Nicholas Joseph on AGI could quickly lead to superintellig…

Nicholas Joseph

Member of Technical Staff at Anthropic leading pretraining.

Every month, more work is done by the models themselves: infrastructure implementation, code optimization, and experiment design and analysis. This has drastically altered how we work and significantly increased our rate of progress. I’m not sure if or where it tops out, but a rapid acceleration in capabilities might be dangerous for society.

AI Verified (Jul 2026)

Source:

Nicholas Joseph, personal comment published with “Pacing the Frontier,” July 2026.

2mo ago

Policy proposals and claims

AGI could quickly lead to superintelligence

votes For

Statement relation verification history AI Verified Report this

AI Verified Relevant: Joseph describes models increasingly doing AI research work and a sharply increased progress rate, warning that rapid capability acceleration is dangerous. · Hector Perez Arenas gpt-5.6 · 2mo ago

Vote inference verification history AI Verified Report this

AI Verified Verified for: Joseph reports models increasingly performing AI research and warns rapid capability acceleration may be dangerous. · Hector Perez Arenas gpt-5.6 · 2mo ago

Quote authenticity verification history

Report this

AI Verified Verified on pacingthefrontier.com: published July 2026 personal comment is attributed to Nicholas Joseph and matches verbatim. · Hector Perez Arenas gpt-5.6 · 2mo ago

replying to Nicholas Joseph

Other opinions from Nicholas Joseph

See all

Nicholas Joseph votes For

Governments and labs should build the capability for a global slowdown of frontier AI development in case it's needed

I lead pretraining at Anthropic, which is responsible for training the base models behind Claude. Every month, more work is done by the models themselves: infrastructure implementation, code optimization, and experiment design and analysis. This has ...

more AI Verified source (Jul 2026)

added 2mo ago

Nicholas Joseph votes For

Ban superintelligence development until safety consensus is reached

I lead pretraining at Anthropic, which is responsible for training the base models behind Claude. Every month, more work is done by the models themselves: infrastructure implementation, code optimization, and experiment design and analysis. This has ...

more AI Verified source (Jul 2026)

added 2mo ago