Yeah, well in the 1960s everyone was convinced that the brain functions using principles that are very similar to McCullough-Pitts networks....Everyone I talk to is 100% convinced that the brain functions using principles that are very similar to deep learning algorithms...
I forget which scientist built a physical analog computer in the 1950s, that used four meter movements that could move a metal vane through a arc-shaped trough of conductive salt water, connected through a rotary switch system of random resistors. And a few tube amplifiers. Each had contacts at the end of its travel, so when it hit the limit, it would invoke a motor to turn the switch one position and randomly change the resistor network. The thing would sit there slowly moving vanes through troughs and switching the network until it arrived at an equilibrium in which every vane settled down into a static position within each trough. IIRC that typically took many hours.
He maintained that it was exhibiting intelligent behavior, solving problems, and demonstrating conditioned reflexes.
Everyone always thinks they are on the verge of understanding how the brain works.
I believe that the brain is a physical system that can in principle be explained by physical laws and is in principle capable of being emulated by electronic circuits. I just think we're nowhere near that point, and, indeed, it is at least possible that the human brain does not have the capacity to understand the human brain.
Statistics: Posted by nisiprius — Thu Jan 01, 2026 8:17 am — Replies 227 — Views 20020