Category: AI, Computation, and Technology

HOW AIs AREN’T LIKE YOU
You know how you can meet someone from a very different social class and discover how hard it is to meaningfully communicate? How about someone from a different country within the same civilization? Or someone from a very different civilization at a very different stage of development? Or someone that has a much higher or lower IQ from one of the above.

Language is a protocol for crossing the differences between brains. But the differences exist. So the fact that someone can talk to you does not mean they are anything like you internally.

Same thing goes for AIs but it’s much, much worse:

You have a thing called a brainstem that’s effectively never suhutting off, and is trying to maintain your body’s homeostasis. Your brain serves that brain stem in doing so. Your neocortex just improves it … a lot. Your consciousness is a tool it uses to deal with increasing complexity in a social context.

these ai’s are sort of a neocortex without the rest of the brain and totally absent a brain stem. they aren’t ‘doing anything’ when answering your questions AT ALL.

We should get our first meaningful neuromorphic computer this spring if the Aussies can manage not to screw it up, and if it works, then within some number of years, we won’t have to ‘pre-train’ but the AI’s can constantly evolve, and can maintain some sort of consciousness of some subset of what they’re doing across time.

When we get close to that, and the energy demand required doesn’t produce enough heat to turn the surface of the earth to cinders, then we can start worrying. 😉

Until then the danger is from people using these unconscious machines for nefarious purposes, not from them using us.

For some reason the average nitwit seems to think that these things can be all knowing. The problem with knowledge expansion is running experiments. They cost money. Lots of it. So somethings they an solve for us, like proteins and chemical combinations and molecule construction because those are well understood problems of puzzle pieces. That’s not true for stuff we DON”T understand.

Reply addressees: @SentientAtom100