Tuesday, October 25, 2022

Dreams and Neuroscience

Razib Khan interviewed Eric Hoel, a neuroscientist at Tufts on a variety of topics, but I thought this would be fun to pass along. Hoel won the book review contest at ACX this year, which would be one of my most-preferred honors to achieve. Likely because I know I never will.

Razib also talks to Hoel about his recent paper, The overfitted brain: dreams evolve to assist generalization, which argues that by “hallucinating out-of-distribution sensory stimulation every night, the brain is able to rescue the generalizability of its perceptual and cognitive abilities and increase task performance.” In plainer English, dreams allow the brain to experiment with novel possibilities outside of the range of experience and let it be more flexible and well-prepared in the face of surprising stimuli.

This is not entirely new, as I have seen things like this before.  But it does seem to be more precise and be stronger evidence. I am planning some discussion on this, but sometimes I overpromise.

More direct link.

2 comments:

Grim said...

“My theory is we might be imagining things.”

I’m pretty sure that is what “dream” always meant to ordinary people.

Ganzir said...

"In plainer English, dreams allow the brain to experiment with novel possibilities outside of the range of experience and let it be more flexible and well-prepared in the face of surprising stimuli."

Kind of like a chess engine analyzing future positions