AI pet peeve: everyone equates artificial neural networks and gradient-based optimization with brains, minds, and thinking.
ANNs are big, parameterized math equations that we configure with an algorithm. Living neurons are intelligent agents that manage their own behavior and relationships autonomously. Human brains definitely aren't attempting to differentiate through their interactions in the physical world, because that isn't possible. They don't do backprop, either.
Deep learning is its own thing. Brains are something else. It's hard to figure out how they compare when we keep pretending they're the same.
