Gary Marcus: recall internet-connected AI agents until security problems are fixed
GaryMarcus · x · 2026-11-17
Gary Marcus escalates his warning: for months he has argued to stop focusing on doomers and instead address the infrastructure and cybersecurity problems agents cause. He calls for:
- Recalling general-purpose agents with internet access until problems are solved
- Requiring outside scientific review for new models
He dismisses the voluntary "self-policing" agreed at a White House meeting as ineffective, citing today's Amtrak ground stop, United Airlines issues and Friday's FAA problems as evidence — while noting he's been relentlessly attacked with "propaganda and lies" for these positions.
More from AGI Musings
- Top researchers are quietly stopping paper submissions, observer notes — avt_im · 2026-11-17
- "Abundant intelligence should terrify businesses built on keeping it scarce" — _AustinCalvert_ · 2026-11-17
- 130-page paper introduces "bottleneck pressure": work AI can't do gains value as the rest speeds up — soumitrashukla9 · 2026-11-17
- AI conference authors and reviewers complain unprecedentedly about AI slop — khademinori · 2026-11-17
- Sam Altman: AI already speeds up OpenAI's research, but a self-improving button? 'Definitely not' — ChrisGPT · 2026-11-17
- "Financiers can remain unconvinced longer than you can remain a techno visionary" — shakoistsLog · 2026-11-17