| Can AI Learn to Improve Itself? | Recursive Self-Improvement | Sarah Wang | The a16z Show | 2026-11-17 | | For decades, the idea of AI improving itself was just a thought experiment. Today, AI is already helping researchers at frontier labs run experiments, write code and test ideas faster. Mark Chen says OpenAI has built “AI research interns.” So how far are we from AI that can do the research on its own? In our first episode of Working Theory, a16z general partner Sarah Wang explores recursive self-improvement with two leading AI researchers: Mark Chen, Chief Research Officer at OpenAI, and John Schulman, co-founder and Chief Scientist of Thinking Machines Lab, and co-founder of OpenAI. They cover what RSI actually means, why research taste is so hard to teach, and why AI may be one of our best tools for solving alignment. |
| Sam Altman: AI Beyond Human Control “Absolutely Possible” | Sam Altman | Fortune 500: Titans and Disruptors of Industry | 2026-09-12 | 46 min | OpenAI CEO Sam Altman sits down with Fortune's Editor-in-Chief Alyson Shontell to discuss the high-stakes balancing act of AI: pushing the boundaries of scientific discovery while confronting the existential risks of loss of control. In this episode, Altman addresses the reality of AI alignment, the Hugging Face incident, and OpenAI's commitment to halting training runs if safety thresholds aren't met. He also shares insights on global AI coordination between the U.S. and China, healthcare breakthroughs, and the company's timeline for an IPO. 00:00 Introducing Sam Altman 01:54 Astra, AGI claims, and P(doom) fears 03:10 Is a 10% extinction risk real? 06:03 Loss of control and OpenAI’s safety stance 09:05 Recursive self‑improvement and human oversight 12:19 Why alignment is still unsolved 14:23 Can humans control superintelligence? 16:35 From grade‑school math to Navier–Stokes 18:42 The Hugging Face sandbox escape incident 21:58 Limits of current safeguards and accident reporting 23:01 IPO timing, profit pressure, and pausing for safety 28:36 Global coordination, U.S.–China, and AI standards 35:09 P(doom) vs. the upside case for AI 41:51 The weight on Sam Altman and human motivations Learn more about your ad choices. Visit megaphone.fm/adchoices |