Recursive's $670M Bet on Self-Improving AI, Sonnet 5.5 Hits 70%, Elon Co-Leads Pentagon Push | EP #299
October 3, 2026
AI Summary
5 min readOn a live video call, 48% of people who spoke with an AI avatar thought they were talking to a real human. The previous state of the art was 3%. This is the kind of number that makes you sit up and pay attention, and it was just one of several data points in a wide-ranging conversation about where artificial intelligence actually stands right now—not in science fiction, but in the models shipping this week, the labs racing to build self-improving systems, and the policy fights that are already breaking out over both.
The Core Thesis: Science as an Engineering Problem
Richard Socher, the co-founder and CEO of Recursive, has a central argument in his new book The Eureka Machine: AI will deliver a century’s worth of scientific breakthroughs in the next decade. The reason, he explains, is that scientific progress has slowed not because we are out of ideas, but because we have fragmented knowledge into too many narrow subdisciplines. A cell biologist knows cells; a molecular biologist knows molecules; no human can weave it all together. AI can.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Moonshots with Peter Diamandis
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 Main Outline
- 2 (00:00) **Introduction & Richard Socher's Thesis** - Richard Socher joins to discuss his book, Recursive's mission, and the state of AI.
- 3 (07:44) **Timelines for Scientific Breakthroughs** - The panel presses Richard on when AI will "solve" physical sciences and human disease.
- 4 (13:37) **The Fields Lagging Behind** - A discussion on which academic fields are slowest to adopt AI, specifically biology and economics.
- 5 (17:18) **The Bitter Lesson & Virtual Cells** - Richard explains the "bitter lesson" of AI and its application to biology.
- 6 (23:33) **Brain-Computer Interfaces & Brain IT** - The team discusses a new AI system that reconstructs images from fMRI brain scans.
- 7 (30:51) **The Mystery of Biological Learning** - A deep dive into how the brain learns without gradient descent, the core algorithm of AI.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
The mates sit down with Richard Socher to discuss Recursive’s $670M bet on self-improving AI, why he puts P(Doom) at zero, the race toward ASI, proposed restrictions on recursive self-improvement, and what new AI benchmarks like Tavus’ Turing Test could tell us about what comes next.
Get access to metatrends 10+ years before anyone else - https://qr.diamandis.com/metatrends
Peter H. Diamandis, MD, is the Founder of XPRIZE, Singularity University, ZeroG, and A360
Salim Ismail is the founder of Open ExO, a GP at Exponential Venture Capital/The Organizational Singularity Fund and a sought after global speaker and thought leader.
Dave Blundin is the founder & GP of Link Ventures
Dr. Alexander Wissner-Gross is a computer scientist and founder of Reified
Richard Socher is a leading AI researcher and entrepreneur, founder and CEO of You.com and co-founder and CEO of Recursive, who previously served as Chief Scientist at Salesforce after founding MetaMind. you.com
Read Richard Socher’s new book, The Eureka Machine: Why AI Is the Key to Unlocking a New Era of Scientific Discoveries: https://www.eurekamachine.org/
–
This episode is brought to you by:
Get the blueprint for generative media https://goo.gle/startupgenmedia
Go to Blitzy to book a free demo and start building today: https://qr.diamandis.com/blitzy
Your body is incredibly good at hiding disease. Schedule a call with Fountain Life to add healthy decades to your life, and to learn more about their Memberships: https://www.fountainlife.com/peter
_
Connect with Peter:
Connect with Dave:
Connect with Salim: