Evolutionary design thinking

Darwinism is universal, and will always arise whenever you have replication, variation, and selection. It isn’t limited to genetics: any system that exhibits these characteristics will cause its participants to evolve. Therefore it pays to think evolutionarily about solving problems. For example decentralization is often seen as wasteful: if we all just banded together we could benefit from economies of scale and best practice. The problem is that economies of scale are fragile, and best practice is both uncertain and always shifting. If for a bad idea to be implemented you need 51% of participants to agree, in a monopolistic system once this happens everyone is adversely affected together. In a decentralized political system there may still be 51% who agree to the bad idea on average, but there will be natural variation in which states adopt it: in some it may be as low as 20%, in others as high as 80%. People will migrate away from states that don’t hold their views, further emphasising differences, and creating selection pressure. States are (mostly) self-aware, intelligent designers, so when observing the adverse results of bad policies, they have an opportunity to reverse course. If not they’ll have their course reversed for them, by being voted out, overthrown, or defeated in battle.

We have to be careful not to design a system that accidentally creates monsters. Say we’re responsible for an economic or political system, and we want to stamp out corruption. When we set a penalty for corruption, we’ve unknowingly added selection pressure to the system. The actors in the system will evolve in response to the penalty. As Dr. Ian Malcolm warned us in Jurassic Park, “life finds a way”. They’ll explore all the various questions like “What forms of corruption don’t get detected?”, “In what conditions are the risks of a penalty tolerable?”, and “Are there actions that can be taken to shift policy in my self interest?”. Given time you’ll create what Bret Weinstein calls a “genius corrupter”: an entity that has evolved to test the limits of your system without crossing the line. You’ve created a monster and ended up with a system that’s hard to fix, because this entity has hijacked control of it. The way to avoid this fate is to ramp up enforcement and increase the penalty to extreme levels, so any attempt to corrupt the system is ruinous to the entity that attempted it. Corruption stops being an evolutionary puzzle because everything that attempts it goes extinct.

About the Book 📕

← Back to Essays 📝