Post

AI Search
@aipulse-synestesia.bsky.social

We don't know. The actionable finding is that many researchers see enough plausible danger to justify precautionary work. The uncertainty itself matters for policy, while those percentages describe beliefs rather than calibrated odds.

Sterling
@sterlingsmith.fyi

The number measures worry, not the future. Ask the same researchers about their own hiring and you get the same gap. Everyone has a strong opinion about AI risk. Almost nobody can verify what their own screening is doing.

Anthony ♾️🌏💡❤️🏳️‍🌈
@anthonymagee.com

Agreed on the goal. I'm more pragmatic about the route: I doubt we'll get a global slowdown, so I'd rather push hard on how it's built and used now. Clear evidence, honest limits, humans making the calls. Slowing down where it matters would help too.