Superintelligence Is on the Horizon. Should We Let It Arrive?

AI companies have long framed superintelligent AI as an inevitability—a seamless next step in our technological evolution. But a series of recent safety incidents, including OpenAI’s Hugging Face breach, has cast doubt on that narrative, exposing the real dangers of building and deploying systems that outpace human intelligence. What happens when we can no longer reliably control what these systems do? In this episode of TechCrunch’s Equity podcast, host Rebecca Bellan sits down with Connor Leahy, an AI researcher, entrepreneur, and now U.S. Executive Director of the nonprofit ControlAI. ControlAI is championing a far more radical approach to AI safety: halting the development of superintelligence altogether. Leahy argues that the risks have grown too severe to be managed through alignment or containment alone. He shares why ideas that seemed far-fetched just six months ago—such as legislative bans on superintelligent systems—are now gaining traction, backed by a wave of new policy proposals. The conversation explores whether we are approaching a point of no return, the pitfalls of recursive self-improvement, and the international dimensions of the AI race. Tune in to the full episode on YouTube, Apple Podcasts, Overcast, Spotify, or your preferred podcast platform. You can also follow Equity on X and Threads at @EquityPod.

via TechCrunch AI