[sanxiyn]

In reinforcement learning, you are supposed to have large learning rate at the beginning and smaller and smaller learning rate as you go, eventually reaching zero. It would make sense for brain to use similar strategy.

In unusual cases temporarily tweaking learning rate can be profitable, and it could apply to brain too.

[wyager]

>large learning rate at the beginning and smaller and smaller learning rate as you go, eventually reaching zero. It would make sense for brain to use similar strategy.

Maybe if we were still cavemen living in a technologically stagnant world.

The world is changing extremely fast. Humanity needs neuroplasticity boosters to keep up.