Pinker argues doomer extinction narratives starve real AI risks like cybersecurity

Dr_Atoosa · x · 2026-09-06

Weighing in on criticism around Cal Newport's piece, the author offers a charitable reading of Pinker's post: the classic doomer AI-extinction narrative — a singleton or family of superintelligent AIs wiping out humanity — has monopolized scarce attention that could have gone to concrete, nearer-term risks like cybersecurity and biosecurity. The author doesn't fully endorse Newport but says the EA and Yudkowskian rationalist communities did a disservice (alongside service) to AI governance and its discourse, something careful people should acknowledge.

Original post →

More from AGI Musings

AGI Musings channel →