Ben Oldroyd is breeding anarchist bees. There's a particular genetic mutation in honeybees that switches the drones from policing and destroying one another's eggs, to a situation in which egg-laying becomes more common and is not punished*. Usually the queen is promiscuous: you can calculate the genetic relationship coefficients and from there understand why the hive works as it does (why the workers will sacrifice themselves for the whole, why worker eggs are eaten). But once this change occurs (perhaps the drones can't tell the difference between the anarchist eggs and the queen's eggs), the initial conditions have changed and the solutions to the game theory equations change... the dynamic of the hive changes. It becomes an anarchist collective. It makes me wonder how stable our own ethics are [to small perturbations]. Are they the best available iterative solution to the game theory equations that describe society, physics, our behaviour? If so, shouldn't we see the same ethics arising in any system with the same basic rules, just like we see the Fibonacci sequence in the skin of a pineapple and rabbit populations over time?
There's more about the honeybees in Anarchist Bees in The Economist, where I seem to remember it saying Oldroyd's team had pinned the mutation down to a single gene, called alien. (If anyone has a subscription and can send me the full text, I'd be very grateful. [Update. Someone did. Thanks!])
(* Originally I'd said the mutation makes the drones fertile. This isn't true, honeybee drones are fertile anyway, and the cause is somewhat different. Thanks Mark Ward for the correction, and the email on honeybees and anarchy from his experience.)