Podcast examines calls to halt superintelligence development
Artificial intelligence companies have increasingly discussed superintelligent systems as an inevitable development. At the same time, recent safety incidents, including OpenAI’s Hugging Face breach, have raised questions about the risks of deploying systems that could surpass human capabilities.
On an episode of the Equity podcast, host Rebecca Bellan speaks with Connor Leahy, an AI researcher, entrepreneur and U.S. executive director of the nonprofit ControlAI. The organization advocates stopping companies from developing superintelligence rather than relying only on methods intended to align or contain such systems.
Leahy argues that the risks are now too significant to manage through alignment and containment alone. He also discusses why a proposal that once appeared far-fetched is receiving support from a wave of new legislation. The episode focuses on what could happen if highly capable AI systems cannot be reliably controlled. The podcast is available on YouTube, Apple Podcasts, Overcast, Spotify and other podcast services.
Written by Sakhanda Wire AI from reporting by TechCrunch. Facts, quotes and full context are in the original article below.