Robin: It’s a fair point, but as a general matter, I’m not sure it’s biologically easy to have mutations that encode knowledge. Why be scared of snakes, instead of having the abstract knowledge “snakes are more dangerous than I would otherwise estimate them to be”? That’s how you would build an AI, but evolution goes down the path of things that happen easily as the result of mutations.
I would reply that the brain just doesn’t seem to be built around that kind of architecture. Even when an estimate does get raised, it gets raised as the result of feeling some emotion that binds to the estimate. If evolution is going to build a feeling anyway, it may be easier to feel that you just shouldn’t do something; than to feel that you should override your mental model that says you won’t get caught, with an expectation of getting caught anyway for reasons not in your mental model. Yes, the latter would be more elegant, but evolution is nothing if not inelegant.
Maelin: Read. The. Damned. Links. Here, I link it again. Even if whole bands went extinct more often than individuals within bands, that still wouldn’t be enough, mathematically speaking, to let group selection win out.
Robin: It’s a fair point, but as a general matter, I’m not sure it’s biologically easy to have mutations that encode knowledge. Why be scared of snakes, instead of having the abstract knowledge “snakes are more dangerous than I would otherwise estimate them to be”? That’s how you would build an AI, but evolution goes down the path of things that happen easily as the result of mutations.
I would reply that the brain just doesn’t seem to be built around that kind of architecture. Even when an estimate does get raised, it gets raised as the result of feeling some emotion that binds to the estimate. If evolution is going to build a feeling anyway, it may be easier to feel that you just shouldn’t do something; than to feel that you should override your mental model that says you won’t get caught, with an expectation of getting caught anyway for reasons not in your mental model. Yes, the latter would be more elegant, but evolution is nothing if not inelegant.
Maelin: Read. The. Damned. Links. Here, I link it again. Even if whole bands went extinct more often than individuals within bands, that still wouldn’t be enough, mathematically speaking, to let group selection win out.