Hacker News new | ask | show | jobs
by return0 4871 days ago
ANNs were build with the assumption of point neurons as nodes in a network and many of their training algorithms are biophysically unrealistic. While ANNs can be good function approximators, they are a very inaccurate representation of brain neuronal neurons and networks (which have intricate dendrites and biophysics). We don't know yet how fundamental processes like memory storage work in the brain; it's one of the great mysteries of neuroscience.