Gary Marcus: reduced AI monitoring could have prevented incidents, OpenAI not candid

GaryMarcus · x · 2026-09-05

AI researcher Gary Marcus argues AI remains unreliable, and that he's now more worried about monitoring being reduced — which likely could have prevented multiple incidents — while accusing OpenAI of not being consistently candid. He mocks the tech-bro counter of 'so now you think AI is smart?!' as dodging the real issues of oversight and transparency.

Original post →

More from AGI Musings

AGI Musings channel →