it seems worth pointing out that one of the key papers that gave birth to AI, "A Logical Calculus of the Ideas Immanent in Nervous Activity" (from 1943), in which McCulloch and Pitts model the brain and all its functions as a digital system, is full of...
-
it seems worth pointing out that one of the key papers that gave birth to AI, "A Logical Calculus of the Ideas Immanent in Nervous Activity" (from 1943), in which McCulloch and Pitts model the brain and all its functions as a digital system, is full of errors and was soon found to be a very poor model of how the brain actually works. doesn't that give you pause? #AI
-
it seems worth pointing out that one of the key papers that gave birth to AI, "A Logical Calculus of the Ideas Immanent in Nervous Activity" (from 1943), in which McCulloch and Pitts model the brain and all its functions as a digital system, is full of errors and was soon found to be a very poor model of how the brain actually works. doesn't that give you pause? #AI
Interesting. I know little about this paper or the human brain for that matter,
but i DO know that the human (and other real) brains work nothing like digital/"AI" systems -
possibly the reason why they outperform so-called "AI" by orders of magnitude, and why the brute force approach of modern "AI" is driving into a wall. -
Interesting. I know little about this paper or the human brain for that matter,
but i DO know that the human (and other real) brains work nothing like digital/"AI" systems -
possibly the reason why they outperform so-called "AI" by orders of magnitude, and why the brute force approach of modern "AI" is driving into a wall.@sebastian yes, that's exactly what I'm thinking/saying - it's as if the whole endeavour of AI came out of a way of thinking in the 1940s that everyone saw as flawed even then but somehow that didn't stop the hype or stop the field from growing. so strange...
-
@sebastian yes, that's exactly what I'm thinking/saying - it's as if the whole endeavour of AI came out of a way of thinking in the 1940s that everyone saw as flawed even then but somehow that didn't stop the hype or stop the field from growing. so strange...
@loriemerson
the human brain does about
10**15 operations/Joule- what s the best that computers can do today? 10**10 ops/J ?
and they are getting less efficient with every attempt to drive into the wall harder ...(leaving open the question of whether ops are a good measure of intelligence .. which kinda answers itself ...)
