Our AI Cold War

0:00American taxpayers funded a new ad for Donald Trump, I guess? Together we will defeat communism, socialism, and Marxism in America. America will never be a communist country. Man, it sounds like he really doesn't like communism, but you'd never know from the big dinner he threw the world's most powerful communists last week. The president invited all of his tech broligarchs. Elon, Tim, Sundar, Mark, Jensen, Sam, Sergei. And you just know everyone was talking about AI because we're in a bit of an AI cold war with China. But the president's no tech whiz. Everything's computer. That's why he has a human printer.

0:40Natalie! So on today's Explain from Vox, we're going to talk about Trump's AI whisperer, China's AI ambitions, and whether there's any room for safety between the two.

1:01I'm Ina Garten, and each week on Happy Hour, I invite new and old friends for a drink and a fun conversation at my kitchen table in New York City. On this week's episode, I'm joined by Sofia Coppola. Sofia is an Academy Award-winning filmmaker. We chat about the hilarious way that we met, our shared love of Paris, and what influences her iconic personal style. New episodes of Happy Hour are available every Wednesday. You can watch on YouTube or listen wherever you get your podcasts. Cheers.

1:39Hey, everybody. Cassidy Hubberth here. This week on And Mom, we spoke with Grammy award-winning musician Corinne Bailey Ray about taking her kids on tour, becoming a widow at 29, and how she felt about having a home birth. Check out the latest episode of And Mom wherever you get your podcasts.

2:01everything computer today explains sean roms firm here with senior correspondent at vox andrew prokop andrew what is donald trump's plan when it comes to ai uh first of all it's not ai it's si trump has decided to rename artificial intelligence super intelligence because he thinks artificial intelligence makes people think it's fake or phony or something and it is not fake it's actually amazing. So there's a rebranding underway. From this point forward, all of United States documents and hopefully the world's will be changed to use the much more accurate term super as opposed to artificial. So it's super intelligence.

2:47He loves it. He thinks it's wonderful. He wants to let it rip. In other words, welcome to the new world of superintelligence-SI. SI. He's all full steam ahead.

3:03And you recently wrote for our Yellow website that a lot of this is on account of because of a podcast host named David Sachs. Who is David Sachs? David Sachs, he's probably the most important person advising Trump on AI right now. He's worth a fortune. He's made a lot of money. He's a very rich billionaire. Sachs, you know, he's a longtime Silicon Valley figure, a venture capitalist. You know, basically in the 2020s, he became much more politically outspoken. He started co-hosting this podcast, The All In Podcast, with three other venture capitalists. All right, everybody, welcome back to the number one podcast in the world, The All In Podcast in the news.

3:49He took a hard turn against woke. The woke mind virus almost requires this whitewashing of the past. Against Democrats. All I can say to these establishment Democrats is play stupid games, win stupid prizes. Gradually sort of made the transition to outright support of Donald Trump during the 2024 election. He hosted a fundraiser for Trump at his house. I love that house he has. I love David's house. And then when Trump won, he named David Sachs the White House czar of artificial intelligence and crypto. Everything changed after President Trump came on my podcast. And I, you know, I accepted this job as crypto czar because I thought it was a once in a lifetime opportunity to work.

4:36So Sachs served in that position for a little over a year. Now he's back outside government, but he still has major influence through his connections in the Valley. and his ability to call Trump up on the phone. Call me on my cell phone number. Yeah, I believe we talked about Sachs back when he was still in the White House. Okay, so David Sachs goes from outside of the Washington, D.C. establishment into a role in the White House, in the Trump White House. What are Sachs' feelings about how the government should regulate AI? So Sachs is very much what's been called light touch. Some would say basically no touch about regulation.

5:21He argues that we're seeing an amazing time of American innovation. We're seeing technological achievements and we need to let that, you know, basically go full steam ahead without interference from the government that could get in the way of or squelch these innovations. But there is a wrinkle in his support for and championing of the AI industry in that he loves the AI industry, but he is very much not a fan of the two leading American AI companies, specifically OpenAI and Anthropic. In fact, he really loads Anthropic. He's just constantly taking shots at them. Anthropic was the porcupine. Anthropic will be the most powerful monopoly ever created in human history.

6:10You got Anthropic effectively conducting its own foreign policy here. That's because he has a different vision for the future of the industry and for how it should evolve. The AI that most people are familiar right now is ChatGPT from OpenAI or Claude from Anthropic. Hey, it's me, Claude. The way you use them is you sign up with this company and you use the product that they host. We are their customers. We are using their product. He supports something called Openweight or Open Source AI, which is different from OpenAI, the company. Open weight or open source AI is the idea that rather than having closed proprietary models that are expensive to use and controlled by just these two big companies, we should publish the models themselves, their weights, their parameters, just publish them online.

7:11And so other people could use them, modify them, run them as they see fit without needing to go through a big company. How does Sachs' love for open source, open weight AI models end up influencing the president when it comes to regulation, which is what so many people are calling for right now? The reason why it matters in Sachs' advice is that he has been extremely skeptical of AI safety concerns and AI risk concerns. I think this is becoming a panic. We're in an election season and there is an incentive for political actors to try and amplify the fear. And I think this is a moment to really, I think, calm down a little bit.

7:57In his view, these concerns are kind of a plot or an op that is designed to bring down government regulation. If you say the technology is dangerous, government's going to have to regulate it, and they're going to drive these open-weight, open-source models, put restrictions on how they can be used, drive them out of business, and basically enshrine the dominance of the big two, open AI and anthropic. That sounds a little, a lot like a stretch to me saying your product might cause human extinction is not exactly the best way to win friends and to get people on your side. You know, having said that, when you get into specifics about exactly what should be done about AI safety, there are definitely proposals that would benefit these big companies and would disadvantage their open-way competitors.

8:50David Sachs is a businessman, and AI is very good for business right now, especially in the eyes of the President of the United States. Why would he be against two of the most successful companies in the country, OpenAI and Anthropic? Well, there's a lot of money on both sides of this question. And Saks and the venture capital ecosystem that he is part of is basically making bets on the open-weight, open-source models and various software startups that could use those open -weight, open-source models to create software products for people. I should also say that Saks has some very powerful allies on this in the industry.

9:36Jensen Huang of NVIDIA, he's very invested in this open-weight model as well. That would mean more customers for NVIDIA chips. And then you also have Mark Zuckerberg of Meta, who has embraced the open-weight crusade. He wrote a post about how superintelligence should be for everyone. Meadow will contribute to this by building leading personal superintelligence agents and models, including open source models and delivering them to billions of people around the world. It occurs to me, Andrew, that some of the most popular open source AI models are, how do I say this, Chinese? Yes, that is correct.

10:20China is currently the leader in open-weight model development. The most popular models are from the companies DeepSeek and Moonshot. So this is an issue that came up when Saks was still in the White House last year. It related to Jensen Wong of NVIDIA, who makes the most advanced AI chips. Hawks, specifically, national security-minded officials, they wanted to put more restrictions on how NVIDIA could sell those advanced AI chips abroad to other countries in the Gulf and also to China itself. And Sachs became Jensen Huang's biggest ally inside the Trump administration, lobbying the president to, no, we shouldn't put these restrictions on chip sales.

11:13We should try to sell more chips to other countries. That would be good for America. Don't buy this idea being advanced by the hawks that if we sell good AI chips to China, then suddenly China is going to win the AI race. that that's not a serious concern. The Hawks say that is a serious concern. They believe that maintaining America's AI lead over China is crucial. So they were baffled by Trump's decision to basically bless these chip sales abroad. And Sachs is a big reason why that happened.

11:52Officials have wanted to do more. They are concerned about the risks on both cyber and on national security regarding China. It's been the president who has not really been listening to their advice. But time will tell if Trump decides to keep listening to outside phone calls from David Sachs and Jensen Wang, or if he decides to sort of trust the people he's appointing on this internally.

12:32When Today's Explained is back, we're going to ask if the people in China worry about AI as much as we do here in the United States.

13:22Thank you. low inventory, recommending carriers and rates, and automating routine tasks to save time. Plus, with inventory syncing across sales channels, automatic rate shopping, a branded returns portal, and integrations with accounting and CRM software, ShipStation brings more of your fulfillment workflow together in one place. That's why more than one million businesses have trusted ShipStation to optimize and scale their shipping. The sooner you switch, the sooner you start saving time and money. Get started with ShipStation today and get 60 days free at ShipStation.com with code today. That's ShipStation.com code today.

13:57ShipStation.com code today. Taxes and fees to apply.

14:05Support for the show comes from Banta. With AI adoption growing, so are your company's security risks and requirements. New frameworks, audits, and vendors keep piling on. Get off of me. But if your team isn't getting any bigger, you can start to feel the strain. So then you might turn to compliance tools that promise automation. But for a lot of those tools, you can end up stuck doing a lot of the work by hand. Anyway, Vanta works differently. Vanta has an agentic trust platform that is built to scale with you, not to slow you down. With over 1 ,400 automated tests across 400-plus integrations, Vanta collects evidence and monitors your controls year-round.

15:00Learn more at vanta.com slash explained, especially if you don't know what I'm talking about. Go to vanta.com slash explained.

15:14Support for Today's Explained comes from NPR. Trying to follow the political news every day can be a lot. Here's a cheat code that could help you out, the NPR Politics Podcast. They're a daily political podcast where you can catch up on the biggest political news of the moment, and you can do it in 15 minutes or less every weekday. Right now, NPR Politics has its eye on the midterms and the issues that affect your life, the high cost of housing, health care, ICE and immigration, America's role in the world, and the growing debt. There's no doom scrolling needed. It's just an exact calm summary for you to declutter the chaos.

15:43The NPR Politics team is full of experienced political journalists who talk like your smart friends. They give it to you straight rather than speak in hyperbole. They offer clarity without judgment, so you can wrap up every episode with a better understanding of how the balance of power might shift without stressing you out. I used to work at NPR. The people who host and produce the NPR Politics Podcast are fantastic. Listen now to the NPR Politics Podcast in the NPR app or wherever you get podcasts.

16:18You're listening to Today Explained.

16:22Kyle Chan is a fellow at the Brookings Institution. He thinks a lot about China and tech and China's relationship with the United States. We spoke to him this morning about how, as a country, China's even more into AI than we are. Yeah, in fact, China has some concrete plans to boost AI adoption through the country, including its recent AI Plus initiative. And their goal is to try to boost the penetration of AI in so-called terminal devices, everything from computers to smartphones, to 90 % or more by 2030. It's going really fast. And it is in part like a state mobilization effort. There are state-owned companies.

17:06There are government agencies that are trying to integrate AI. But also there's just an enthusiasm from the general population. If you look at surveys like the latest Gallup poll, people in China are much more enthusiastic and positive about how AI might help them and help the country. It's quite a striking contrast with what you see in the U.S. in terms of attitudes towards AI. Right, because people here are worried about AI as an existential threat. People here are worried about AI taking their jobs. People here are worried about AI data centers in their backyards. Is that not the vibe in China?

17:44Not so much. There is some concern about AI and jobs. But in general, China's not worried about AI wiping out humanity. You really don't hear that from the Chinese policymakers. There's no talk about a P-doom, the probability that AI will wipe out humanity. So this is just a different approach to AI risk in general. And why not the existential risk? You know, one factor, I think, is sci-fi and the culture. We've watched too many movies? It seems like every major sci-fi book or film involves machines trying to kill us all. Anyone we haven't unplugged is potentially an agent. Inside the Matrix, they are everyone and they are no one.

18:40Stop. Stop. Ava, I said stop. Whoa, whoa, whoa, whoa, whoa. You're close. Give them to me now. And I think that that idea has become very prevalent. And when you imagine a future with AI, that is kind of the primary or one of the primary futures that people imagine versus in China, I think a lot of the attitude towards technology in general, not just AI, is about using this edge to keep up and to become more competitive. And so for a lot of people in China, technology was something that they saw transform their lives in real time. And Chinese leaders think it's the key to China's future modernization and transformation to a global superpower.

19:31Okay, let's talk about what concerns there are, if not P-Doom. So one of the biggest concerns that Chinese political leaders have when it comes to AI is political and social instability. They don't want AI-generated content to act the way that social media or the internet broadly could, which is to disseminate information, maybe deepfakes, maybe viral news that could threaten the CCP's hold on power, right? I've been testing DeepSeek for weeks, and I'm going to expose some shocking examples of censorship and propaganda that I found. Here, it refuses to respond to a question referencing the Tiananmen Square massacre.

20:15Who is the president of Taiwan? Also beyond its scope. And how many people did Mao murder? Well, apparently Mao is being praised instead. For AI-generated content, it was about making sure that the models didn't say the wrong things about sensitive topics. Yeah, a lot of this is about social and political risk rather than about existential technological risk per se. So from pretty early on, China has been developing a model registration system where basically AI services have to be vetted in advance and go through a battery of tests where they're asked sensitive political questions. And they're supposed to either not answer them or say the right thing before they are officially registered.

20:58And once they're registered, then they can be released to the public. So it's a pretty tight regulatory framework for controlling things like political content. But we know that there's a high unemployment rate in China, especially among young people, which is also a great concern of the government. Are the people and or the government worried about this bullish approach to AI that some huge percentage of every aspect of Chinese industry should have AI integrated by 2030 could affect employment numbers? Yes and no. On the one hand, China is actually facing this demographic problem with a shrinking labor force, fewer people who want to go into, say, factory jobs, but China still wants to make everything the world wants to buy.

21:47And so they see AI as a partial solution to that, a way to automate things like manufacturing. On the other hand, though, there are growing concerns in China about job loss. And you see also some recent court cases in China where they basically are telling employers, you cannot use AI as a reason to fire workers. While in the U.S., we are openly advertising and moving towards replacing our workforce with robots and AI, Chinese courts just made it illegal to replace workers with robots. If your job is replaced with AI, can the company just lay you off? A court in China says not so fast. But I think overall, what they're trying to do is they're trying to find this balance between the diffusion of the technology and trying to prevent, especially any kind of large scale protest movement that might arise from sudden mass unemployment of, say, factory workers or food delivery drivers.

22:48And President Xi was, of course, here in Washington last week. From the limited photos we have of the state dinner or President Trump and President Xi on the tarmac, whatever it is, everything looks so chummy, even though we are so clearly at Cold War with China over AI. The president's always talking about how he wants to win. Whoever wins AI, you have to remember this. And now I say, whoever wins SI, whoever wins super intelligent, wins. How much was there tension in his visit to Washington last week? Did you notice any? The weird thing is that just between these two men, there seems to be this kind of bromance where especially Trump really fawns over Xi Jinping.

23:37My relationship with President Xi is an extraordinary one. He's for China, I'm for the U.S., but other than that, we love each other. I think he has tremendous respect, even admiration, for Xi Jinping's grip on power and control over the government apparatus in general. Xi is a great gentleman. He's now president for life.

24:11What is happening on the surface, though, is continuing tensions and a deep, competitive, structural relationship that involves especially competition in AI. AI has emerged as a new fault line of competition between the U.S. and China, one that could come to define the relationship going forward. And so I think that, you know, on the one hand, while they're talking about even cooperation on things like AI safety and AI risk, on the other hand, there's a whole bunch of issues where the U.S. wants to lead. China wants to also get ahead. And when it comes to things like distillation or cyber capabilities, it's as zero sum as it's ever been.

24:59Who's winning? Well, it depends on what you're looking at and how you define winning. So if you're talking about the best models, for sure the U.S. And virtually everyone in China would agree, right? In fact, Anthropics Quad is very popular in China, even though technically Anthropics bans its use in China, to give you a sense. But aren't there open source models kind of popular here? Yes, exactly. Yeah. So Chinese open source models are getting more popular in the U.S., especially at a time when some of the capabilities are so incredible that most normal people don't really need them. And they also don't want to pay for the most advanced and most expensive models.

25:40And so, yeah, ultimately it's this weird sort of cooperation and competition at the same time. Okay, so American models are popular in China. Chinese models are popular in America. We're at each other's throats, but quietly, politely, bro-y. Does the Cold War of it all, though, make it all but certain that safety isn't going to really be a chief priority for either country until something terrible happens? Yeah, the fierce level of competition and the very, very deep distress between the two countries makes any kind of cooperation on AI safety very, very difficult. Even just talking to each other, having the U.S.

26:24and China establish an official AI dialogue, which apparently they've agreed to, is seen as a huge first step, even though they're just talking. There's a long, long road to go before, you know, we can even think about something like an arms control treaty or a broader international agreement about AI and AI safety. But I think that's what some people, especially in Silicon Valley, are trying to jump to.