The former Google China chief and longtime AI investor on Beijing’s open-model advantage, the future of work and why CEOs still underestimate AI.
https://www.bloomberg.com/features/2026-kai-fu-lee-weekend-interview/?accessToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzb3VyY2UiOiJTdWJzY3JpYmVyR2lmdGVkQXJ0aWNsZSIsImlhdCI6MTc4ODU5OTAyOCwiZXhwIjoxNzg5MjAzODI4LCJhcnRpY2xlSWQiOiJUS1RQV0tLSVVQUzgwMCIsImJjb25uZWN0SWQiOiJEMzU0MUJFQjhBQUY0QkUwQkFBOUQzNkI3QjlCRjI4OCJ9.AWy7tstJTITSSm4hhMmD8Oo_TIM0tc6_8eiRbf5yAwM
15 Comments
*Editor-at-Large Mishal Husain for Bloomberg News*
Chinese AI companies are rapidly closing the gap with US rivals, even after years of restrictions on their access to advanced chips. But the race between the world’s two AI superpowers is only one part of a much bigger transformation — of companies, jobs and even how people think about work. Few have watched it unfold from as many vantage points as AI pioneer and investor Kai-Fu Lee, who has worked at Apple and led Microsoft and Google teams in China. His backing of dozens of tech startups has helped create billion-dollar Chinese companies, while his own AI company is 01.ai.
🎁 Gift Link: [The full interview is available to read here.](https://www.bloomberg.com/features/2026-kai-fu-lee-weekend-interview/?accessToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzb3VyY2UiOiJTdWJzY3JpYmVyR2lmdGVkQXJ0aWNsZSIsImlhdCI6MTc4ODU5OTAyOCwiZXhwIjoxNzg5MjAzODI4LCJhcnRpY2xlSWQiOiJUS1RQV0tLSVVQUzgwMCIsImJjb25uZWN0SWQiOiJEMzU0MUJFQjhBQUY0QkUwQkFBOUQzNkI3QjlCRjI4OCJ9.AWy7tstJTITSSm4hhMmD8Oo_TIM0tc6_8eiRbf5yAwM) You can also [listen to this interview](https://omny.fm/shows/the-mishal-husain-show/us-vs-china-cheap-ai-human-love-ai-pioneer-kai-fu-lee) and follow The Mishal Husain Show on [iHeart Podcasts](https://www.iheart.com/podcast/1119-the-mishal-husain-show-300204707/), [Apple Podcasts](https://podcasts.apple.com/us/podcast/the-mishal-husain-show/id1845840408), [Spotify](https://open.spotify.com/show/4cJNIDbIEyNyaS7EWVh5Hh) or wherever you get your podcasts.
I’m fairly convinced the anti AI/Luddite propaganda in the US is being done by China to great effect.
Easy to win a race if the other side lacks education, ultra selfish, and looking for a team to latch on to make it their personality.
Edit: see the regurgitated talking points each of these cultists recite ad nauseum below.
They assume that everywhere in the world has had all layers of elected government fail them and allow cheapest built cost cutting environmental blight in residential areas – while blaming technology and not those responsible for the problem.
Thanks for validating my post drones!
Well, I don’t think there are any intrinsic advantages to open models in the LLM field.
Open LLMs are a concept entirely different from traditional open-source software. They are, after all, not “open source” but “open weights.” AI companies simply distribute the LLMs they’ve developed in their internal labs along with user manuals, while the source code, training data, training methods, and intermediate checkpoints remain confidential.
The approach traditional open-source software took—where a collective effort by an unspecified number of developers worldwide led to better products—and the resulting benefits do not apply to “open-weight” LLMs. There is also no pathway for an unspecified number of external developers to participate in or contribute to the training process of the next generation of LLMs.
Just because open-source software has been successful does not mean that “open-weight” LLMs can be successful in the same way.
P.S. The following are examples of truly open-source LLMs.
[https://ifm.ai/k2/](https://ifm.ai/k2/)
However, even in these cases, there is no pathway for an unspecified number of external developers to participate in the development of IFM’s (the developer of K2 Horizon) next model. They can simply take the source code and use it as a basis to develop new AI if they wish.
I hope they “win” this race just to silence the leaders of our braindead country.
“win”? That’s what we care about? I’m being rhetorical.
It’ll be cat and mouse. When US has been ahead, China has copied things we do and technologies we’ve had to catch up and close the gap over time. There’s been studies about this. My guess is with AI the time in between those gaps will be much shorter.
All so we can achieve ASI that’ll be a million times more intelligent than us to see if we’ll achieve immortality or extinction.
For China, the chance to win long term on AI is winning through electricity price & efficiency and AI Deployment in daily life and infrastructure
Win what? And why? For what perceived purpose or benefit to the “winner”?
No one fucking knows, because the benefits of rapid A.I. development are theoretical at best, and fantasy at worst, so this is nothing but a dick-waving contest, a race to nothing, so China in particular can continue to say how great it is with nothing to show for it. It’s a propaganda machine, with infinite output, because they’re selling theoretical possibility, making up what can happen, and when you don’t deliver on possibility, you can just say those promised results will come later, and be even better then. What are they going to do if it amounts to nothing of value, say “whoops, guess we misjudged that one”? Of course fucking not; they’ll just promise more, placating the masses.
Nothing but small men padding their egos, as they dupe the average moron that never fucking questions it.
As with any war or conflict, there needs to be a DEFINITION of what counts as victory.
The lesson from Vietnam was, DoD defined “victory” as the North Vietnamese having a higher body count than the U.S.
The NVA and the Viet Cong were focusing on morale and mobilizing the population against the U.S.
It was clear who chose the right metric.
I agree with him, The US currently holds the lead in raw, cutting edge innovation and computing power, however, innovation without scaling is just a lab experiment. China excel at rapid, hyper-efficient commercialization. and If a country can deploy slightly less capable AI into its factories, supply chains, and public infrastructure at half the energy costs, then it achieves a massive economic multiplier.
also the American $20/month SaaS subscription model is very limiting for mass adoption and It treats AI like an elite corporate tool or a luxury hobby rather than a foundational public utility
don’t care if China win or not, China needs to do something to crash the stock market like last time, it’s damn too high
China doesn’t have to win, the US just have to lose. That’s because China has not invested as much as the USA have.
When the AI bubble pops, the USA will be damaged more economically. It might not be tomorrow, but it will eventually pop. As all overvalued markets always do.
Doesn’t help that open weight models kinda fucked over the plan to monopolize AI models. AI companies heavily subsidized the usage of AI to increase the market share, only for people to have the option to switch to Open Weight when the subsidy ends and the frontier models become unaffordable.
I’m unclear if anyone here has actually read any of his writings but he’s basically been wrong about everything from the get go. Pre OpenAI he predicted china would crush the AI race and that hasn’t happened at all. He basically completely failed to predict that china would continue to be second place as its copying what American labs are doing.
His book was called AI superpowers or something like that
This is genuinely the most valid and helpful article I’ve read for anyone trying to figure AI out. His points on AI mastery, which is not coding but strategic deployment, is 100% on the money.
USA has been so firm on winning the race, China simply has to sit back and see what is working and what is failing.
I know nothing of brain surgery, but if I am able to watch several videos of surgeons performaning, I am more likely to do it better and make less mistake than if I didn’t.
Now factor in counter intelligence and false bread crumbs, control over firmware or hardware, education and personnel.
China also went in on automating and improving on smaller scale application which reduces the impact to human jobs.
The dude with hige investments in China AI labs says China AI labs will win, color me surprised.