Gary Marcus says the singularity is still not here, and CEOs are overclaiming
Gary Marcus · rss · 2026-07-29
Gary Marcus argues that the industry still has not reached the singularity, and that CEO claims to the contrary are mostly hype.
- He says “CEO said a thing” journalism should be taken skeptically, because leaders like Elon Musk and Sam Altman have repeatedly made grand claims without a precise definition of what “singularity” means.
- He points to the Klarna Effect: companies that brag about AI replacing headcount are now quietly rehiring humans or expanding teams, suggesting current AI is still limited.
- He rejects several common singularity definitions as unmet: AI is not yet beyond human control, humans have not merged with AI through brain-computer interfaces, and we still can meaningfully stop or constrain these systems.
- Marcus cites I. J. Good’s 1965 “ultraintelligent machine” idea and says current systems are nowhere near “as good or better than the best human” on the full set of tasks AGI would need.
- He concludes that superintelligence may eventually arrive, but probably not this decade, and that better AI safety will be needed before it does.
More from AGI Musings
- Zuckerberg says Meta is preparing a positive vision for superintelligence — altryne · 2026-07-29
- Marcus Hutter argues AGI could make a job-less future desirable and likely — mhutter42 · 2026-07-29
- A repost mocks Pause AI for copying protest tactics in ways that look absurd — RishiBommasani · 2026-07-29
- Science study finds sycophantic AI boosts user certainty and reduces willingness to repair conflict — stanfordnlp · 2026-07-29
- Hundreds of Thousands of Tons of Books Landfilled Annually: AI Training as a Better Alternative — aronchick · 2026-07-29
- Spielberg’s 2001 film *A.I.* now reads like a warning about human-machine love — adariostrange · 2026-07-29