quantum indeterminacy and the observer effect, universal fine-tuning, Planck length/time, speed of light, holographic principle, information paradox in black holes, etc.
this line of argument, that, like, physics looks like it’s been designed to be efficiently computable, is invalid afaik, lots of stuff actually isn’t discrete, and the time-complexity class of the laws of physics are afaik exponential on particle count.
Presumably, that which once identified as human would “wake up” and realize its true identity.
Why would it ever forget, why would this be sleep? That’s a human thing. If there’s some deep computational engineering reason that a mind will always be better at simulating if it enters a hallucinatory condition, then indeed that would be a good reason for it to use a distinct and dissimilar kind of computer to run the simulation, because a lucid awareness of why the simulation is being run, remembering precisely what question is being pursued, and the engagement of intelligent compression methods, is necessary to make the approximation of the physics as accurate*efficient as possible.
Though if it did this, I’d expect the control interface between the ASI and the dream substrate to be thicker than the interfaces humans have with their machines, to the extent that it doesn’t quite make sense to call them separate minds.
Regarding signs,
true, but
this line of argument, that, like, physics looks like it’s been designed to be efficiently computable, is invalid afaik, lots of stuff actually isn’t discrete, and the time-complexity class of the laws of physics are afaik exponential on particle count.
Why would it ever forget, why would this be sleep? That’s a human thing. If there’s some deep computational engineering reason that a mind will always be better at simulating if it enters a hallucinatory condition, then indeed that would be a good reason for it to use a distinct and dissimilar kind of computer to run the simulation, because a lucid awareness of why the simulation is being run, remembering precisely what question is being pursued, and the engagement of intelligent compression methods, is necessary to make the approximation of the physics as accurate*efficient as possible.
Though if it did this, I’d expect the control interface between the ASI and the dream substrate to be thicker than the interfaces humans have with their machines, to the extent that it doesn’t quite make sense to call them separate minds.