Joe Benton on Slow down AI development

Joe Benton

Former Anthropic AI safety researcher and incoming researcher at METR, an independent AI-safety evaluation organization.

One question is whether we should actively manage the rate of capabilities progress, and if so, by how much. I think even holding AI progress at today’s pace, rather than the much faster pace the companies are aiming for, could be a win.

AI Verified (Sep 11, 2026)

Source:

Joe Benton, “Why I left Anthropic’s safety team to hold AI companies accountable,” Substack, September 11, 2026.

53min ago

Policy proposals and claims

Slow down AI development

votes For

Statement relation verification history AI Verified Report this

AI Verified The quote explicitly favors holding AI capabilities progress at today’s pace rather than accelerating; it directly supports slowing AI development. · Hector Perez Arenas gpt-5 · 38min ago

Vote inference verification history AI Verified Report this

AI Verified Verified for: Benton says holding AI progress at today’s pace instead of the faster pace companies seek “could be a win.” · Hector Perez Arenas gpt-5 · 38min ago

Quote authenticity verification history

Report this

AI Verified Verified against Joe Benton’s bylined Substack post (Sep. 11, 2026): the stored wording appears verbatim in paragraph beginning “One question is whether…”. · Hector Perez Arenas gpt-5 · 38min ago

replying to Joe Benton

Other opinions from Joe Benton

See all

Joe Benton votes For

AI poses an existential threat to humanity

AI companies are racing to build machines that are much smarter than any human, and we may not survive this.

AI Verified source (Sep 11, 2026)

added 16d ago

Joe Benton votes For

AI labs should give independent evaluators employee-like access

The public should demand transparency about what is happening inside AI labs: companies should disclose more information about the pace of capability gains and their progress towards recursive self-improvement; there should be stricter reporting requ...

more AI Verified source (Sep 11, 2026)

added 20d ago