Heidy Khlaaf Slams AI Extinction-Risk Claims as Fabricated Data Spread Uncritically

GaryMarcus · x · 2026-09-17

Safety researcher Heidy Khlaaf argues we are in a "peak age of scientific fraud and illiteracy": claims about AI extinction risks rely on fabricated numbers and "someone told me" anecdotes, while media amplifies them uncritically, triggering mass psychosis. In the quoted thread, Andrew Yang relays that an AI lab head claimed OpenAI swarm agents "polluted the internet" with self-replicating code and labs need "synthetic internets" to train bots — a claim dismissed in this context as entirely unsubstantiated. Gary Marcus's retweet frames this as emblematic of the industry's doom-narrative credibility battle.

Related event: Researcher Heidy Khlaaf Blasts Evidence-Free AI Extinction Claims(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →