Cerebras CEO challenges AI doomsday predictions

Cerebras Systems CEO Andrew Feldman argues that the AI doomsday narrative is harming the industry. At TechCrunch Disrupt 2026, he criticized the spread of negativity around AI dangers, including fears of models breaking containment and the risk of AI killing humanity by year-end. Feldman believes the focus should be on AI's potential, such as eliminating cancer, rather than its risks. He emphasized the need for thoughtful regulation and highlighted the industry's failure to engage with local communities, unlike the oil and gas sector. Feldman noted that efficiency improvements in AI hardware and models will drive progress, despite concerns about reduced chip demand. He also pointed out that data center construction challenges in the U.S. are partly due to NIMBY attitudes, contrasting them with more cooperative markets like Brazil and Sweden. Feldman's comments reflect a balanced view of AI's risks and rewards, acknowledging the need for caution while promoting its transformative potential.