Hugging Face CEO: We are in a "large model bubble," not an "AI bubble," and this bubble is about to burst.

```

Hugging Face co-founder and CEO Clem Delangue believes that the current market is experiencing a "large model bubble" rather than an artificial intelligence bubble, and that this bubble may be about to burst. However, he emphasizes that even if the bubble bursts, it will not pose a major threat to the AI industry as a whole.

Delangue said at an Axios event on Tuesday that large models have received excessive attention, with all focus and funding concentrated on the idea of building a single model with massive computing power to solve every problem. He expects this bubble could burst next year.

This AI platform leader points out that large models are only a subset of AI, and applications of AI in fields such as biology, chemistry, imaging, audio, and video are just beginning. In the future, more customized and specialized small models will emerge, solving specific problems at lower cost and faster speed.

Delangue stated that Hugging Face still retains half of its financing (about $200 million) in the bank, a cautious capital strategy that stands in stark contrast to other AI companies that spend billions of dollars.

Large Models Are Not a Universal Solution

Delangue believes that large models are not suitable for every scenario, and that smaller, more specialized models will see broader use in the future.

He illustrated with the example of a bank customer chatbot:

You don’t need it to tell you the meaning of life. You can use a smaller, more specialized model that’s cheaper and faster, and perhaps you can even run it on your own enterprise infrastructure. I think that’s the future of AI.

He points out that the current market's problem is that all resources are bet on one assumption—that a single model built with massive computing power can solve every problem for every company and individual. But in reality, in the coming months to years, the market will see a proliferation of customized, specialized models addressing different problems.

AI Industry’s Diversity Provides Cushioning Space

Delangue admits that the bursting of the large model bubble could affect Hugging Face to some degree, but he emphasizes that the AI industry is vast and has already diversified. Even if one part of the industry (such as large language models) is overvalued, it won’t have a huge impact on the overall AI field or its business.

He says that large models are only AI’s application in text processing, while applications of AI in biology, chemistry, images, audio, and video are still in their infancy, with further development expected in the coming years.

Capital Efficiency Oriented Long-term Strategy

Hugging Face has adopted a capital strategy distinctly different from other AI companies. The company has raised $400 million but retains half of that amount. Delangue said: “By AI industry standards, that’s called profitability, because other companies aren’t spending hundreds of millions—they’re spending billions.”

“I think a lot of people are acting in haste—maybe even in panic—adopting very short-term strategies,” Delangue said. “I’ve been in AI for 15 years and have seen some cycles. We’re learning from that, trying to build a company that has a long-term, sustainable, far-reaching impact on the world.” This approach of capital efficiency represents a more prudent path of development.

Risk Warning and DisclaimerThe market comes with risks, investment requires caution. This article does not constitute personal investment advice, nor does it take into account the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investment decisions are at your own risk. ```