I, scientist with Balazs Kegl Michael Levin part 2
Sep 25, 2025
In this engaging discussion, Michael Levin, a developmental biologist at Tufts University, delves into the fascinating interplay between Platonic patterns and biological systems. He challenges listeners to consider if mathematical structures might guide biology, as seen in cicada cycles. Levin posits that patterns themselves may possess agency, suggesting a new way to view consciousness in AI. He emphasizes the need for understanding and communicating with diverse minds, urging us to rethink the boundaries of organism and machine.
AI Snips
Chapters
Transcript
Episode notes
Patterns Underlie Biological Form
- Biological forms reflect deeper mathematical and computational patterns beyond genes and physics.
- Michael Levin calls this a structured latent (Platonic) space that constrains which forms appear when we build interfaces.
Cicadas Point To Math, Not Just Biology
- Levin uses cicadas as an example: 13 and 17 year cycles reflect prime-number structure.
- That explanation moves the question from biology into mathematics to explain timing choices.
Systematically Study The Latent Space
- Treat the latent space as a research program and systematically map its structure.
- Levin urges labs to study which patterns appear for specific interfaces to gain predictive control.