Inner Cosmos with David Eagleman Ep121 "What’s the secret to intelligence (in brains and AI)?" with Ramesh Raskar
27 snips
Sep 15, 2025 
In this engaging discussion, Ramesh Raskar, a professor at MIT Media Lab and an innovator in decentralized AI, shares fascinating insights about the evolution of intelligence in both humans and artificial systems. He explores the parallels between complex neural networks of the brain and decentralized AI, highlighting the transformative potential of small AI agents in healthcare. Raskar emphasizes the importance of ethical frameworks as these technologies develop and reflects on how cooperative and competitive dynamics shape both human cognition and AI advancements.
AI Snips
Chapters
Books
Transcript
Episode notes
Brains As Rival Teams
- The brain is a team of rival specialized modules that compete and cooperate to produce behavior.
- Decentralized AI could mirror this by using many small agents instead of one giant model.
Share Intelligence, Not Raw Data
- Decentralized AI can extract insights from local data without moving raw records.
- Hospitals could let micro-agents learn on-site and return only learned model updates.
Favor Small Specialized Agents
- Build small, task-specific agents rather than one massive model for every job.
- Let agents improve by interacting with local data and other agents over time.
