Understanding the Panic About Chinese AI

The release of Kimi’s latest AI model from Chinese company Moonshot AI has reignited debates about U.S. competitiveness and open versus proprietary AI.

There’s been a lot of conversation on social media, but the debate also seems to be brewing in Washington, D.C., where OpenAI and Anthropic have reportedly lobbied regulators with concerns about the open China model.

In the latest episode of TechCrunch’s Equity podcast, Kirsten Korosec, Sean O’Kane and I discuss why this is such an important issue. In addition to suggesting that certain people should “touch the grass” rather than spend their weekends arguing about X, Sean pointed out that in many ways this “feels like a repeat of the previous monster.” Everyone in Silicon Valley “expects something to arrive and blow everything else away.”

And Kirsten pointed out that imposing strict restrictions on Chinese AI models could primarily benefit a small number of companies. “Are we accelerating and ensuring that Americans win the AI ​​race, or are we ensuring that certain pioneering labs perform better than others?

Continue reading for excerpts of our conversation, edited for length and clarity.

Anthony Ha: For those who have been following the discourse on Chinese AI, this will probably be very familiar from the DeepSeek launch, which featured Chinese models by default. In some benchmarks it is competitive, or at least appears to be competitive, even with some Frontier models. Certain parts of the tech industry are losing their minds.

Some of these (discussions) received additional scrutiny because one of the people who posted about them was an (executive) at OpenAI. However, there are usually recurring questions like: Can Chinese companies beat American companies in at least some aspects, and do it in a much cheaper and much more open way?

Sean O’Kane: Yes, there are a lot of elements that feel like repeats of previous monsters. One of my favorites is: Everyone is ready and expecting something to arrive and blow everything else away. And my favorite example from last week was when people were bragging, “Oh my god, Kimi created a full clone of macOS in 30 minutes.” And yes, it replicates the look of macOS with very impressive graphics, but it’s not an OS.

We’re seeing this happen over and over again, something that’s so disturbing to everyone in the tech industry. And I think there’s this expectation, especially with some of the Chinese models that are coming out, and I think that’s the key to why people reacted the way they did last weekend. (By the way, go outside and touch the grass. It’s the weekend. Everyone in the industry has been trading barbs on Twitter all weekend.) But this anxiety was really interesting to me. Because now it’s been a week and no one feels as close to the end as they did a week ago.

Kirsten Korosek: Here’s a really great story from one of our reporters, Tim Fernholz. He’s trying to unravel the psychosis on this issue here in America. He points out several reasons. And then we conclude. I don’t want to jump to conclusions for him, but I think there are some that rise to the top more than others.

There are concerns that these China open weight models may have an implicit bias against China, and there are other concerns about security risks and guardrails. But there is also a very big idea here: protectionism. And who will quote “win the race”? Will it be America or China? And that seems to be driving a lot of what fear is.

I don’t know, Anthony, do you agree?

Anthony: I completely agree. I think the Chinese side always adds this level of hysteria. And that doesn’t mean people shouldn’t worry about how the United States competes with China in various industries. But it’s amplified too much.

Another thing this reminds me of is the debate surrounding TikTok from a few years ago. And again, it’s not that I thought the concern about TikTok was entirely made up, it’s the level of panic that people had. As soon as you add the word China to the discussion, things seem to escalate dramatically. And in this case, it ties into the discussion of openness (weight) and the idea that AI is so powerful and dangerous that the only way we can control it is through the proprietary models of American pioneering companies.

Obviously, most people who say this have their own reasons for wanting to say it. David Sacks, the Trump administration’s AI czar and now serving in other roles in the Trump administration, exclaimed about This is how they argue the position they already have on AI. “Oh my God, if China beat us, that would be ridiculous, so I have to do what I want anyway.”

Kirsten: you’re right. I’m not saying there aren’t real concerns here if there’s a blanket ban on Chinese made open weight models, but I’ll just give it a try. Doing so would benefit models created by OpenAI, for example, and force companies to use them instead of using models like Kimi.

So we really need to ask the question: Are we accelerating and ensuring that Americans win the AI ​​race, or are we ensuring that certain pioneering labs perform better than others?

Sean: At this point we should say that these discussions were actually initiated by Dean Ball, Head of Strategic Futures at OpenAI. Dean Ball was the first to publish a very lengthy post addressing some of these concerns.

I think the reaction to this was because people didn’t agree with what Dean had written. Part of me thinks the reaction was driven by the fact that he just said those words out loud. He basically said that the United States should create regulatory FUD (fear, uncertainty, doubt) that would ruin the ability of these open models to compete with the United States. (Ball later backed down from this claim.)

And I think you can read people’s reactions like, “Dean, you shouldn’t say that out loud.”