Artificial Intelligence and Great Power Competition, With Paul Scharre

LINDSAY:
Welcome to The President's Inbox, a CFR podcast about the foreign policy challenges facing the United States. I'm Jim Lindsay, director of Studies at the Council on Foreign Relations. This week's topic is artificial intelligence in the era of great power competition. With me to discuss artificial intelligence in its role in competition with China is Paul Scharre. Paul is vice president and director of Studies at the Center for a New American Security. He is a former army ranger with multiple tours in Iraq and Afghanistan. He worked in the office of the secretary of defense in the Bush and Obama administrations where he played a leading role in establishing policies on emerging weapons technologies. The Economist named his first book Army of None: Autonomous Weapons in the Future of War as one of the top five books to understand modern warfare. Paul was kind enough to come on The President's Inbox back in 2018 to discuss it. He's now out with a new book Four Battlegrounds: Power in The Age of Artificial Intelligence, which is getting rave reviews. Paul, thanks for coming back on The President's Inbox.

 

SCHARRE:
Thank you. Thanks for having me.

 

LINDSAY:
Paul I'd like to begin with a basic question in that everyone these days is talking about artificial intelligence or AI. What exactly is it?

 

SCHARRE:
Yeah, that's a great question. AI is everywhere, but pinning down, what do we mean when we say artificial intelligence is really critical because we've been able to build machines that have a very narrow form of intelligence. They're able to perform various tasks, whether it's generating art images, generating text, doing facial recognition, identifying objects in videos and images better than people in many cases, but they lack the general purpose intelligence that humans have, the ability to learn a whole wide range of tasks. 

 

What's been behind the recent explosion in AI has been largely driven by machine learning and in particular, deep learning, a type of machine learning that uses deep neural networks. And these AI systems like Chat GPT for example, that have garnered so much attention, one of the things that's interesting about them is they learn based on data. So rather than older rule-based systems, like for example, a commercial airline autopilot where there's a set of rules for how the airplane has to behave, these newer systems are trained on data and then they learn from the data to identify patterns, and that governs their behavior.

 

LINDSAY:
Okay, so Four Battlegrounds opens with an epigraph from Russian President Vladimir Putin, let me quote it, "Artificial intelligence is the future not only for Russia, but for all humankind. Whoever becomes the leader in this sphere will become the ruler of the world." I take it that you share President Putin's assessment of AI's central importance. Indeed, you write that it is changing global security and power dynamics. How so?

 

SCHARRE:
Artificial intelligence is a general purpose technology, much like electricity or the internal combustion engine or computer networks. And we saw that prior general purpose technologies led to the first and second industrial revolution where nations rose and fell on the global stage based on how rapidly they industrialized, and even the key metrics of power changed. So coal and steel production became key inputs of national power. Oil became a geo-strategic resource that countries were willing to fight wars over, and I think it's the case and we'll see, but it's a hypothesis, that AI is likely to have similarly transformative effects across society and a whole wide range of industries, increasing economic productivity, accelerating scientific progress, and transforming military power.

 

LINDSAY:
So if AI is going to become a central ingredient of power in the geopolitical sphere, which countries are best poised to take advantage of this new source of power?

 

SCHARRE:
Right now, the three main centers of AI power are the United States, China and Europe. And the U.S. has been in the lead, China is rapidly catching up and depending on the metrics that you're looking at, in many cases China may be on track to overtake the U.S. in the next decade. China has said that their intention is be the global leader in AI by 2030. That's certainly gotten a lot of attention and raised concern here in Washington. I take that quite seriously. And while it remains the case that U.S. labs are at the frontier of AI research, AI technology spreads so rapidly that it's a very level playing field and China's in many cases only a few months behind some of the most cutting edge breakthroughs like Chat GPT, for example.

 

LINDSAY:
Paul, I'm curious, what does it mean in this context to talk about winning the AI race? Because it seems to me that technology is always changing, it's always adapting. Technology has a tendency to leak, to spread, to diffuse across societies, across countries. And indeed, if I think back over the course of my own life, companies that were at the forefront of technology back in the sixties and seventies, many of them are no longer in business. I think, for example, of companies like Polaroid and Kodak, you can think of Lucent, a lot of companies that made a big splash, but then were disrupted by others. So how do we think about this race in which what is the cutting edge technology today may be supplanted by a different technology or a different approach tomorrow?

 

SCHARRE:
Well, as you pointed out, with these companies, this kind of disruptive change can be merciless to companies or to countries that aren't able to adapt to both maintain a leadership position in the technology itself, but then figure out the best ways of using it. So we saw during the industrial revolution, for example, that Great Britain and Germany industrialized faster than other nations. They raced ahead in economic power and then military power. As a result, Russia was a laggard industrializing, and by the end of the 19th century, they'd fallen far behind Great Britain and Germany. And so there's major costs in moving slowly. Technology's a key enabler of political, economic and military power, but it's not enough to be in a leadership position, countries also have to figure out the best ways of using technology.

 

LINDSAY:
So lets talk about how one succeeds in this AI space. You argue that there are really four main ingredients to success in AI. Walk me through them.

 

SCHARRE:
So AI in the form of machine learning is dominated by three key technical inputs, data, computing hardware and algorithms. Now, algorithms are really difficult to control, and so the competitive advantage that companies or countries have are going to come from either advantaging data or computing hardware. But that's not enough. Companies and countries also have to have the ability to translate these key technical inputs into useful applications. And so human talent is really essential. There's a fierce competition for human talent around the globe for AI researchers. They're in very short supply, and the institutions that are used to translate data computing hardware and human talent into useful AI applications are essential for maintaining a leadership position in AI.

 

LINDSAY:
So let's talk about those components. Let's talk first off about data. For several years, I've heard about how China has the advantage when it comes to data because there are more Chinese people, and so there's more data, that we have a government in China that is not concerned about privacy issues. So China is destined to win the AI race because it just has more data that it can use. Is that true?

 

SCHARRE:
Well, this is the fascinating thing, is we've all heard this and it seems on the surface like it should be true, but when you dig into it, it turns out that that's not a great way to think about data advantage. So yes, China has more internet users than the United States or Europe, and of course the Chinese Communist Party doesn't have the same kind of restrictions on its ability to collect data about its citizens. But those are not the right metrics because when you look at national user base, that's not as important as the user base of companies. U.S. tech companies have global reach. Facebook and YouTube both have over 2 billion users globally. WeChat by comparison, has only 1.2 billion users, and in fact other than TikTok, Chinese companies have really struggled to break into the market outside of China. It's also not the case that the same freedom that the Chinese Communist Party has to collect data on its citizens applies to Chinese companies.

 

And in fact, the party has been pretty proactive in regulating consumer data privacy inside China, in part because of a series of consumer data privacy scandals, but also because the party doesn't want anyone else to have the same spying powers that they have. Whereas here in the U.S., we don't have federal data privacy regulations, and American citizens have basically acquiesced at companies hoovering up a lot of their personal data without a lot of pushback. So when you actually look at the type of data that companies have access to, it's not quite as black and white as it might appear. But most importantly, and this is where people may have heard, data is the new oil, like any comparison, there's good or bad things to that in play, but data's not a fungible asset. That way it's not like oil at all. And so better data, for example, on Chinese faces, doesn't translate to non-Chinese faces, and it's definitely not going to translate to some other application like training an AI fighter product.

 

LINDSAY:
And you're making this reference to Chinese faces because China has been at the cutting edge of developing surveillance technologies that rely on facial recognition, correct?

 

SCHARRE:
That's right. So half of the world's 1 billion surveillance cameras are in China, and they're increasingly using AI tools to empower the surveillance network that China's building, whether it's facial recognition, gait recognition, voice recognition, it's being merged with geolocation data from phones with biometric data to monitor and track Chinese citizens. And so China is building this very dystopian tech enhanced authoritarianism inside China. They're exporting it abroad. Eighty countries around the world have purchased Chinese policing and surveillance technology, but I think most troubling, we're seeing the export of the social software behind China's model of surveillance, the laws and norms that China is using, other countries adopting that.

 

LINDSAY:
But if countries are not interested in going down the surveillance road the way the Chinese have, it's less important to be at the cutting edge of AI in this space. Correct?

 

SCHARRE:
I think it's absolutely important to be at the forefront of AI, both to have an advantage in the technology that could be translated for a whole variety of things. So whether it's economic productivity, military power intelligence operations, but also because being in a leadership position in the technology can also help you set global norms. And the spread of surveillance technology globally and China's model of tech enhanced surveillance is threatening to human freedom around the world. And I think it's absolutely important that democratic countries push back with an alternative model for how AI technology should be used.

 

LINDSAY:
Okay. So let's talk about computing power, or I guess the term of art is compute. It's a little awkward to talk about it in that way, but let's talk about compute. Who has the lead when it comes to compute?

 

SCHARRE:
Well, in computing hardware or compute as folks in the AI industry talk about it, the U.S. has a commanding lead over China. China has a huge vulnerability in that they are highly dependent on foreign chips. China imports over 400 billion dollars a year in foreign chips, and none of the leading edge ships are made inside China, none of the most advanced chips are made here in the U.S. either, but they're made using U.S. technology. And so U.S. companies have controlled over key choke points in the semiconductor supply chain, and the U.S. has used this to cut off China's access to chips. It did this to Huawei, kneecapping Huawei's 5G business. And it did it most recently in October with the Biden administrations export controls on chips that effectively shut China out of the most advanced graphics processing units, or GPUs, that are used for some of the most advanced AI applications.

 

And if China's not able to access these chips, they can't get the computing hardware needed to train these very, very large models on massive datasets. So models like Chat GPT use thousands of GPUs running for weeks at a time. Without the computing hardware, China simply can't compete in the most advanced AI capability.

 

LINDSAY:
But how effective will this semiconductor and semiconductor technology ban be, Paul? Because one, the Chinese can look for ways to get chips through back channels, through front companies, as we've seen that in all kinds of sanctions cases. But there's also the question of can't the Chinese eventually catch up to the West, to the United States, in terms of the hardware part of the equation?

 

SCHARRE:
Well, I think it remains an open question, but the really key linchpin that's making all of this work is the restrictions on the manufacturing technology, the tooling and software that's needed to make chips. And that's almost the more important aspect of the export controls that the administration put in place, which will effectively freeze China out of the ability to build advanced semiconductors.

 

LINDSAY:
But that depends upon other countries agreeing to go along with the United States, particularly I guess the Japanese and the Dutch. The Dutch company ASML is on the cutting edge of lithography, the equipment used to cut these incredibly small transistors.

 

SCHARRE:
Exactly. So the U.S., Japan and the Netherlands collectively control 90 percent of the global market for semiconductor manufacturing equipment. So if the three of them say that the most advanced equipment doesn't go to China, then China is going to have a real challenge to find a way to indigenize this technology, because it's extremely advanced technology, and recently-

 

LINDSAY:
It's almost like barriers to entry.

 

SCHARRE:
It's a massive barrier to entry. And the technology's very, very difficult to build. So for example, in extreme ultraviolet lithography, the most advanced use for the most cutting edge chips, one company in the world, ASML, has a complete monopoly on this. So more recently, we've seen Japan and Netherlands say that they're coming on board with these restrictions. There's a lot that we still don't know about the details of them, although more details are leaking out. But if the three countries are able to stay together, China will really struggle to catch up in chip production.

 

LINDSAY:
The one country we haven't mentioned here is Taiwan, and you've likened Taiwan to being the Saudi Arabia of compute. Help me understand that.

 

SCHARRE:
Sure. So Taiwan occupies this really unique central role in the AI ecosystem, which is that 90 percent of the most advanced chips in the world are made in Taiwan. And so Taiwan has this incredible place in controlling the hardware infrastructure that's being produced to access the most advanced AI capabilities. Now, obviously from a geopolitical standpoint, that's deeply concerning given that the Chinese Communist Party has pledged to absorb Taiwan by force if necessary. And so it is yet another reason why Taiwan occupies such a central geopolitical flashpoint, and one where Taiwan's technological power has really significant implications for power in the 21st century.

 

LINDSAY:
So the third of your four pillars of power in an AI world is talent. Where do we stand in the talent competition? Where is the talent? Where might it be going?

 

SCHARRE:
Well, the U.S. has a tremendous advantage over China in the talent competition. What's fascinating is most of the top AI researchers in the world come from China, but they don't stay in China, they leave China and they come to the United States. And the U.S. is a magnet for global talent. So the best AI researchers in the world, including from China, they want to come to the United States, they want to study at U.S. universities, they want to work at U.S. companies, and the numbers are staggering. The brain drain from China and the gain to the United States is massive. So over half of the top undergraduates in China studying computer science come to the United States for graduate school. And of those doing a PhD in computer science here in the U.S., 90 percent of them stay in the United States. So the U.S. is able to bring China's best and brightest over to the U.S. and to keep them, and that's a massive advantage for the United States that China simply cannot compete with.

 

LINDSAY:
What about concerns I often hear, Paul, that that's actually a double-edged sword, that Chinese talent is coming to the United States learning science here in the United States, getting access to cutting edge technology that then gets returned to China even if those researchers themselves don't relocate to China?

 

SCHARRE:
Well, China's been very active in trying to tap into the diaspora of Chinese scientists around the world and bring that knowledge back to China. China has over 200 talent recruitment plans, the most notable of which Thousand Talents Program. And this is a problem. It's one where we need greater investments in screening upfront and then investigations to crack down on academic espionage, intellectual property stuff. So there's a level playing field for competition, but the U.S. nets out in a massive way from this talent flow. And one of the biggest mistakes the U.S. could make in this AI competition would be to cut off this talent flow coming from China to the United States.

 

LINDSAY:
Or the Chinese may decide not to allow their citizens to come to the United States, or Chinese citizens may worry that if they come to the United States, they may not have the opportunities they once did or be discriminated against. Isn't that a real threat?

 

SCHARRE:
It is a real threat, and it's really essential that the U.S. manages this in a way that we continue to be an inviting place for talented scientists and engineers from China and other countries to come to the U.S. and to stay here to study, to start their lives, to found a business. That's really core to sustaining American competitiveness. At the end of the day, the U.S. as a country of 330 million people is always going to be restricted against a country of 1.4 billion if we restrict ourselves to homegrown talent. If we tap into the best and brightest of the 8 billion people around the world, the U.S. will have an advantage that China cannot match.

 

LINDSAY:
Okay. Let's talk about the fourth pillar, institutions. What do you mean by institutions, and how are they essential to the outcome of this competition?

 

SCHARRE:
So institutions are the organizations that are going to affect how countries adopt AI and employ them. And we can see throughout history that what matters more than getting technology first or even having the best technology is finding the best ways of using it. So if you look at aircraft technology, the fact that airplanes were invented in the United States gave the U.S. no meaningful advantage by the time you get to World War II. What mattered much more was figuring out, what do you do with an airplane? How do you use airplanes effectively? And all of the great powers at the time, they had access to airplane technology, there were lots of different experimentations about how to employ airplanes effectively. If you look at carrier aviation, the U.S. and Japan were able to innovate effectively and employ aircraft on aircraft carriers to change naval warfare. Great Britain had access to the same technology, and they stumbled not because of their technology, but because of bureaucratic and cultural squabbles within the British military, and they fell behind in carrier aviation.

 

LINDSAY:
Let's pick up on that point about the impact of technology on warfare and on the military, because obviously AI has significant potential to scramble what we know about warfare, and again, you've written quite a lot about the changing nature of warfare. So walk me through how you see AI affecting the battlefield.

 

SCHARRE:
If you think about AI as another industrial revolution, the process of industrialization transformed war in very profound ways in World War I and World War II, where in total, the aggregate use of industrialized technology in terms of machine guns and railroads and submarines and aircraft increased the physical scale of destructiveness of warfare. AI is likely to do the same to the cognitive aspects of warfare. AI is a cognitive revolution. It will change the ability of militaries to process information on the battlefield and increase their ability to gain information to make sense of it, and likely over time, accelerate the decision tempo of military operations allowing militaries to move faster.

 

LINDSAY:
Okay. So help me think of that in a very practical way, because you have the issue of AI being this general technological revolution that makes more things possible. And as you point out, institutions can either accelerate that change or inhibit that change. And when I look at the Pentagon, I think about what I've been reading for four decades now that the Pentagon can be rather hidebound. It has a complex, opaque, difficult, sclerotic acquisition process. Is the defense Department positioned to take advantage of what AI enabled or assisted technologies provide.

 

SCHARRE:
The DOD has really struggled because one of the things that's difficult about AI is it's coming out of the commercial sector. It's in some ways like the opposite of stealth technology that was invented in Secret Defense Labs and the U.S. had a major lead over competitors. The playing field for AI is extremely level, so U.S. tech companies are in leadership positions, but so are Chinese tech companies. And so both the U.S. and Chinese militaries are working to then bring this technology from the commercial sector into their militaries. Now, the U.S. has been on an innovation building spree building organizations like the Defense Innovation Unit, DIU, but China's doing the same thing. China built a rapid response small group in Shenzhen that was referred to by Chinese commentators as China's DIU. And so finding ways to bring this technology in, break through the red tape, and then figuring out how do you use AI effectively, is very much a level playing field. I mean, it's one where the U.S. is going to have to move faster and be more effective than adversaries like China.

 

LINDSAY:
Ooh, that's actually a pretty complex thing to do because some of the new work being done is being done by small startup organizations that may have a great technological advantage however, they may have no experience or understanding of how to navigate the DOD process. So you can have really good ideas, they can't get from a lab to a battlefield because in essence, the people generating the technology can't fit with the organization that needs that. Now, you've given a bunch of examples of how that has worked or not worked.

 

SCHARRE:
Yeah. The DOD's acquisition system is particularly difficult for startups. So major tech companies like Microsoft and Google or Amazon, they can weather the challenges of the duties' acquisition system. They're big enough to absorb that. But for startups, there's this whole host of challenges that when I talked with startups, they said that they faced. Some of it was just compliance with all of the Defense Department's regulations. One startup I spoke with that was right at the forefront of machine learning, did some of the earliest deep learning applications, taking stuff out of the deep learning revolution and applying it to neural nets for synthetic aperture radar for identifying objects. One of their first hires had to be a compliance officer rather than investing in the technology. And eventually, they sold the company to General Dynamics because General Dynamics had the overhead just to manage compliance with DOD's acquisition system. So that's a real problem because it inhibits the ability of these startups to scale on their own, to grow rapidly, and then to make a more competitive marketplace for the DOD.

 

LINDSAY:
But I would also imagine these technologies can go to the heart of the organizational culture of particular services. You mentioned World War II and the question of developing aircraft carriers. I know part of the inhibition was for navies. There was a element of the Navy that liked battle ships and saw aircraft carriers as basically something new and trivial, not new and transformational. And you tell this wonderful story about a startup company that ends up competing in a DOD competition with fighter aircraft. And at the end, they pit this algorithm, machine learning, whatever it is, against a human pilot. And the algorithm basically defeated one of the best pilots the Pentagon had in a face-to-face simulated match. But that would suggest that if machine learning is so good that it can basically outthink humans, then why do you need a man in the loop or a man in your F-35? But I could imagine if I go to my friends in the Air Force and say, "Hey, we don't need pilots anymore," that seems to run counter to their organization.

 

SCHARRE:
That's right. That's right. And pilots get a bad rap for this because there have been lots of instances. I think the naval aviation community has been unfortunately very effective in strangling the development of drones within the U.S. Navy. But these AI are coming for everyone because the AI technology's really good. And in this instance, DARPA created an Alpha Dog fight challenge taking a page from Alpha Go, that achieved the superhuman performance in a game of Go. They had a competition for companies to submit their algorithms. The winner was a previously unheard of company called Heron Systems, small defense company, that was really early getting a jump on machine learning. They beat up Lockheed Martin in the finals, went head-to-head against a human, and they absolutely crushed the human pilot. Fifteen to zero. Human didn't get a single shot off.

 

LINDSAY:
Say that again. Not only didn't get a kill, didn't get a shot off?

 

SCHARRE:
Well, didn't get any kills. Didn't get any kills. I mean shooting, but didn't actually hit anything. And so the wild thing to me though was that the AI was able to use tactics that humans can't do. So it wasn't just that it was better, it's that it fights differently than people. Now in this case, what the AI did was make these superhuman gunshots when the aircraft are racing at each other head to head, for aviation enthusiasts, forward-quarter gunshots, which are not only basically impossible for humans because there's a split second where there's an opportunity to make these shots, they're actually banned in training because they're dangerous for humans to even try because the air crafts are racing each other hundreds of miles an hour. So that gives an example of how AI has an opportunity to not just be better than people, but open up new ways of operating, new ways of war fighting. And that kind of disruptive change is exactly the kind of thing that U.S. military needs to be in the forefront of.

 

LINDSAY:
But as I think about this, and you've touched on this in your last book, as you start to rely on systems that can in essence, outthink, out anticipate, out act human beings, you can potentially create a whole set of pitfalls. I think you have written about how AI can be simultaneously transformative and brittle, and it can do things that you don't. How do you think about those trade-offs?

 

SCHARRE:
Well, it's a real challenge. One of the problems with AI systems is their intelligence can be very narrow, and so they can be very good at a task that they were trained to do, but if the environment or their context for use changes ever so slightly, their performance can drop off fairly dramatically. So one of the earlier versions of Alpha Go reportedly, if you slightly change the size of the board, its performance will drop off. It didn't have the ability to generalize its knowledge about Go even to just a differently sized board. That's a real problem when you think about, for example, a military environment where we don't get to control where we're going to operate or who we're going to fight against, the enemy's not going to give you a preview of their tactics and the enemy's adaptive. 

 

And so there was this wild story that I heard from a DARPA program manager about a time when humans were able to defeat an AI system that they were training, and they had trained this AI system to detect people walking. Very simple task that's very doable with AI today. The recorded video of a bunch of Marines walking around in a site here in the U.S. and then they trained the AI, this is what a person walking looks like. So at the end of a week of this, they told the Marines, "Okay, now it's your job to defeat the AI." They park this robot right in the middle of an intersection, told the Marines they have to get 300 meters away, to get all the way to this robot and touch it without getting detected. And all the Marines made it doing all sorts of clever things. Two of them summersaulted the whole way. So the AI system didn't know that that was a person, it wasn't trained on people summersaulting. One of them hid under a cardboard box the whole way, just scooting under the box. Another one stripped a fir tree and covered himself in branches and then walked up to the robot.

 

And so obviously any human would notice these things. The human would see the box moving and be like, "Oh, there's somebody under the box," but the AI wasn't trained on that. And so that is a real problem when we think about AI being used in competitive environments because it can be so easily manipulated and people are clever. Our adversaries are clever, and that's a limitation when we think about how we're going to use AI in the real world.

 

LINDSAY:
One other thing about this issue of AI being adapted to the battlefield is whether or not companies are going to want to work with the Pentagon. You've talked about Project Maven. So explain a little bit what happened there and what the fallout has been from the controversy over Project Maven.

 

SCHARRE:
When Google discontinued its work on Project Maven, military leaders panicked that they would be cut out of this game-changing technology that was coming out of the commercial sector. And it wasn't just Google. We saw employees at Amazon and Microsoft also protest working with the U.S. military. Now, needless to say, we're not going to see that in Chinese tech companies. Chinese tech company employees write an open letter of protesting working with the government, they're going to go to jail. U.S. national security leaders were really worried they'd be locked out of this technology and fall behind China. And I think what we've seen is that hasn't been the case, that in fact, there's been a whole slew of new defense-oriented AI startups that are very eager to work with the Defense Department, and all of the major tech companies, including Google, are now working with the military.

 

LINDSAY:
So Paul, as you survey the terrain and there's a lot of things happening, a lot of churn, and obviously in the context of the return of geopolitical competition, the U.S.-China rivalry maps over this technical competition in rivalry as well. What is your advice for the Biden administration, for policy makers here in the United States generally? Are there things that you think in particular they should start doing that they're not doing? Are there things that they're doing that they should stop doing? Are there things that we're doing right that we ought to continue? What's the basic policy advice?

 

SCHARRE:
Yeah, I think that we are starting to see the early components of a competitive strategy in artificial intelligence, but we have a long way to go still. So I'm very excited by Congress's energy behind the 280 billion dollars in the CHIPS and Science Act. Of that, the 52 billion for semiconductor industry here in the United States. I think the Biden administration's move on restricting the tooling to China for semiconductor production was a good one, although I have some concerns about the chip restrictions themselves. But I think that one of the challenges we're going to face is that we can't do everything, we need to be strategic about our approach. 

 

We need to focus on the things that really matter. The two things that matter where the U.S. has a huge asymmetric advantage over China are hardware and talent, and we need to find ways to maintain those advantages and to double down on them when we can. So expanding high skilled immigration, making sure that we're bringing the best and brightest from around the world, and keeping U.S. companies in the lead in some of these critical choke points for semiconductor technology. We don't need to compete on everything with semiconductors with China. We don't need to worry about trying to ... Competing at every single kind of chip at every component. What we need to focus on is maintaining a hub of leading edge manufacturing here in the United States so that we can have U.S. companies at these critical choke points, so that we can control China's access to the most advanced hardware in the future.

 

LINDSAY:
On that note, I'll close up the President's Inbox for this week. My guest has been Paul Scharre, the vice president and director of Studies at the Center for a New American Security, and the author of the new book, Four Battlegrounds: Power in The Age of Artificial Intelligence. Paul, thank you for joining me.

 

SCHARRE:
Thank you. Thanks so much for having me.

 

LINDSAY:
Please subscribe to The President's Inbox on Apple Podcast, Google Podcast, Spotify, or wherever you listen, and leave us a review. We love the feedback. The publications mentioned in this episode and a transcript of our conversation are available on the podcast page for The President's Inbox on cfr.org. As always, opinions expressed on The President's Inbox are solely those of the host or our guest, not as CFR, which takes no institutional positions on matters of policy. 

 

Today's episode was produced by Ester Fang with Director of Podcasting Gabrielle Sierra. Special thanks go out to Michelle Kurilla for her assistance. This is Jim Lindsay. Thanks for listening.