For Humanity: An AI Risk Podcast We’re Racing Toward AI We Can’t Control | For Humanity #79
13 snips
Feb 14, 2026 David Krueger, AI professor turned safety advocate who founded Evitable, discusses the race toward uncontrollable superintelligence. He covers why research alone won’t save us, the geopolitics of chip supply chains, job displacement as a political lever, and the need to build public pressure for governance and pauses in development.
AI Snips
Chapters
Transcript
Episode notes
Mobilize Public Pressure
- Inform and organize the public to create political pressure for safety measures.
- Relying only on whispering to elites or policymakers won't produce the necessary, enforceable will to act.
Real-World Harms Are Converting Attention
- Harms from AI have already shifted from theoretical to real, including job displacement and fraud scandals.
- These emerging, tangible harms create openings to connect public concern to existential risk conversations.
Target Chip Supply Chains
- Target choke points in the AI supply chain, especially advanced chips, to create verifiable pauses.
- A hardware-focused approach is easier to monitor than trying to detect secret software development.