3 ms·

And as a scientist working in quantum physics I can confirm that the physics part of the article was fishy as well. It uses handwavy metaphors and flawed compar

And as a scientist working in quantum physics I can confirm that the physics part of the article was fishy as well. It uses handwavy metaphors and flawed comparisons. It would work great for a scifi story, but I would not take it too seriously as popsci explanation of real science.

jerf 5y ago
The philosophy aspect is questionable as well. That the fundamental layer of the universe may be something amenable to quantum simulation is way less impressive than this author thinks; it is not clear to me what "fundamental physics" would not be so amenable to quantum computation. If such a fundamental theory of everything wasn't amenable to quantum computation I'm not sure we could ever discover it. If you hike over to the latest work that materials science has been doing you can also see that once you have a "qubit" it's pretty natural to make it sing and dance and do all sorts of related mathematical tricks. It's actually unsurprising there are things that might theoretically be able to simulate qubits in a more complicated universe. Materials science is past the point where it's proving these things exist and is into the process of engineering these states to spec, even if they haven't quite come up with a generalized quantum computer yet. Or, to put it another way, I'm not updating my Bayesian probability of "the universe is a simulation" simply because it seems like the universe seems to be something that could be simulated on a particular type of computer, on the grounds that I don't have and never had any probability assigned to "the universe is fundamentally not something that could be simulated", again on the grounds that if that statement is true somehow it is likely that for the same reason it is true we could never prove it, or even necessarily gather significant evidence for it. (How would I do that? By finding some event in the universe that fails to match my own computational models? The probabilities that this is evidence for the non-computability of the universe would be smaller than "my model is wrong", "my data fed into the model is wrong", and "the physical process of computing my model failed" almost no matter what I did.) I already had effectively 100% assigned to "any observation I can make with my senses, no matter what machines may be augmenting it, is something that could be simulated with a sufficiently advanced simulator"... in fact, a "sufficiently advanced" simulator could be faking most of the "quantum" nature of reality, even with my machines I only can ever observe a very few qubits directly at a time. (And I personally am observing basically none! I'm just reading articles about it. It is easy for the Great Simulator to be forging those articles for me, in a reality that is "generally" non-quantum except when it absolutely has to be for me.)
- whatshisface 5y ago
  That's a common misconception, the great simulation doesn't forge articles for you, it forges the memory of having read them. Since people forget almost everything they read, we found that it was much easier to do it that way. Disclaimer: I work at Great Simulation Metaincorporated.
- tshaddox 5y ago
  Couldn’t a quantum computer simulate a Universe that appears (to any simulated observers within) to be impossible to be simulated on a quantum computer? And doesn’t that mean that, even if we observed something in our Universe that can’t be simulated by a quantum computer, that wouldn’t provide any “evidence” or any “reason to update our priors” regarding whether our Universe is being simulated?
  - jerf 5y ago
    "Couldn’t a quantum computer simulate a Universe that appears (to any simulated observers within) to be impossible to be simulated on a quantum computer?" Given the Church-Turing thesis, what would that look like, though? How would we even identify it? We ask "hey, universe, what's 2 + 3" and it says "C'thulu"? Even the quantumness of the simulation is an efficiency argument, not a possibility argument. A classical computer can simulate our quantum universe, it just has even worse resource requirements that simulating the universe seems to. But there would be no experiment you could run from within that would show you're running on a classical computer or a quantum one. You might be able to create experiments that consume yet vaster amounts of computation time, but you couldn't tell from the inside. And it's not as if I can just stroll down to my kitchen and perform an experiment that will go one way if the universe is computing to a 10^-20 accuracy, and another if it's using 10^-23 accuracy. Most of your personal experience could be simulated with a classical approach with ad-hoc minor quantum fixups here and there.

jerf 5y ago

The philosophy aspect is questionable as well. That the fundamental layer of the universe may be something amenable to quantum simulation is way less impressive than this author thinks; it is not clear to me what "fundamental physics" would not be so amenable to quantum computation. If such a fundamental theory of everything wasn't amenable to quantum computation I'm not sure we could ever discover it. If you hike over to the latest work that materials science has been doing you can also see that once you have a "qubit" it's pretty natural to make it sing and dance and do all sorts of related mathematical tricks. It's actually unsurprising there are things that might theoretically be able to simulate qubits in a more complicated universe. Materials science is past the point where it's proving these things exist and is into the process of engineering these states to spec, even if they haven't quite come up with a generalized quantum computer yet. Or, to put it another way, I'm not updating my Bayesian probability of "the universe is a simulation" simply because it seems like the universe seems to be something that could be simulated on a particular type of computer, on the grounds that I don't have and never had any probability assigned to "the universe is fundamentally not something that could be simulated", again on the grounds that if that statement is true somehow it is likely that for the same reason it is true we could never prove it, or even necessarily gather significant evidence for it. (How would I do that? By finding some event in the universe that fails to match my own computational models? The probabilities that this is evidence for the non-computability of the universe would be smaller than "my model is wrong", "my data fed into the model is wrong", and "the physical process of computing my model failed" almost no matter what I did.) I already had effectively 100% assigned to "any observation I can make with my senses, no matter what machines may be augmenting it, is something that could be simulated with a sufficiently advanced simulator"... in fact, a "sufficiently advanced" simulator could be faking most of the "quantum" nature of reality, even with my machines I only can ever observe a very few qubits directly at a time. (And I personally am observing basically none! I'm just reading articles about it. It is easy for the Great Simulator to be forging those articles for me, in a reality that is "generally" non-quantum except when it absolutely has to be for me.)

whatshisface 5y ago
That's a common misconception, the great simulation doesn't forge articles for you, it forges the memory of having read them. Since people forget almost everything they read, we found that it was much easier to do it that way. Disclaimer: I work at Great Simulation Metaincorporated.
tshaddox 5y ago
Couldn’t a quantum computer simulate a Universe that appears (to any simulated observers within) to be impossible to be simulated on a quantum computer? And doesn’t that mean that, even if we observed something in our Universe that can’t be simulated by a quantum computer, that wouldn’t provide any “evidence” or any “reason to update our priors” regarding whether our Universe is being simulated?
- jerf 5y ago
  "Couldn’t a quantum computer simulate a Universe that appears (to any simulated observers within) to be impossible to be simulated on a quantum computer?" Given the Church-Turing thesis, what would that look like, though? How would we even identify it? We ask "hey, universe, what's 2 + 3" and it says "C'thulu"? Even the quantumness of the simulation is an efficiency argument, not a possibility argument. A classical computer can simulate our quantum universe, it just has even worse resource requirements that simulating the universe seems to. But there would be no experiment you could run from within that would show you're running on a classical computer or a quantum one. You might be able to create experiments that consume yet vaster amounts of computation time, but you couldn't tell from the inside. And it's not as if I can just stroll down to my kitchen and perform an experiment that will go one way if the universe is computing to a 10^-20 accuracy, and another if it's using 10^-23 accuracy. Most of your personal experience could be simulated with a classical approach with ad-hoc minor quantum fixups here and there.