Cerebras CEO Pushes Back On AI Doomsday Warnings logo

Cerebras CEO Pushes Back On AI Doomsday Warnings

Cerebras CEO Andrew Feldman told TechCrunch Disrupt that AI leaders' public doomsday warnings are overshadowing the technology's potential benefits, arguing some executives 'shouldn't be allowed to talk in public.'

ShareXLinkedInEmail

THE RUNDOWN

1

Feldman's comments directly respond to warnings from AI safety researchers, including a widely discussed Anthropic researcher resignation over self-improving AI risk.

2

Cerebras counts OpenAI, Figma, Block, GSK and Lovable as customers, giving Feldman's framing real commercial weight with model-layer and enterprise buyers.

3

Feldman argues data-center community opposition stems from poor local engagement, not technology risk -- a framing GPs backing compute infrastructure should track.

4

The remarks land the same week Pulse covered Anthropic's warning that small models may soon self-replicate, a direct tension worth watching.

The VC Read

Value Add VC analysis

Feldman's comments are as much investor relations as AI-safety commentary -- Cerebras just went public and has every incentive to frame AI risk as manageable rather than existential. The diligence point for infrastructure investors: track whether his 'reasonable regulation' framing survives contact with an actual rulemaking process, since chip vendors tend to support regulation in the abstract and fight it in the specifics.

Analysis

Cerebras Systems CEO Andrew Feldman told TechCrunch Disrupt 2026 that the AI industry's doomsday warnings are drowning out its potential benefits, arguing some AI leaders "maybe shouldn't be allowed to talk in public" given how often their warnings dominate coverage over the technology's upside, TechCrunch reported.

Feldman's comments respond to a summer and fall of increasingly public AI-safety alarm, including a September 2026 case of an Anthropic researcher who quit while warning that self-improving AI systems could soon escape meaningful human control -- a concern Pulse covered separately in its report on Anthropic's self-replication warning. The debate has become mainstream enough that a recent "Saturday Night Live" sketch parodied Anthropic CEO Dario Amodei with a fictional line: "We do not condone what we are doing."

The case Feldman is making

Feldman argues every powerful technology carries dual-use risk and reward together -- "I think you can't make one and not have the other," he said -- and that the right response is "reasonable" regulation rather than public alarm. He pointed to AI's potential to help cure diseases, saying he wants a future where children "never meet anybody who dies of cancer," and argued the data-center industry has hurt its own image by failing to engage local communities the way the oil and gas industry historically has.

Cerebras, which competes with Nvidia in AI training and inference chips, counts OpenAI, Figma, Block, pharmaceutical company GSK and Lovable among its customers -- a list that gives Feldman's framing real commercial weight rather than pure punditry. Cerebras went public earlier in 2026, and Pulse has since covered its second-quarter earnings, where the company's guidance swings became a live test of investor appetite for AI-infrastructure volatility.

Feldman's remarks don't resolve the underlying safety debate so much as stake out a competing public-relations position -- one that chip and infrastructure vendors with the most to lose from a slowdown have strong incentive to take. Worth noting: Feldman did not name the researchers or labs he believes should talk less, leaving open whether he means direct competitors or allies making him uncomfortable.

ShareXLinkedInEmail

Key Sources

2 sources

Reported by TechCrunch · Analysis by Value Add Pulse.

← Back to Pulse

THE WIRE in your inbox— Tech, startup & VC news with The VC Read, a few times a week. Free to subscribe, no spam.

No spam, unsubscribe anytime. Sent via Substack · Privacy · or subscribe on Substack ↗