3 ms·
> Even for isolated systems where, in principle, we could have zero surprisal and access to all possible information... This makes no sense. How could you hav
by lisper 2y ago
> Even for isolated systems where, in principle, we could have zero surprisal and access to all possible information...
This makes no sense. How could you have access to "all possible information" in an isolated system? You obviously can't make any measurements, and if the system is prepared, then it's entangled with the system used to prepare it and again cannot be isolated. The whole notion of "an isolated system" is a theoretical fiction that doesn't actually exist in physical reality, but even in theory one cannot access all of the information in an isolated system because of the no-cloning theorem. So this really feels to me like the old joke about spherical chickens.
Furthermore, this seems like an already-solved problem. Constructing classical reality requires copying classical information, and the only way to make that happen is to discard quantum information [1]. That is the source of the Second Law and the arrow of time [2].
[1] https://arxiv.org/abs/quant-ph/9512022 https://arxiv.org/abs/quant-ph/9512022
[2] https://blog.rongarret.info/2014/10/parallel-universes-and-arrow-of-time.html https://blog.rongarret.info/2014/10/parallel-universes-and-a...
- 11101010001100 2y agoAs you know, all possible information for an isolated system is obtained via solutions to the Schrodinger equation. This is standard many-body physics.
- lisper 2y agoWell, yeah, but that seems like a vacuous observation to me. In order to find solutions to the SE you have to know the initial conditions. How are you going to obtain those for an isolated system? You haven't solved the problem, you have just pushed it backwards in time.
- 11101010001100 2y agoThermalization (excluding systems which exhibit many body localization and the like) will occur regardless of initial conditions.
- lisper 2y agoThat is manifestly untrue because the SE is time-reversible. That is the whole problem.
- 11101010001100 2y agoYes, and at the same time, the statistics of the system can still satisfy thermalization.