Evolving Prosocial AI: A conversation with Peter Fenton and David Sloan Wilson

Unknown Source October 22, 2025 64 min

artificial-intelligence startup generative-ai investment nvidia google openai anthropic

🎧 Listen to Original

63 Companies

84 Key Quotes

4 Topics

1 Insights

🎯 Summary

Evolving Prosocial AI: A Conversation with Peter Fenton and David Sloan Wilson

This 64-minute podcast episode features venture capitalist Peter Fenton (Benchmark) and evolutionary biologist/philosopher David Sloan Wilson discussing the intersection of technological evolution (specifically AI) and evolutionary theory, focusing on how to engineer prosocial outcomes in rapidly evolving systems.

---

1. Focus Area: The primary focus is the evolution of Artificial Intelligence (AI), framed through the lens of evolutionary biology and multi-level selection theory. The discussion bridges Peter Fenton’s experience in Silicon Valley venture capital (tracking tech cycles like the internet, social/mobile, crypto, and now generative AI) with David Sloan Wilson’s expertise on how selection pressures shape cooperation and potential dysfunction (like cancer) in complex systems.

2. Key Technical Insights:

• Historical Context of AI Models: Fenton noted the early 1990s tension between symbolic AI (favored on the East Coast) and neural networks (favored in California), highlighting that biological models proved effective even when they lacked simple, top-down heuristic functions.
• Rapid Ecosystem Responsiveness: The speed at which Silicon Valley adapted to the launch of ChatGPT (2022) is unprecedented, spawning hundreds of LLM-based startups almost instantly, demonstrating the system’s extreme adaptability to disruption.
• Alignment as Multi-Agent Competition: The conversation touched upon emerging alignment methodologies, specifically mentioning Emmett Shear’s alignment competition where multiple competing agents/teams collaborate to achieve a complex, shared outcome, suggesting alignment is not a single-machine problem but an ecosystem challenge.

3. Market/Investment Angle:

• Silicon Valley as the Epicenter: Fenton asserts that Silicon Valley remains the most likely location (90% probability) for the next trillion-dollar market cap company when a new technology segment emerges, due to its unique cultural and structural advantages.
• Venture Capital Cycles: The discussion mapped recent major tech cycles: Internet Boom (peak 1999), Social/Mobile (2007-2012), Crypto (2013-2022), and the current, largest wave: Generative AI (starting 2022).
• Mania as Necessary Fuel: Fenton argues that the manic cycles and bubbles inherent to Silicon Valley are paradoxically required to drive innovation at the necessary pace, even though they lead to inevitable depressive cycles.

4. Notable Companies/People:

• David Sloan Wilson: Intellectual hero to Fenton, known for his work on evolutionary biology, altruism, and multi-level selection.
• Peter Fenton (Benchmark): VC with a strong background in philosophy, involved in early investments in Uber, Snap, Instagram, and Twitter.
• Elliot Sober: Philosopher of biology and mentor to Fenton, whose work on selection units and altruism frames the discussion on cooperation.
• Emmett Shear (Softmax): Mentioned for pioneering alignment competitions that use multi-agent competition to drive prosocial outcomes.
• Jonathan Haidt: Referenced regarding his work on the negative societal impacts of social media algorithms on youth well-being (The Anxious Generation).

5. Regulatory/Policy Discussion:

• The Three Models of Governance: The conversation contrasted three approaches to managing technological evolution: Laissez-faire (Silicon Valley mantra, leading to entropy/cancerous self-interest), Top-Down Regulation (often leading to regulatory capture, as seen in healthcare), and a Third Way informed by evolutionary biology.
• Perverse Incentives: Fenton suggested that current industry calls for minimal regulation are often driven by incumbent AI leaders (OpenAI, Anthropic) seeking to block disruptive startups through regulatory capture.
• The Need for Prosocial Alignment: Wilson emphasized that evolution does not automatically yield beneficial outcomes; just as cancer is evolution at a lower scale destroying the larger organism, unaligned AI optimization risks systemic destruction.

6. Future Implications:

• Co-evolution at Unprecedented Scale: The rapid penetration of AI (ChatGPT reaching over a billion users quickly) means humanity is co-evolving with this technology at a pace never before witnessed, raising critical questions about the health of that evolution.
• The Third Way for AI Alignment: The future likely requires moving beyond the binary of pure deregulation or heavy bureaucracy. The solution lies in engineering systems that foster cooperation and alignment across multiple levels, mirroring successful biological structures (like multicellularity).

7. Target Audience: Tech Professionals, Venture Capitalists, AI Researchers, and Philosophers/Ethicists interested in the structural and cultural drivers of technological innovation and the application of evolutionary theory to AI alignment and governance.

🏢 Companies Mentioned

Twitter ✅ General Tech/Social Media

Reddit ✅ General Tech/Social Media

Softmax ✅ General Tech (AI)

Meta ✅ Tech/Social Media (Known Web3 Player)

Facebook ✅ Tech/Social Media (Known Web3 Player)

Reddit Japan ✅ unknown

And John Schulman ✅ unknown

Nick Fiedler ✅ unknown

Howard T ✅ unknown

Emmett Shear ✅ unknown

Lynn Margulis ✅ unknown

Washington D ✅ unknown

Silicon Valley Silicon Valley ✅ unknown

Bill Gurley ✅ unknown

The Anxious Generation ✅ unknown

💬 Key Insights

"One reason it works is that the objective function, the purpose, the raison d'être, is not profit. It's a repository of knowledge, or it's a software development system that everyone can contribute to."

Impact Score: 10

"how do you allow for an objective function to be aligned but in a system that's inscrutable? And in the absence of observability inside of the AI, it's not like you could see, 'Pull this lever, pull that lever.'"

Impact Score: 10

"If the objective function is profit maximization, we're going to have outcomes that are at real odds with the high functioning of this technology alongside humanity."

Impact Score: 10

"if you launch a lot of recent grads with advanced degrees into the economy without work, especially if they're lawyers—lead overproduction—and well for those familiar with Peter Turchin, an elite overproduction, which is sort of now amplified with the elimination, the in is full, the seats are taken, and half the seats are demonetized by AI."

Impact Score: 10

"is to whether this is going to be a cancerous form of evolution or it's going to be one that leads to worldwide pro-sociality. And of course, currently, we're very much into cancerous mode."

Impact Score: 10

"When a new species comes online, 90% of the phenotypic change occurs in the first 10% of the species' existence."

Impact Score: 10

📊 Topics

#artificialintelligence 112 #startup 32 #generativeai 18 #investment 1

🧠 Key Takeaways

💡 sublimating all the other things that are in your life