Evolution and Computation

on YouTube ↗01h 03m

" summary

Claim: under weak selection classical Fisher‑Wright population genetics is equivalent to a repeated game run by genes using the multiplicative weights update rule, so sex evolves to maximize mixability (alleles that pair well broadly) rather than lone genotype fitness. Tools: toy two‑locus fitness matrices, simulations, and a theorem showing any Boolean trait with a small 1+ε advantage becomes fixed with high probability. Result appears at ITCS and explains preservation of genetic variation and rapid adaptation.

AI-written summary · about

" chapters (24)

1. 0:00Intro
2. 0:47The Algorithm as a Lens
3. 1:16Evolution before Darwin
4. 4:56The Wallace-Darwin papers: Exponential Growth
5. 7:28Cryptography against Lamarck
6. 8:12Genetics
7. 8:47The crisis in Evolution 1900 - 1920
8. 9:45The "Modern Synthesis" 1920 - 1950
9. 13:44Disbelief, algorithmic version
10. 15:05Valiant's Evolvability
11. 17:39And in this Corner... Simulated Annealing
12. 18:17The Mystery of Sex Deepens
13. 19:28A Radical Thought
14. 19:52Mixability!
15. 21:31Explaining Mixability (cont)
16. 22:38Pointer Dogs
17. 24:11Waddington's Experiment (1952)
18. 25:27Genetic Assimilation
19. 25:42Is There a Genetic Explanation?
20. 27:20Arbitrary Boolean Functions
21. 32:13Changing the subject: The experts problem
22. 36:10Multiplicative weights update
23. 38:43Theorem: Under weak selection, evolution of a
24. 42:31Finally...