Gary Marcus calls for boycotting Generative AI, citing Anthropic and OpenAI scientists' own admissions

GaryMarcus · x · 2026-09-10

Gary Marcus published an essay arguing "the time has come" to boycott generative AI. His core claims:

He cites two pieces of evidence: Anthropic researcher Evan Hubinger publicly stated he sincerely believes AI could kill all humans, putting >10% probability within the decade, and that alignment for superintelligence remains unsolved and Anthropic is "not clearly on track" (28M views); days earlier, OpenAI chief scientist Jakub Pachocki was similarly candid about the lack of a solution to the alignment problem.

Marcus concedes some fear-mongering may be marketing, but insists real risks exist with no plan behind them — while noting he personally doesn't worry much about extinction, given human geographic and genetic resilience.

Related event: Gary Marcus Calls for Boycott of Generative AI, Warning of Frontier Lab Risks(4 posts)→

Original post →

More from AGI Musings

AGI Musings channel →