OpenAI Employee Claims 'AGI Achieved', Sparks Criticism

GaryMarcus · x · 2026-08-16

Vahid Kazemi, a technical staff member at OpenAI, claimed that the company has "already achieved AGI," stating that the o1 model makes this even clearer. However, he immediately qualified this: they have not achieved "better than any human at any task," but rather "better than most humans at most tasks."

Critics argue this relies on a convenient, unconventional definition of AGI focused on breadth of tasks rather than peak expertise. Gary Marcus mocked the claim. This follows the full release of OpenAI's o1 model.

Related event: Gary Marcus Mocks OpenAI Employee's AGI Claim(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →