Superintelligent AI Risks and the Push for Regulation: A Conversation with Connor Leahy

TechCrunch's Equity podcast hosts Connor Leahy, US executive director of the nonprofit Control AI, for a discussion on the existential risks posed by superintelligent AI. The episode examines recent AI safety incidents, the limitations of technical alignment, and the growing push for legislation in the US and UK to prevent the development of superintelligent systems.

■ Background and technical perspective
- From GPT-2 reverse engineering to founding EleutherAI and Conjecture
- Why technical alignment alone is insufficient

■ Superintelligence risks and incidents
- The Hugging Face breach and autonomous AI agents
- Superintelligence as an adversary, not a tool

■ Policy, legislation, and action
- US and UK legislative efforts to ban superintelligence
- International cooperation and citizen advocacy

This episode is suited for listeners interested in AI safety, technology policy, and the societal implications of advanced AI. It provides an overview of the arguments for regulating superintelligent AI and the current state of policy efforts.