Jan 17
1/17/22
“The goal of this paper is to argue that the brain faces … [the] challenge of overfitting and that nightly dreams evolved to combat the brain’s overfitting during its daily learning. That is, dreams are a biological mechanism for increasing generalizability via the creation of corrupted sensory inputs from stochastic activity across the hierarchy of neural structures.”
This is great. It reminds me of ideas I’ve had (and written about) regarding the capacity of psychedelic experience to disrupt overdetermined (i.e. overfit) mental processes, using machine learning as a reference point. The author, Erik Hoel, also researches the neurological function of literature.
TL;DR: we need dreams/stories/aesthetic encounters to break away from the predictive mechanisms that stifle our apprehension of the world and imagination in general. Certainly they’re an antidote for algorithmic homogenization.
