you suggested, is that there are a lot of social worries that are coming from the ways in which AI has been implemented. And I guess in particular, because we have really focused, and this is, I think, an implicit part of your story, large language models have become the way forward, as opposed perhaps to some of the more reinforcement learning aimed at sort of trying to figure out how AI would work in an environment, perspectives that predominated earlier. So if we were to actually pay attention, and here I'm asking you to put on your sociological hat, if you were to pay attention as a sociologist to the kinds of changes, the kinds of challenges, the kinds of worries that we have in society as AI becomes increasingly woven into the systems that we use. What are the things we should be paying attention to that we are not paying enough attention to right now?
- [16:11]
D
Well, I think if just, you know, I'll just segue from the robotics piece. I mean, you know, there is work today that remains quite dangerous. I mean, you know, there are ways in that we could use AI in those ways. I think to the extent that we have negative sentiment and we want to have, you know, more positive sentiment or people to feel like AI might actually be something that can help improve their lives, there's that kind of hilarious viral tweet of the, you know, the woman saying, I want AI that can fold my laundry. Well, you know, there might be AI that can fold your laundry, but, you know, as well as any of the. What we kind of think about as often as the kind of AI for good pursuits, you know, many of them, you know, may involve sort of things like advanced manufacturing. And so, you know, I think that we've got to begin to think in sectors about what it is that that would be good to automate, that we want to automate that we agree as a community should be automated and then think about, you know, combination of brains and embodiment that might do that. But there's lots of other things besides. And I think one of the challenges and where the imagery of the brain in a vat gets us is to making claims and statements about in an AI for good space that says, if we have AI, we're going to have drug development. If we have AI, we're going to have. We're going to cure cancer. And, you know, not only sociologically, I think, just materially, there are so many steps between that aspiration and that outcome. And part of what needs to happen there is the actual stewardship of those outcomes. That stewardship requires kind of clunky things. Institutions. Maybe they don't look like institutions of the 20th century, but, like, what are the structures that are gonna help us get from an aspiration for drug development to, you know, more drugs for more people, for example. So certainly thinking about new institutional forums, thinking about people's concerns about their privacy, people's concerns about access to data. I think the whole suite of issues around people's rights and access and opportunities remain really important, even as we're told, I think increasingly that they don't matter or shouldn't matter if we're going to not lose the race in the space of AI.
- [18:29]
A
Okay, So I think that's really important because as I see it, and of course feel free to disagree if I get this wrong, it seems to me that there are two major failure modes in the existing major debates. One of them you've already referred to, which is when the AI safety thing goes into Eliezer Yudkowski type speculations where we think that we are all doomed because we're going to press the button and intelligent AI is going to turn us into corgis or maybe just turn the plane.
- [18:58]
D
I do like a corgi.
- [18:59]
A
Yeah, corgis are awesome. Corgis are awesome. But the other failure mode is maybe the Mark Andreessen failure mode, which is the we don't need institutions. We will just have a glorious effective acceleration into a future where AI is going to do everything for us and sort of, who cares about that stuff? Doge that stuff away. We don't need any of the state or the traditional institutions that we used to have. And it sounds like you are pointing towards something which is thinking in a forward looking way about what kinds of institutional structures we might have in the future. And I'm just wondering, and this is also a selfish request, where do we look to for the people who are beginning to think about those questions in interesting ways?
- [19:47]
D
I think it's a kind of a strange bedfellows moment because I think there are, I think the sort of administrative state disorganization that we're living in right now, to put it politely, and things like the doge effort offer on the one hand an opportunity to say I didn't like these things anyway, shut them all down, burn them all down and we want to do something else. So that's the one bedfellows on the other side, the other are folks who are like, well, maybe these are opportunities for those of us who are reform minded, who believe in the importance of the administrative state, but knew that we were not getting everything right. And that part of the reason we went to government or we work at a think tank or work in the policy space is that we know that reforms are needed and institutions have to evolve. And so I think it's probably an opportunity for a conversation between strange bedfellows in that space. What I don't hope will happen in this moment is a kind of hearty defense of the status quo, because I think that's not. I think that's just wrong. In the space of science policy in particular, I kind of use the phrase a soft defense or things we want to defend, that there should be rigorous evidence based policymaking and research and these sorts of things, but I don't think we have to defend the status of the institution. So what does that mean? It means that we need to, I think, experiment with different kinds of forms of institutions that may or may not be comfortable to us. You might need perhaps fewer people doing a task, but I think we have to ask questions about who those people are and whether or not you need them. So if we're looking in how you need them to do the work. So thinking about private institutions or private companies and corporations, one of the things we're seeing in the news quite a lot is that lots of folks are getting laid off or there's kind of constriction in the employment sector for some, you know, including some of the big tech companies. So one way that you could as an institution respond to that is just keep doing all the same things you're doing with the same people. But I think that we, if we miss the, if you miss a generation of young people. So like, if, like, that's actually not good for institutional design. So I think we need to be thinking as we're transforming institutions. How do you keep young, fresh thinking people who, for whom things like AI, you know, are much more intuitive than other generations of workers? That's a question for government and it's also a question for the private sector. I think in government we have thought about really important new institutional forms in the Obama administration, the U.S. digital Service, 18F. It's not clear to me that those are the institutional forms we need right now, in part because I think it was indicated in the last panel, you're just not going to get a bunch of AI engineers working in government. There's just not enough of them, particularly if we've got immigration constraints. And, you know, not only would you be taking a pay cut, you're taking like an extreme pay cut. I think it's just a different moment. And also I think to do, potentially to do AI well in the private sector, in the public sector space, means asking strategic questions. It means certainly, as a colleague was mentioning earlier about having more people on the Hill who know more about the technology. But I don't think that's all that we need. I think that we need people who can think strategically about AI as an infrastructure for organizational structures and how and where you deploy tools and systems and models in different spaces in an organization. So that's a very different way of thinking about technological expertise and where you plug it in.
- [23:28]
A
Okay. So that also, I think that's an extremely attractive vision and I'd like to return to it right at the end. But before that, I'd like to talk perhaps maybe a little bit to get a sense from you about what you referred to in passing, which is this idea that the Trump administration is all about deregulating AI. And in fact, as you have argued in a piece in Science a couple of months ago, you think that it is not actually deregulating in any real sense at all. So I think that's a really important and subtle point, and I'd love if you could develop it maybe for people here.
- [24:02]
D
Yeah. So, you know, I worked in the Biden administration for a bit over two years. We did a lot of AI policy and I think also attempted to not be heavily regulatory, but I think philosophically wanted to simply say that existing laws apply to AI to get AI out of the sort of part of what the brain in the VAT idea about AI does, is put it in this sort of state of exception in the policy space, such that things can't possibly apply. So, and I think Lina Khan was, and the colleagues who were working in the civil rights space were very good at this. They put out a collective memo in 2023 or 2024 that said, if you're breaking the law using AI, you are breaking the law. The sort of mechanism for the violation of the law doesn't change whether or not the law is broken and whether or not there should be the penalties that we would normally use. So, so it was. I was. So I was mindful of the fact that without making any new laws, that the, you know, President Biden's AI executive order was basically calling on the interagency to actually do just do their workaday work and not take AI as an exception to all the work that you might normally do, just does a philosophical, philosophical approach to doing AI policy. So it was very curious to me, and, you know, I was interested to see what the Trump administration was going to do. You know, not surprised to see that some of the work in the Biden administration was rolled back. Of course, that's what administrations do. But it became curious to me that it was framed as deregulatory, because as the portfolio, the suite of things that the Trump administration was doing around AI was expanding, it was actually very clear that this current administration is actually very tightly stewarding the direction of travel of AI. So in the science piece I write about, so it may not be deregulatory. And this is where I think a new institutional framework is helpful. So I think those of us, of us with MPAs and MPPs or even have taken civics classes, are told that this is how government works and these are how the branches are supposed to work, et cetera. That's clearly, it's not been true in Washington in a very long time. We act like it's still true. It's certainly not true in this administration. But we need to keep our eye on all the other things that are happening. And so in the space of AI, just for a few examples, you have, of course, export controls. The Biden administration did some of that, of course, around semiconductor chips and going back and forth with that in this administration, I think now we're just. The H200 is the one that will allow and others we won't. But we also had this golden share, this, that we were taking equities in private companies, including nearly $9 billion in intel, and not, as we had done in the past, sort of offering incentives. If you reach a certain kind of benchmark in the development of the technology, we're going to offer you grants that have, you know, various kinds of milestones that a company has had us to hit. We are, I believe, the United States, we taxpayers, are the largest stakeholders, shareholders in intel, single shareholders. And so that means we have veto power, this golden share over how intel does its operations. I don't know if that's. Some people might think that's a good thing, Some others might think it's a bad thing. I mean, it's certainly industrial policy. You can shape the conditionalities that come out of industrial policy in a lot of different ways and maybe not make this choice. I think the concern for me and the choices that we make about the conditions the government sets on things like an investment in a company like intel are about where you still have areas for democracy to be there. Right. So if you are, you know, a golden share says, you know, the, the United States can sort of have a lever here, but who gets to have a say in that lever? It's that. That becomes very unclear. So it becomes also a narrowing of the democratic process. But so, so that's one example in which we don't call that regulation. We, you know, that we don't say that the Trump administration is making, really picking winners and losers, making very decisive choices about who's going to be in this growing AI ecosystem and who's not going to be and who's going to succeed and who's going to fail. That's one, I think, certainly U.S. immigration policy and it's the stalling of it, both what it means for bringing students from China, Iran elsewhere to the United States who work in the space of science and technology, you know, that is in the way, a kind of AI policy. So there, so what I would say is that sort of, there's the sort of Trump administration sort of has its finger on the scale of sort of a lot of different pieces that are a part of what we might typically call industrial policy, but lots of other things besides that are giving shape to what we currently call AI policy. And this shouldn't be a surprise, because the transform, if we believe that the transformation that AI is partly going to usher in is really an infrastructural one, if it's a one that's just not about new software in our offices, but it's going to change the way that we think it's going to change the social order, then because it is that kind of powerful tool, then we shouldn't be surprised that it offers new levers for managing it, for directing the way that it moves in the world.
- [29:37]
A
Okay, well, that brings me to the last question I'd like to ask, which maybe refers also a little bit back to what you were talking about when you referred to the need for institutional change within government. And that reminded me very much of the Mark Carney statement that nostalgia is not a strategy, that we can't and shouldn't be hearkening back to some mystical golden age because we're not going to get there. And clearly, if the Trump administration has really changed the way in which the state and the economy relate to each other, that opens up both problems and possibilities. So I guess my question would be if you were to think about this going forward, say there is a small d democracy oriented administration which comes in after this one and has to think about this really gets back to what Kat was saying, has to think about what would be the best way forward, given all of the possible possibilities, all of the opportunities, also the huge challenges that any administration is going to have to face, what are the broad contours or guidelines that you would offer to them as to how they ought to think about how even to begin to embark upon this herculean labor that they would have to embark upon.
- [30:53]
D
So I think any new administration is going to have to I think it will be, as a colleague said in I think the second panel, AI is going to be a campaign issue. It's already been a campaign issue. And so I think that any new administration coming in is probably going to have to have made some commitments about that. And what I think some of those should be is one I Think helping. I mean, I've actually been quite encouraged by the fact that we are having a conversation about data centers in part because I think broadly what I call the state of exception, that special exceptional place that we put AI in that, you know, says that other policies can apply to it, existing policies can apply, you know, that we can get a sense of disenchantment around that when we think about AI as being in a supply chain that includes data centers, which is to say it is not just a brain in a vat or this sort of, you know, ethereal ephemeral thing that like, you know, exists as a kind of epiphenomenon, but in fact takes electricity, produces heat, may cause pollution, sits in communities, et cetera. So that's only one piece. But if you think about critical minerals, not only the kind of weaponized interdependence of your amazing work in that space, Henry, but the fact that the places from which critical minerals must come. So I think that part of what this moment is preparing is for an administration. I think that can take a much more practical and material approach to AI policy, which is that it is a tool, it is software, that sort of thing. And I think in the policy space that will be very helpful because it allows more people to think that they can have levers to help shape and steward these tools. I think certainly we've got to deal with a lot of the negative sentiment. I think Kat was actually right in suggesting that it's going to take some policies and some guidelines. I think, think that there will be contested for sure by the time we get to, you know, a future administration, a lot more of state level policy, some of which will be good and will have worked and some of which will be problematic, but at least it will give us some data to actually think about, like what's worked, what's not worked, which, you know, what's scalable and what should we kind of toss to the side. So I think that will be very important. We will also, we will need, I think, different kinds of workers to work on AI and AI policy and government. Some of those, I think may come from the states actually. So I think that you could imagine, because that's where a lot of action and activity is happening, in addition to kind of companies and the private sector. But I hope it will be, and you're seeing this in some of the responses. So, you know, Representative Ocasio Cortez and Senator Sanders released this sort of data center moratorium bill a few days ago and it is using, and this again is the strange bedfellows that goes back to, I think, this distinctive, curious American way. So it's using this very populist language, which is that there should be a moratorium on the creation, development of data centers until such time as we know that they can be sort of advancing human flourishing in the cause of humanity. I mean, like, wow. But what I do appreciate about that is the high level and it was kind of what we tried to accomplish with the AI Bill of Rights, which is like, what are the kind of high level aspirations for a community or for a national community? And how do you think about the ways that you can offer guidelines around particular technologies that get us there? Not that the technologies themselves ensure our rights or ensure certain outcomes, but that you can use the technology that you can create these as kind of high watermarks as North Stars for how we think about procurement practices, interagency practices to the extent that we'll have an interagency around AI.
- [34:58]
A
Well, thank you so much. That was just an extraordinary conversation and I think an extraordinary place to finish on this idea of having a combination of some kind of North Star to aim towards and an understanding of the practical material consequences of the technology rather than just reading some fantastical future in as given would be extraordinary progress. So thank you so so much. Thank.
- [35:30]
B
You.
- [35:30]
E
This is on here. What a day it has been. Thank you all so much for being here with us for this really important series of conversations. I'm not sure that we have found all the answers, but we have opened up Pandora's box perhaps and hope that you will return and continue to interrogate these really fundamental questions with us in the months and years to come. I want to thank to all of our panelists for the incredible insights and the array of out of the box thinking that we engaged in today in such a rigorous but also civil and creative way. Some of the sort of the founding values of the Institute. For more and in fact many of these were showcased on the stage today. We have our ACF Insight series where we will continue to publish sort of some of this rigorous foundational thinking, not only about key assumptions underpinning U.S. policy toward China and the array of intersecting issues, but also specific policy recommendations. So please visit our website, stay tuned for that and more. And then thanks not least to our growing team here at the Institute, Managing Director Margaret Myers. I also see here in the audience we have Kate as well as Thomas and many junior fellows here, students at sais, who are helping us to launch all this incredible activity. So, so please tell everybody about our growing institute. We will be of course recruiting new Junior Fellows in the fall. I know this is admissions season and acceptance, so we hope you will to join us here as we continue to build this growing community of thinkers and doers, scholars, experts, dedicated to a different kind of conversation about China, about US Policy grounded and the sort of rigorous, sort of evidence based policy making and conversation that we hope will suffuse all of Washington. So thank you again and please join me in thanking each other and for this wonderful day.
- [37:44]
D
Sam.