Cerebras CEO Andrew Feldman: AI Leaders "Maybe Shouldn't Be Allowed to Talk"

By Olivia Hart

Cerebras Systems CEO Andrew Feldman told the TechCrunch Disrupt 2026 audience that "some of the leaders in the AI world maybe shouldn't be allowed to talk in public." His target: the industry's own doomsday chorus. Feldman, who co-founded and runs the AI chipmaker competing with Nvidia, argued the existential-risk narrative now dominating headlines is crowding out the technology's tangible upside.

The doomer discourse he is pushing back against has genuine reach. AI insiders now debate the percentage chance that AI kills all of humanity by the end of the decade. Anthropic CEO Dario Amodei was parodied on Saturday Night Live, where a fictional character summed up the industry's posture with the line: "We do not condone what we are doing."

What does Feldman want AI leaders to say instead?

Feldman thinks executives with concerns about AI risk should spend more time on the potential payoff. His example: a world where your grandkids will "never meet anybody who dies of cancer." That is a promise, not a forecast — and Feldman did not offer a timeline or a mechanism for getting there.

The incentive is not exactly hidden. Cerebras counts OpenAI, Block, Figma, GSK and Lovable among its customers. Every dollar of AI optimism translates, eventually, into chip demand.

Is he dismissing the risks?

Not quite. Feldman framed risk as inseparable from capability.

"I don't think there are powerful technologies that don't come with the challenge of 'what happens when powerful technologies are in the hands of bad people?' I think you can't make one and not have the other," Feldman said at Disrupt.

"I think what we have to say is that this technology is powerful," he added. "We have to be thoughtful about it. We have to regulate it reasonably."

Note what he did not say: which regulations, enforced by whom, or what Cerebras itself does to police misuse of its hardware. "Reasonably" is doing a lot of work in that sentence.

Why does he think the public is souring on AI?

Feldman laid part of the blame on his own industry's handling of data center buildouts driven by AI demand.

"As an industry, we weren't super good for the past few years at working with local communities, particularly in the U.S. We — the industry — didn't do what others have done in the power industry in oil and gas. They didn't work with the community. They didn't engage the community as they should have, and that was a problem," Feldman said.

The comparison is instructive, if odd: oil and gas is not widely held up as a model of community relations. Feldman appears to mean that energy companies at least built a playbook for local engagement, one the data center sector skipped.

He contrasted American NIMBY-ism — "not in my back yard!" — with countries actively courting data center construction. He named Brazil, Canada, Finland, Iceland, Norway, Mexico and Sweden as markets encouraging such investment. That is seven national governments rolling out the welcome mat while U.S. communities push back.

Could efficiency gains kill chip demand?

Feldman dismissed the idea that better efficiency reduces the need for his silicon.

"We can do our job as hardware vendors to make our compute more efficient, and then the model makers can do improved work in making their models more efficient to achieve the same goals. All of this has to come together to make meaningful strides forward," he said.

The claim rests on an assumption he left unstated: that demand for AI capability keeps expanding fast enough to absorb every efficiency gain. History in computing mostly supports him — better chips have repeatedly widened rather than shrunk the market.

The takeaway: Feldman is selling optimism, but his call for reasonable regulation and better community engagement signals that at least one chip executive reads the public mood as a business problem, not just a public-relations one.