Don't Worry About the Vase Podcast The Week in AI Governance
Aug 1, 2025
The podcast dives into the latest developments in AI governance, discussing how major companies like Google and OpenAI are aligning with the EU AI Code of Practice. It raises alarms about the risks of AI deregulation tools on public safety. Listeners will find insights into China's AI strategy and its international collaboration efforts amid criticisms. The urgent need for verification in AI capabilities, especially for national security, is highlighted. The discussion also critiques tech export strategies and addresses ongoing chip supply constraints impacting AI development.
AI Snips
Chapters
Transcript
Episode notes
Major AI Firms Back EU Safety Code
- Most major AI companies, except Meta, have signed the EU AI Code of Practice's safety chapter, marking a step up in safety commitments.
- Meta remains the sole major company refusing any commitment, showcasing a rogue stance on AI safety.
AI for Federal Deregulation
- The US government is exploring AI tools to rapidly identify and cut unnecessary federal regulations, promising trillions in savings.
- This approach risks oversimplification and errors due to AI's imperfect understanding and lack of human nuance.
China's Open AI Cooperation Vision
- China's AI Action Plan emphasizes open models, international cooperation, and broad AI diffusion with a focus on the Global South.
- It mostly articulates general aspirations lacking strong specifics on serious safety measures.