3 ms·

Perhaps it is falsifiable. Perhaps there is some object with a fractal property that would be enormously expensive to compute, but that nature can determine wit

by rm445 14y ago

Perhaps it is falsifiable. Perhaps there is some object with a fractal property that would be enormously expensive to compute, but that nature can determine without computation. Perhaps such an object could be produced, and it could be shown that its property would be of such near-infinite complexity to rule out us running in a simulation. Or perhaps I've just spouted some technobabble. The point is that we are at such an early stage of thinking about such things that they may yet be put on a scientific footing. And if we do find answers to age-old philosophical conundra (the universe-simulation one not being much different to older questions about what we can determine about reality) then any answers that come will come from science, not philosophy.

dexen 14y ago
> and it could be shown that its property would be of such near-infinite complexity to rule out us running in a simulation. I believe a fractal is not a viable experiment. You hinge the experiment on assumption any simulation would be bound by memory size or time taken to compute. Practical constraints aside (matter and space is quantized at very small scale, so the depth of computation is finite), I don't believe we can just assume such limitations. I believe memory shouldn't be a problem for a fractal, because properties of the fragment you are trying to observe can be computed directly from equations describing the fractal, never mind the scale. As for time taken to compute, it is not observable. We don't know of any reliable timing source outside of the hypervisor[1] to compare time taken for computation to; the supposed VM is free to take as long as it needs to to advance to compute state, before advancing one http://en.wikipedia.org/wiki/Planck_time http://en.wikipedia.org/wiki/Planck_time * * * It could be argued the http://en.wikipedia.org/wiki/N-body_problem http://en.wikipedia.org/wiki/N-body_problem of all the celestial bodies already puts our supposed VM to a good test ;-) --- [1] if we knew one, it would make the experiment moot, as we would have a direct proof we're inside of a simulation ;-)

by rm445 14y ago

dexen 14y ago

> and it could be shown that its property would be of such near-infinite complexity to rule out us running in a simulation. I believe a fractal is not a viable experiment. You hinge the experiment on assumption any simulation would be bound by memory size or time taken to compute. Practical constraints aside (matter and space is quantized at very small scale, so the depth of computation is finite), I don't believe we can just assume such limitations. I believe memory shouldn't be a problem for a fractal, because properties of the fragment you are trying to observe can be computed directly from equations describing the fractal, never mind the scale. As for time taken to compute, it is not observable. We don't know of any reliable timing source outside of the hypervisor[1] to compare time taken for computation to; the supposed VM is free to take as long as it needs to to advance to compute state, before advancing one http://en.wikipedia.org/wiki/Planck_time http://en.wikipedia.org/wiki/Planck_time * * * It could be argued the http://en.wikipedia.org/wiki/N-body_problem http://en.wikipedia.org/wiki/N-body_problem of all the celestial bodies already puts our supposed VM to a good test ;-) --- [1] if we knew one, it would make the experiment moot, as we would have a direct proof we're inside of a simulation ;-)