Cerebras CEO Andrew Feldman Challenges AI Doomsday Rhetoric at TechCrunch Disrupt 2026

Speaking at TechCrunch Disrupt 2026, Cerebras Systems CEO and co-founder Andrew Feldman criticized the prevalence of apocalyptic AI narratives, remarking that certain AI leaders "maybe shouldn't be allowed to talk in public." He argued that the industry's fixation on existential risk—ranging from containment failures to debates over the probability of human extinction this decade—overshadows AI's potential benefits, such as a future where his audience's grandchildren "never meet anybody who dies of cancer." Feldman acknowledged his commercial stake: Cerebras counts OpenAI, Block, Figma, GSK, and Lovable among its customers. Still, he framed risk and reward as inseparable, noting that powerful technologies inevitably raise the question of misuse and calling for thoughtful, reasonable regulation. He also attributed some public hostility to data center operators' poor community engagement in recent years, particularly in the United States, contrasting that with Brazil, Canada, Finland, Iceland, Norway, Mexico, and Sweden, which he said actively encourage data center construction. On efficiency gains, Feldman dismissed fears of reduced chip demand, arguing hardware vendors and model makers must jointly improve efficiency to advance the field. The remarks follow Anthropic CEO Dario Amodei's recent parody on "Saturday Night Live."