Superintelligence is coming. Should we let it?
TL;DR
AI companies have been talking about superintelligent AI like it’s inevitable, but recent safety incidents like OpenAI’s Hugging Face breach are demonstrating the potential dangers of deploying AI systems that are more capable than humans. So what happens when we can’t reliably control what these systems do? On this episode of TechCrunch’s Equity podcast, Rebecca Bellan is joined by Connor Leahy, an AI researcher, entrepreneur, and now the U. S.
Nauti's Take
A prominent control skeptic like Connor Leahy getting extended airtime on an industry podcast is a real opportunity for the debate, since control questions usually disappear behind product announcements. The format has clear limits: a podcast delivers claims and no evidence, and the OpenAI Hugging Face incident it cites is an operations failure rather than proof of models slipping control.
Teams should still listen, mainly for the question of which access their own AI systems actually need.
Sources