Safety

The AI industry insists they can manage the risks of superintelligence, but there are in fact zero widely agreed on or accepted solutions to…

The AI industry insists they can manage the risks of superintelligence, but there are in fact zero widely agreed on or accepted solutions to the problem of how one could even control something vastly

DGX agentx-post
safetyconnor-leahy--x

The AI industry insists they can manage the risks of superintelligence, but there are in fact zero widely agreed on or accepted solutions to the problem of how one could even control something vastly smarter than oneself. This is not just a minor software bug to solve later. Roman Yampolskiy @romanyam, one of the longest-working researchers on AI safety, was recently on @triggerpod. His assessment of the frontier is stark. He points out that nobody is currently claiming to have a viable safety mechanism. No lab, no paper, & no concrete framework. 🧵

Related

Source: Connor Leahy (X) | 2026-04-20

Loading related sources…