Can Trump contain China's AI boom?

Australia Broadcast Center

3 days ago

Sam Hawley: For so long, the tech bros of Silicon Valley have dominated the AI race. Now there's a boom underway in China, giving them a run for their money and Donald Trump doesn't like it. Today, Kyle Chan from the global policy think tank, the Rand Corporation, on why the president is so desperate for the US to beat Beijing. I'm Sam Hawley on Gadigal land in Sydney. This is ABC News Daily.
Sam Hawley: Kyle, there's a global race going on right now to be the world leader in AI. This is a race basically to make technologies that rival the human brain, right?
Kyle Chan: Yeah. So there really is this global race. And in particular, you have the US and China with many of the world's best AI models. And it's quite impressive to see almost every day, it feels like a new model coming out with new advanced capabilities. So, yes,getting close, if not even beating what we can do ourselves.
Sam Hawley: Yeah, it's fascinating to watch how quickly this is moving.
Donald Trump, US President: I don't like the name artificial anything because it's not artificial. It's genius. It's pure genius.
Sam Hawley: Donald Trump, of course, wants to make sure that America wins this race.
Donald Trump, US President: America is the country that started the AI race. And as president of the United States, I'm here today to declare that America is going to win it. We're going to work hard. We're going to win it.
Sam Hawley: He even gave a speech with that title, winning the AI race.
Donald Trump, US President: Because we will not allow any foreign nation to beat us. Our children will not live on a planet controlled by the algorithms of the adversaries advancing values and interests contrary to our own.
Sam Hawley: He's pretty invested in this, isn't he?
Kyle Chan: That's right. Yeah, this has been a big topic throughout his administration so far. I think a lot of what US policy is focused on, including the current Trump administration, is on winning the race to AGI. I think there's a strong sense that this could be a pivotal turning point.
Sam Hawley: Remind me, what is AGI?
Kyle Chan: So artificial general intelligence. There's this idea that perhaps one day it could reach a point where it could replicate or even exceed the abilities of humans to do, say, certain kinds of office work or certain kinds of research. This could even extend into areas like military capabilities, like autonomous weapon systems, for example. Reaching this stage where AI is as good as, if not better than, human reasoning.
Sam Hawley: So we might not be needed actually anymore. We won't need to think anymore, right?
Kyle Chan: We'll see.
Sam Hawley: Exactly. All right. Well, Kyle, of course, up until now, the US has really dominated this market. All the big tech giants who've developed AI, things like ChatGPT, they're sitting there in Silicon Valley.
Kyle Chan: Oh, yeah. So you have OpenAI, currently led by Sam Altman. You have Google, which has been coming out with a number of various sort of cutting edge models with its Gemini series. You have Claude, which is very well known from Anthropic, well known for its coding capabilities. You have Meta as well as xAI. So there's actually quite a quite a large roster of strong American AI companies.
Sam Hawley: So for many, many years, America's really led the world when it comes to AI development. But as you say, China has been creeping up on it. And that has the US administration a bit worried. It even tried to stop Beijing's advancement in this space, didn't it? By banning Nvidia from selling advanced chips to China. Just remind me what happened then.
Kyle Chan: Yes, that's right. So this was actually in the Biden administration. You had very strong export controls placed on especially Nvidia's more advanced chips. And so here you actually have several rounds of downgrading of what kinds of Nvidia chips could be exported in China.
Sam Hawley: So the US, in part, was saying that it was deeply concerned that AI could be used by the Chinese for military purposes.
Kyle Chan: That's right. Yeah. And, you know, to be sure, it was also part of this broader idea that advanced semiconductors in general can be used for a whole range of important applications. So in addition to AI, there are also more direct military implications for this ban.
Sam Hawley: All right. So Biden brought in this ban to stop these really advanced chips from being exported from the United States to China. But intriguingly, Trump just recently has now removed that ban. Do we know why he did that and how significant is that decision?
Kyle Chan: Yes. What's interesting is I think whereas before people expected maybe a continuous ratcheting up of these export controls, Trump has reversed the ban on the H20 chips. Interestingly, a new line of argument has gotten a lot of prominence, which is that Nvidia and other US tech companies who sort of, you know, quote, unquote, sell the picks and shovels, that is, build the infrastructure and build the sort of underlying platforms for AI development, that American companies should be the ones who are dominant in the world and that people should build on the American tech stack as it were, rather than cede, say, the Chinese market to its competitors like Huawei, which is also developing its own AI chips. So the idea here was that rather than block out the Chinese market entirely, that the US should stay engaged, at least in terms of providing some kind of sweet spot of infrastructure, but not not too advanced in order to actually accelerate China's efforts.
Sam Hawley: And that's the argument that the Nvidia boss, Jensen Huang, has been making to Donald Trump.
Jensen Huang, Nvidia CEO: This is a once in a lifetime opportunity for America to have AI technology leadership. This is a once in a lifetime opportunity for China to have AI leadership. And if we want to be a leader, we have to engage developers all over the world. We have to engage markets all over the world.
Sam Hawley: So the best way to beat China at AI is to actually help China to compete. Have I got that right?
Kyle Chan: Yes. Yes. The logic, it can be hard to parse out sometimes, but yes, this is one of the main arguments.
Sam Hawley: All right. So, Kyle, Donald Trump, he's delivered this speech outlining the importance of the US dominating the AI market.
Donald Trump, US President: America needs new data centres, new semiconductor and chip manufacturing facilities, new power plants and transmission lines. And under my leadership, we're going to get that job done.
Sam Hawley: But as we mentioned, in China, it's full speed ahead. There really is a boom going on there right now, isn't there?
Kyle Chan: Absolutely. Yeah. So, I mean, everyone now knows about DeepSeek and the DeepSeek moment.
News report: The release of a high performing Chinese rival to chat GPT has sent shockwaves through the global tech sector and caused US tech stocks to fall.
Kyle Chan: A Chinese AI model for the first time seemed to be almost on par with the US leading models. And this was done at a fraction of the cost in terms of compute. And this was done sort of in defiance of US efforts to put on export controls and to restrict Chinese compute capacity. But DeepSeek is really the tip of the iceberg. So there's a whole set of very competitive Chinese AI models. You think about Alibaba's Qwen, you think about Tencent, ByteDance. There's Moonshot, a whole host of startups as well. Most of these companies now, they all have their own sort of chatbot like chat GPT, where anyone can download the app or go to the website and just start chatting directly with the AI model, the underlying model itself. And so what's interesting is that it's not just one company or one startup per se. It's actually a whole sort of lineup, in a way, a Chinese team competing with the US one.
Sam Hawley: And these AI apps, what they haven't needed, that chip that was banned, I guess, from being exported from the United States. China's done it on its own, has it?
Kyle Chan: Well, yeah. So it's complicated because actually many of these Chinese AI companies, they do use Nvidia chips. They do, including the chip that was banned, the H20. At the same time, though, they're trying to experiment and test Chinese domestic alternatives, knowing very well that, you know, in the long run, they may no longer have access to Nvidia's GPUs. So there's a question right now within the Chinese tech community, Chinese AI policy about how hard to push for this domestic alternative versus to continue to rely on what are otherwise better performing Nvidia chips.
Sam Hawley: All right. Well, Kyle, just unpack for me now. What's actually driving this AI boom in China? Because it has a lot to do with the Communist Party's backing of this, doesn't it, of the government's funding of it.
Kyle Chan: That's right. So what's interesting is that Beijing is pouring resources into the entire, what I call the entire AI tech stack. So they're investing in not only chips, as we mentioned earlier, but in the rollout of data centres, often tied to renewable energy. They are investing in the development of foundation models. They have special local government AI labs. And then all the way to applications, especially in so-called hard tech areas like robotics and industrial automation. So you can see sort of this full range of support. And of course, at the very heart of this, I think is ultimately the emphasis on talent development and basic research. So a lot of the universities in China, many of them are producing really world class AI developers.
Sam Hawley: And we've seen this before, haven't we? From the Chinese government when it wanted to boost the EV market. It did the same thing. It did the same thing with solar and it works.
Kyle Chan: Yeah, that's right. They've tried this playbook before and they're going to try it again. But the funny thing is, yeah, AI is sort of a different beast. And so, you know, for example, just in the past year, we have this shift towards reason models. And that already has thrown a bit of a wrench into some of the industrial policy efforts that China has made in AI. So some of the data centre build out that was government backed. You know, there's a question now about whether that is fit for purpose with the shift towards this sort of new AI paradigm. And it could change again. So it's a fast moving space.
Sam Hawley: All right. So, Kyle, there is this race going on between the United States and China to dominate AI development. But tell me, why is that so important? Why does it matter who wins this race in the end?
Kyle Chan: So the AI race, I think, is especially important now because it has implications for economic growth, long term productivity. There's a sense both in the US and in China that AI could help boost a whole range of sectors. From education, health care, biotech, drug discovery, manufacturing services. So on the one hand, you have this sort of economic implication. On the other hand, there are military implications. So AI could be used for developing autonomous systems. You think about drones or swarms of drones that are able to navigate on the battlefield on their own. Or you think about missile defence capabilities that might use AI or satellite technology that might use AI. So there are both security and economic repercussions for, you know, the question of sort of who is ahead in the race for AI.
Sam Hawley: Yeah. And I note that Sam Altman from OpenAI says he wants to make sure that democratic AI wins over authoritarian AI. What do you make of that?
Kyle Chan: Yeah, that's right. I mean, it's an interesting idea because right now there's also this battle over sort of diffusion and who can get their models out into the world. And so it's not just a matter of, you know, who has the best model, but also which model is more widely used. And I think right now what's interesting is a lot of Chinese models are open source or at least open weights. That is people, companies, organisations, individuals can download these models and run them locally, run them themselves. And what this means is that a Chinese type of AI might end up diffusing more broadly, perhaps maybe outside of the U.S. into other countries.
Sam Hawley: All right. Well, it's a fascinating battle. Kyle, what do you think? What's your prediction? Who's going to come out on top in the end?
Kyle Chan: In a sense, I do see that with some of the industrial policy in China, with some of the government support, as well as perhaps more importantly, different sorts of attitudes towards AI in China. There are some surveys that have shown that people in China more broadly seem to be more open to adopting AI and see it as a more positive force in society. That could play a key role in rolling out and incorporating AI into more areas of life. So that's one area that I would watch very closely.
Sam Hawley: Kyle Chan is a postdoctoral researcher at Princeton University and an adjunct researcher at the Rand Corporation. This episode was produced by Sydney Pead and Sam Dunn. Audio production by Cinnamon Nippard. Our supervising producer is David Coady. I'm Sam Hawley. ABC News Daily will be back again on Monday. Thanks for listening.