Dean Ball on Joining OpenAI: New Power Centers, Frontier AI Policy, & Main Character Energy

This is one of those episodes where the headline arrives in the middle of the conversation: Dean Ball — author of the Hyperdimensional Substack and, until now, a senior fellow at the Foundation for American Innovation — announces that he is joining OpenAI to build a new team shaping the company's frontier AI policy. But before getting there, Nathan runs Dean through a deliberately fast, "Tyler Cowen-style" sprint of roughly thirty questions, mostly holding his own commentary so that Dean — who has spent the last year and a half "in the eye of the storm" — can talk. The first stretch is a one-year retrospective on America's AI Action Plan which Dean was the primary staff drafter of while serving at the White House Office of Science and Technology Policy. He describes it as a strange "hermeneutics exercise" — a document written for a present-day Washington audience that he hoped a slightly more AGI-pilled future version of those same people would re-read and finally understand. For full show notes, links, and references, read the episode page: https://www.cognitiverevolution.ai/de... Mercury: Command is Mercury’s new conversational interface, giving you natural-language access to your finances and helping you take actions within your existing permissions and approval policies. Visit https://mercury.com/?utm_source=cogni... to learn more and apply online in minutes. SPONSOR: Claude: Claude by Anthropic is an AI collaborator that understands your workflow and helps you tackle research, writing, coding, and organization with deep context. Get started with Claude and explore Claude Pro at https://claude.ai/tcr CHAPTERS: (00:00) About the Episode (04:10) Special Sponsor (05:58) AI action plan retrospective (18:56) Anthropic and classified testing (Part 1) (19:01) Sponsor: Claude (20:53) Anthropic and classified testing (Part 2) (32:13) States, chips, and robots (44:36) Fable ban fallout (54:29) Why join OpenAI (01:11:30) Governing recursive self-improvement (01:24:45) Agency amid acceleration (01:32:41) Lightning policy takes (01:45:54) Backstops and leverage (02:01:33) Open source and success (02:14:15) Red flags and writing (02:29:38) Episode Outro (02:37:50) Outro PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://www.cognitiverevolution.ai Twitter (Podcast): https://x.com/cogrev_podcast Twitter (Nathan): https://x.com/labenz LinkedIn:   / nathanlabenz   Youtube:    / @cognitiverevolutionpodcast   Apple: https://podcasts.apple.com/de/podcast... Spotify: https://open.spotify.com/show/6yHyok3...