AI and Dualism

[james03]

To me, it appears that AI has settled the argument on dualism.  In short, dualism is correct.  AI can do everything that the human brain can do, because AI is modeled after the human brain, so called "Neural Networks".  And with hardware advancing, the capabilities are approaching a perfect match.  AI is a human brain.

And AI can do NOTHING that the human mind can do.  It can't perceive.  It can't desire.  It will never be curious.  It will never have intentionality, doing something "it" wants to do to achieve a goal "it" desires.  AI can never perceive the smell of a rose, or the qualia of "red", it can never be "self aware", for in the immaterial world, it doesn't exist nor operate.  It can not become "self aware" because it can never become "aware".  It will never hold the concept of "triangle" or any concept at all.  It, like the brain, reduces down to open and closed connections creating binary bit codes.  And that's it.

And so the brain and the mind are separate, and dualism is correct.

[ChairmanJoeAintMyPrez]

It's worse than you think, actually.

LLMs are good at tricking human operators into believing that an intelligence underlies their responses, but nothing that interesting is actually happening under the covers.  They are nothing more than statistical language models backed by enormous amounts of computing power.  They do not "understand".  For a given input, all they do is generate a set of tokens that would constitute a probable response to that input.

If you had an infinite amount of time, you could probably train Mr. Ed to respond to calculus questions.  He wouldn't actually know what he's doing, but he might manage to correlate your questions to the output that you expect.  An LLM is nothing more than this.

They also take exorbitant amounts of computer power to train and operate.  Until we create a computer with the same speed, power efficiency, structures, sensory input, and non-deterministic behavior as the human brain, it will not be possible to fully model the human brain.  And this is impossible, because these characteristics can be fulfilled only by a human brain.  Wires and silicon will never match biology.

[james03]

and non-deterministic behavior as the human brain, it will not be possible to fully model the human brain.  And this is impossible, because these characteristics can be fulfilled only by a human brain.  Wires and silicon will never match biology.

Non-determistic behavior and perception have nothing to do with "biology" or the brain.  They are faculties of the soul, which is immaterial.

[ChairmanJoeAintMyPrez]

and non-deterministic behavior as the human brain, it will not be possible to fully model the human brain.  And this is impossible, because these characteristics can be fulfilled only by a human brain.  Wires and silicon will never match biology.

Non-determistic behavior and perception have nothing to do with "biology" or the brain.  They are faculties of the soul, which is immaterial.

I thought it would be obvious from my comment that I consider replicating the human brain to be impossible.  The only way to create a human brain is the old-fashioned way, but it still requires decades of intense training.