Very long time AI skeptic (since 1980) but I believe you're wrong. It turns out we don't need to figure out exactly how a biological brain works. It's just natu

Very long time AI skeptic (since 1980) but I believe you're wrong. It turns out we don't need to figure out exactly how a biological brain works. It's just nature's solution for implementing matrix operations (greatly simplified take, obviously). Having hit on approaches that produce similar results to biological systems, we now don't need to focus so much on emulating them. I do suspect however that there's something to the biological experience of "being the life support system" for the brain, that significantly affects the training process. It might be challenging to simulate that.