1591: Bell's Theorem
Title text: The no-communication theorem states that no communication about the no-communication theorem can clear up the misunderstanding quickly enough to allow faster-than-light signaling.
Ponytail begins reading Bell's theorem to Cueball, who is standing 5 meters away. Cueball responds with a misunderstanding of Bell's Theorem in 1 nanosecond. The speed of light in a vacuum is 299,792,458 meters per second. In one nanosecond, the light from Ponytail would only have traveled 0.3 meters, thus Cueball misunderstands Bell's Theorem faster than the light from Ponytail reading the theorem can reach him, which implies that some kind of faster-than-light communication occurred to set up the misunderstanding. This would violate locality. Of course no correct information was transferred faster than light (see also the title text explanation below)...
This comic was published on October 16, 2015, five days before an article about the first-ever Loophole-free Bell's Theorem test was published in Nature magazine (DOI:10.1038/nature15759) (see also Bell test experiments). However, the paper was submitted almost two months earlier on the 24th of August and could most likely be found on-line before this comic was released. It was accepted by Nature already on the 28th of September, but was first published online October 21, 2015. Randall may very well have been aware of the imminent release of this paper, although it is peculiar that he did not wait until the paper was released. (This could potentially be a meta-joke, with the joke about Bell's Theorem being released before the paper about the relevant experiment was published)
Bell's Theorem states "No physical theory of (finitely many) local hidden variables can ever reproduce all of the predictions of quantum mechanics." It says that a theoretical treatment that divides the universe up into separate ("local") systems like this will always discard something about those systems' intercorrelations.
"Global hidden variables" are another story: if there is classical information shared across systems (perhaps by superluminal communication) even up to superdeterminism where the universe is just reading off a script, any correlations can be explained away. But this is unsatisfying.
The preferred resolution of the paradox is not to insist (as early physicists did) that the universe's state is a collection of bits (classical information), but treat it as a collection of qubits (quantum information).
In quantum mechanics (QM), "measurement" is the process of allowing a small system to interact with its environment in a controlled way. The interaction allows information about the system's state to escape to the environment, producing an "observation." If the measurement apparatus is governed by classical mechanics (impossible in reality, but a very common simplification for the purposes of calculation), then the observation can be thought of as classical information, a bit (yes/no answer) in the simplest case. While the system may have been in any one of infinitely many states before the measurement (each a superposition of classical states), the fact that the measurement must leave it consistent with the classical result means that it can end up in only finitely many states afterwards. This is the "wave-function collapse" of early QM, popularized by Schrödinger's cat, but unrelated to the Heisenberg Uncertainty Principle, with which lay audiences often confuse it.
Modern quantum mechanics acknowledges that the environment is not classical, and that wave-function collapse happens by a (comparatively) gradual process called "decoherence," where information leaving the system is made up for by information coming from the environment that drives the system closer and closer to one of the finitely many states predicted by the simplified model above. If a "Schrödinger's cat" is in a half-and-half superposition of the states "dead" and "alive", when its liveness is measured, the ratios of "dead" and "alive" will shift rapidly towards (though not quite reach) 0 and 100% or 100 and 0%. For all but the shortest time scales, the cat's post-measurement state might as well be classical.
Entanglement is a situation where the future outcomes of two or more measurements that would be independent in a classical world are nonetheless correlated. For example, two widely separated electrons from one source could be in a state where, considered individually, each is in a superimposed spin-up/spin-down state, but if one is measured as spin-up, the other will necessarily be measured as spin-down. This is untroubling if the two electrons are modeled as a single system, but strange-seeming if we think of them as separate: how did the measurement of the first electron allow information from the environment around it affect the far-away second electron? It seems like the electrons are communicating, potentially at superluminal speeds, which would violate either relativity or causality. In actuality, there's a fairly simple proof (see below) that correlations from entanglement can't be used to communicate, and causality and relativity are safe. But that doesn't make the seemingly faster-than-light effects much less of a surprise.
One can try to address these concerns by considering 'local hidden variables', classical properties of a local system (like a single electron) that could have been observed but were not. For example, perhaps a classical part of the electrons' state lets them "agree" on a future classical state at the moment they are entangled, and then they just reveal that state in the future. But this becomes unwieldy: there are infinitely many possible future observations the electrons would have to agree on, and it seems difficult to do this without infinitely many local hidden variables.
The title text jokes about the No-Communication Theorem. The real theorem states that although determination of the state of one half of an entangled pair immediately determines that of the other half, however far away it may be, there's no way for the observer of the other half to see if he's the first to find out the state or whether it'd already been determined by the first observer. Thus, no information travels from one observer to the other.
Randall's version of the No-Communication Theorem states that no matter how you try to send information about this theorem (no communication about the No-Communication Theorem) then it cannot clear up the misunderstanding about Bell's Theorem quickly enough that any correct information (about Bell's theorem) has actually been transferred faster than light. So the conclusion is the same as the real No-Communication Theorem - faster-than-light signaling is not possible...
- [Ponytail, facing right, is holding a piece of paper with both hands. In a small frame breaking the top of the large frame is a caption:]
- t= 0 nanoseconds
- Ponytail: This is called Bell's Theorem. It was first–
- [A double-headed arrow points to Ponytail and then exits the frame crossing into the next frame where it points to Cueball. The arrow is split in two parts and in the center across the border of the two frames is a label:]
- 5 meters
- [Cueball, facing left. In a small frame breaking the top of the large frame is a caption:]
- t= 1 nanosecond
- Cueball: Wow, faster-than-light communication is possible!
- [Caption below the frames:]
- Bell's Second Theorem:
- Misunderstandings of Bell's Theorem
- happen so fast that they violate locality.
add a comment! ⋅ add a topic (use sparingly)! ⋅ refresh comments!