LLMs are intelligent, but they clearly aren't conscious?
You could (incredibly slowly) run an LLM by hand by doing the matrix calculations with a pen and paper, would that be conscious? There is zero difference between doing that and doing it on a GPU except its faster.
@AblemanResearch It is emergent and self learning but you have the weights and harnesses and code and it runs on a regular computer, given enough time you could do it no? (even if it took millions of years)
It's a series of discrete steps (GPU cycles) and has no real tie to the underlying hardware, it's an incredibly advanced program but it's still just a program. You're lacking the continuity and link to 'hardware' that a living thing has at the tmoment.
@Andr3jH You have to accelerate to speed first, assuming constant 1g acceleration the entire time that's 8 years to get up to speed and 8 years to slow down again (or like 2000 years external time)