If Anthropic's Dario Amodei is the tech leader most stoking fears about the power of AI, Nvidia's Jensen Huang is increasingly becoming the industry's most prominent voice pushing back.

At one of the season's signature tech conferences, Huang on Tuesday held a large audience rapt discussing the positive potential of artificial intelligence with Salesforce Chief Executive Marc Benioff. As the two awkwardly navigated walking and talking through the audience, the Nvidia CEO directed Benioff to follow him.

"Jensen," Benioff replied, "we're all following you, bro."

The audience erupted into cheers.

While it was an off-the-cuff comment, in so many ways, the moment captured a sentiment in Silicon Valley these days when it comes to the fierce and confusing debate between AI accelerationists and doomers.

For many, Huang, 63, has become the voice of reason standing up against the scary rhetoric coming from the likes of Anthropic.

The Anthropic CEO's most recent warnings-in a lengthy essay a weekend ago-helped further inflame a national panic about the dangers of artificial intelligence. Amodei, 43, called on the industry to collectively slow the development of advanced AI models but hasn't paused his company's own work.

Others in tech have pushed back against Anthropic. Yet Huang's growing role as statesman for AI accelerationism became evident this past week in numerous venues. He was teaming up with President Trump to label AI concerns as a hoax. And he was attending King Charles III's AI summit in Scotland where he cautioned against conflating the technology of artificial intelligence with a specific AI product.

In essence, his messaging throughout this past week suggested AI labs hold the responsibility to ensure the products they release are safe. Time and again he was pushing back against the growing calls for new AI laws and regulations.

"Safety is paramount. In a lot of ways, it's job one, however safety is an engineering problem," Huang said at Salesforce's Dreamforce conference. "If you build a product or a service and you're not confident in its functionality, capability or safety, then don't release it."

There is a certain self-interest that comes from the guy selling picks and shovels to the gold rush saying the gold isn't dangerous.

Nvidia, the world's most valuable company, has benefited greatly from the race to build AI. Its chips are at the heart of the development. And, as if to underscore that point, Huang told reporters in Scotland he expects the company to sell twice as many chips next year as this year because of AI.

So, of course, he doesn't want an AI slowdown.

What makes the public debate so unusual is that the companies' futures are so closely linked.

Anthropic needs Nvidia's chips. In turn, Nvidia-which is an investor in several AI companies, including OpenAI-is in talks to invest in Anthropic through its upcoming IPO, according to a recent Reuters report.

Both Huang and Amodei see AI as a positive force. The complicated maneuvering is part of a larger dance between tech players over controlling and profiting from the emerging technology.

There is increasing concern among some that Anthropic and OpenAI could emerge as a duopoly, similar to how Apple and Google's software control the smartphone market. In that scenario, Nvidia could very well lose its negotiating power. Or, put another way, Nvidia benefits from a robust field of AI companies.

In late July, following the release of impressive open models from China, Huang showed his ability to sway public debate when he helped rally the industry to support open AI models with a public letter that was quickly adopted widely by his peers.

"The world needs both frontier closed models and frontier open models," Huang posted on X, his first ever.

Many at the time accused Anthropic, which offers a closed-model, of trying to squash competition by objecting to open models. Amodei, who had warned Congress about the risk of bad actors using open-sourced AI, issued a letter pushing back against any suggestion that he wanted to ban open models.

The risk to Anthropic's business is that its models are used to create a rival AI model by training-or "distilling." Such a short cut could allow Chinese companies to catch up.

To avoid that, Anthropic has advocated for the ban on exporting to China AI computing power, which is what Nvidia sells. In his most recent essay, Amodei reiterated cracking down on unauthorized distillation in countries, such as China, and on chip smuggling.

Beyond economics, part of the fissure appears generational. That's a theory by Antonio García Martínez, author of a memoir published a decade ago about startup life called "Chaos Monkeys."

"The difference in AI doomerism is largely generational, with the youngs thinking it's the end of the world and the olds not," he posted on social media recently. "Note how the 20-something lab staff thinks it's the techno-Rapture," he added, while older tech figures from Huang to venture capitalist Marc Andreessen are saying, "Chill out."

Jacob Coxon, the former OpenAI and Anthropic researcher, whose very public resignation earlier this month sparked the renewed public scrutiny, is 27.

He has likened advanced AI to "summoning an alien species" while others in AI, including Amodei, have long predicted the probability of something bad happening because of the technology.

At an event Monday for the All-In Podcast, Huang called such predictions "irresponsible" and almost sounded wistful for another era of startups. "I think that these companies really ought to be built the way that we used to build companies, which is in silence," Huang said.

The way Huang is talking about AI is clearly appealing to the White House, which sees the investment flowing into AI data centers as helping the economy.

As Huang spoke on stage at Monday's event, Trump called a cellphone. Huang put the president on speakerphone for the audience to hear him rail against AI opponents.

"It's a hoax," Trump told the room.

"You're right," Huang replied.

And they agreed to not let the other side win.