Thank you.
Thank you.
Reproductive dynamics as a game.
Loci <-> Players
Alleles <-> Strategies
Genotype <-> Strategy profile.
Common interests (all loci get same payoff/fitness).
Sexual reproduction converges to monomorphic pure NE genotype.
Mutations add alleles/strategies. Transition to new NE with higher fitness.
...there is a limit to how many children you can sacrifice without significant objective average fitness effects for those within your culture. Anyway, plenty to think about.
Can imagine arms race in crazy beliefs lead to selection of super crazy beliefs which lead to selection of mega crazy beliefs and so on. At some point should bump into objective reality. E.g. If an intermediate goal is to please a bloodthirsty god, then child sacrifice is justified, but...
Tools (mind etc.) that I use for subjective selection, my evaluative technology so to speak, have been selected for. Of course, they are imperfect & I can end up believing in sorcery because it "seems" to be effective, but such effect should be bounded by objective reality.
Cultural evolution driven by subjective selection. Enjoyed the paper, but not wholly convinced by the subjective/objective dichotomy used. Yes, selection can occur towards intermediate goals, but it still seems like a kind of messy objective selection.
onlinelibrary.wiley.com/doi/10.1002/...
PS. Contrast zero-sum games, where player *outcomes* are opposed, but *incentives* might be aligned. A game like matching pennies is both harmonic & zero sum (opposed outcomes, opposed incentives), but you can also construct a zero-sum potential game (opposed outcomes, aligned incentives).
Harmonic games are the opposite of potential games. In potential games, player incentives are aligned, in harmonic games, player incentives are opposed. Turns out that no-regret learning is PoincarΓ© recurrent in harmonic games. i.e. dynamic returns to close to its starting point.
Hello world.