Cerebras CEO Counters AI Doom Narrative, Highlights Infrastructure Opportunities

Cerebras Systems co‑founder and CEO Andrew Feldman used his stage time at TechCrunch Disrupt 2026 to push back against the prevailing doomsday narrative surrounding artificial intelligence. He suggested that some prominent AI voices “maybe shouldn’t be allowed to talk in public” because their focus on existential risk drowns out the tangible benefits the technology can deliver.

Feldman emphasized that AI leaders should spend more energy illustrating positive outcomes — such as a future where cancer deaths become a memory — while still acknowledging legitimate safety concerns. He framed the debate as a balance between risk and reward, noting that every powerful technology inevitably attracts misuse, and that reasonable regulation is the appropriate response.

The executive also linked the negative sentiment to the data‑center industry’s recent community relations. He admitted that U.S. operators have not engaged local stakeholders as effectively as the oil‑and‑gas sector, fueling NIMBY opposition. By contrast, countries including Brazil, Canada, Finland, Iceland, Norway, Mexico, and Sweden are actively courting data‑center builds.

Why it matters for GPU / AI infrastructure

Feldman argued that efficiency gains in chips and models will not shrink demand for high‑performance compute; instead, they will enable larger, more ambitious workloads. For GPU cloud providers and AI hardware vendors, this underscores a sustained need for scalable, energy‑efficient infrastructure that can support both training and inference at scale.