“We only get one chance to build superintelligence right.”
The worldview
You take seriously the possibility that advanced AI could go badly wrong. Not in a science-fiction way, but in the ordinary way powerful technologies fail: faster than anyone can react, and in ways nobody intended.
You think the people building frontier systems don't yet know how to reliably control them, and that competitive pressure pushes them to ship anyway. You support independent testing, government oversight and the ability to pause if needed.
You're not anti-AI. You just believe the downside is too large to learn by trial and error.
Locked
Their average profile
See where AI Safetyists actually land on all 13 dimensions, compared with everyone.
No password. We only receive your name, email and profile photo from Google. See our privacy policy.