OpenAI executive complains about Kimi, berates open-weight AI models, invites mockery

The arrival of Kimi K3, a large language open-weight model developed by a China-based firm, has clearly rattled the executives at OpenAI. One of them just bizarrely accused Beijing of “recklessly” allowing such models to be available to all – and got ridiculed all over the web.

Kimi K3, an open-weight AI system unveiled last week by Chinese AI startup Moonshot, seems to be just as powerful and smart as Anthropic’s Fable and Mythos models or OpenAI’s products – but a lot cheaper.

The industry in the US is reacting just as it did more than a year ago, when DeepSeek, another Chinese AI startup, released its own cheap and effective model. There is widespread concern, if not panic.

This is not exactly surprising. Executives across the country are already unhappy with how much it costs to use the top closed-weight AI models from labs like OpenAI or Anthropic.

And if they see Kimi K3 as a suitable and cheaper alternative, the American companies will find it difficult to attract new customers.

AI models “inherently decelerationist”

One way of helping themselves out seems to be pushing the US government to regulate or even ban Chinese competitors. That’s exactly what OpenAI’s head of strategic futures, Dean Ball, has suggested in a long and pretty rambling X post.

Bean essentially accuses China of allowing models such as Kimi K3 to be open-sourced, “given potential risks.” This position is fair if you’re OpenAI – you need to justify the cost of your pricy systems somehow by implying that true security isn’t cheap.

But then Bean weirdly calls open-weight AI models “inherently decelerationist.” In the world of AI, decelerationism is, of course, an idea that the risks of speedy AI progress outweigh the benefits – that’s why it all needs to be slowed down.

He explains: “One probable outcome of an open-weight-model-dominant world is full AI communism, which is precisely what China proposes: rather than a market product, AI is a ‘public good,’ which will ultimately be provided by the state as a kind of ‘digital public infrastructure.’ This future strikes me as a dystopian hellscape.”

I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models,

Dean Ball

This is a bizarre choice to describe a pretty nice development, we have to say. It’s quite obvious, though, that Ball is addressing the White House.

President Donald Trump and the Republicans hate any kind of socialism and, as it’s become quite clear since the beginning of Trump’s second term, pay meager attention to public infrastructure as well.

“I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models,” Ball thus contemplates.

“You just create enough regulatory risk that every regulated enterprise backs off.”

“Supreme village idiot”

Unfortunately for Ball, the White House also loathes regulation. Yes, Axios just reported that the Trump administration is showing signs it could ban cutting-edge Chinese AI models. This could indeed lock in dominance by OpenAI and Anthropic.

But David Sacks, Trump’s former AI and crypto czar, currently the chair of the President's Council of Advisors on Science and Technology, retweeted a post claiming that even though such a move had been briefly discussed, the US Commerce Department wasn’t actually moving forward on banning Chinese models.

Sacks, who’s still very close to the US President, also said: “The weaponization of regulatory uncertainty as a competitive tool should be completely unacceptable.”

“The leading closed labs, already a duopoly in terms of AI model revenue, want the government to eliminate their open source competition. They have laid their cards on the table. It is time for the rest of Silicon Valley – the vast majority that still values open competition – to do the same,” Sacks added.

The US defense undersecretary Emil Michael was even less kind and called Ball the AI industry’s “supreme village idiot.”

“Dean Ball has perhaps the biggest gap between actual IQ and his own perceived IQ of anyone in the industry (about 40 points),” Michael said.

It’s been tough at OpenAI lately. The company has consistently, for example, claimed it would be making $100 billion a year from ads alone by 2030, but a new analysis has found the projections to be nowhere close to that astonishingly ambitious number.

According to AdWeek, for the bubble not to burst, OpenAI basically needs a series of miracles.

First, advertisers would have to abandon search engines and social media and put all their budgets into chatbots. Then, OpenAI would need to somehow outcompete giants such as Google and Meta. Finally, the whole AI-ad market would have to insanely explode.