Fermi estimate: brain may pack 500k-5M molecular switching units per 'parameter'
JosephJacks_ · x · 2026-09-28
Joseph Jacks argues that comparing AI parameters to synapses misses enormous underlying machinery. At 4-16-bit precision each weight needs 4-16 DRAM transistors plus arithmetic and interconnects. On the biology side, a tubulin dimer carries eight tryptophan residues whose indole chromophores show fluorescence consistent with collective optical behavior. A Fermi estimate suggests 500 quintillion tubulin-associated tryptophan sites in the brain — roughly 500,000 to 5 million candidate molecular switching units per 'parameter' if synapses (100-1000 trillion) stand in as parameters. He proposes measuring and deriving scaling laws for this machinery rather than assuming today's neural-network scaling laws bound biology.
More from AGI Musings
- AI Ethicist Daniel Faggella Lays Out His Cause: Life, Potentia and AGI's Moral Status — danfaggella · 2026-09-28
- 'I'm an AI optimist': founder lays out first-principles case for AI-driven abundance — gajesh · 2026-09-28
- VC pleads: stop sending AI-generated decks, 'they smell like total lack of thought' — saranormous · 2026-09-28
- Red Queen Bio Uses AI to Design Antibodies Before Future AI Creates Pathogens — jachiam0 · 2026-09-28
- Local Models and Bitcoin Are Both Freedom Tools, Argues Gladstein — csuwildcat · 2026-09-28
- Daniel Faggella: AGI alignment fixates on entitlement, humans should join the intelligence ecosystem — danfaggella · 2026-09-28