AI News

British media says Jensen Huang's departure weakens US open models, benefiting Anthropic and OpenAI

cnBetaUpdated Nov 16 21:13

AI summary

British media reports say that Jensen Huang's departure from Nvidia has left US open-weight models without their strongest advocate and financial backer, while closed-source model developers such as Anthropic and OpenAI may benefit. The report mentions that Nvidia had originally been expected to participate in or lead four startup deals, including investments in Thinking Machines Lab and Groq, a potential partnership with Reflection AI, and the acquisition of Hugging Face for $12.9 billion. Industry insiders believe that investors withdrawing from Nvidia may redirect their funds to these labs; Anthropic launched an IPO this week aiming to raise about $100 billion, and OpenAI is also in the final stage of a $20 billion funding round.

Why read this

AI score 60

The original text sorts out the specific chain of events by which, after Jensen Huang's departure, the open model camp loses its main financial and policy champion, and closed-source labs may take over the funding.

Topics

Tags

Body · AI translation

Jensen Huang's departure benefits closed-source model developers

As Nvidia's former chief, Jensen Huang once used Nvidia's balance sheet to build an extensive AI ecosystem, including investments in and acquisitions of startups involved in developing open-source and open-weight models. Users can customize these models and run them on their own hardware.

Jensen Huang has been a strong advocate for promoting such models. He has lobbied Washington against restricting open models built by China, while pushing the United States to develop its own alternatives.

In his first post on X this summer, he wrote: "The measure of American AI leadership should not be whether we have a particular frontier AI model, but whether America can build a strong open ecosystem and bring it to every industry."

Open models lose their standard-bearer

Jensen Huang's unceremonious departure from Nvidia has left the U.S. open-model field without its most outspoken and influential supporter.

"I think Jensen Huang benefited the most economically (from promoting open models), and he was also personally committed to advancing diversity at the model level," said a venture capitalist who has co-invested in AI startups with Nvidia. "Without him, and with Nvidia no longer being as generous with its investments, we could end up facing a world with a high concentration of model providers."

It is not yet clear whether Nvidia will maintain its rapid deal pace under new leadership. The company had been expected to participate in or lead four startup deals currently under discussion: investments in Thinking Machines Lab and Groq, a potential partnership with Reflection AI, and the $12.9 billion acquisition of open-model platform Hugging Face.

Musk posted on X last Saturday: "Jensen Huang did an excellent job driving supply chain construction at Nvidia, but many of those deals were just verbal agreements he reached with the other parties, not formal written contracts. Without Jensen Huang, at least some of the key deals could fall through."

Jensen Huang described these investments as a way to cultivate U.S. dominance in this technology field. But others point out that there are more direct commercial motives behind these investments: the deals helped Nvidia expand its customer base while also building a line of defense to counterbalance the power of OpenAI and Anthropic.

"In a sense, Jensen Huang's departure from Nvidia is bad for the entire industry. But from another perspective, Jensen Huang was actually a competitor to the frontier model labs, because he wanted models to become commoditized and ubiquitous so that he could continue to occupy a central position in the value chain," said an executive at a top Silicon Valley AI company.

A decade ago, Jensen Huang delivered Nvidia's first "AI supercomputer" to OpenAI, setting the then-nascent startup on a development path that ultimately led to ChatGPT.

At the time, OpenAI's leadership team included Musk, Sam Altman and Dario Amodei. Jensen Huang's relationships with these three underpinned the AI boom, helped fuel OpenAI's rise, and drove the development of SpaceX and Anthropic, founded by Musk and Amodei respectively. Nvidia invested $30 billion in OpenAI earlier this year and pledged more than $100 billion to support the lab's flagship data centers.

Model companies break free from Nvidia dependence

However, there has also been friction between the two sides. The founders of frontier model labs are unhappy with their dependence on Nvidia and are trying to reduce it by building other partnerships and investing in developing their own chips.

"That's why you see every lab and frontier company investing in its own chip projects: it's all because Nvidia's profit margins are too high," said an executive at an AI lab. "There is reason to believe that Jensen Huang's departure will concentrate more computing power in the hands of frontier labs and hyperscale cloud providers. That's good for the labs."

OpenAI developed its own custom chip, Jalapeño, in partnership with Broadcom. OpenAI CFO Sarah Friar has often talked about wanting to ensure the company does not "put all its eggs in one basket" or rely on a single chip supplier.

Anthropic has been diversifying its AI infrastructure suppliers for longer than its competitors, having previously received early backing from Google and Amazon, which enabled it to use their chips. Although Anthropic has disclosed plans to accept a $10 billion investment from Nvidia alongside its IPO, company executives still strongly emphasized during the IPO roadshow that its AI models can switch to rival AMD's chips within days—a process that usually takes months.

Closed-source models benefit

At an event last month, while promoting OpenShell, Nvidia's self-developed open-source security platform for AI agents, Jensen Huang took a veiled swipe at the two frontier model labs' secretive approach to technology.

"How can an opaque agent system, wrapped in an opaque box and locked up by an opaque company, possibly work? Clearly, that model is unsustainable," he said.

In the short term, these closed-source model developers may also gain tangible benefits from Nvidia's crisis. This week, Anthropic will hold one of the largest IPOs in history, aiming to raise a total of about $100 billion. OpenAI is in the final stages of a $20 billion funding round. Industry insiders say investors pulling out of the chipmaker may shift their funds to these labs.

"I don't think open-weight models will disappear... but losing Jensen Huang, their biggest financial backer and advocate, will slow their progress. And it won't hurt Anthropic's IPO," one industry insider said.

Some people inside and around Nvidia insist that despite Jensen Huang's departure, the company's support for open systems will continue.

Thomas Wolf, co-founder of open-source AI platform Hugging Face, said: "If open models had only one supporter, then this week would mean the end of their era. But in reality, open models have millions of developers, and many supporters are inside Nvidia itself."

Source

cnBeta