1/14/2015

01-14-15 - Spooky Action at a Distance

I loved "Only Lovers Left Alive". There's not much to the plot, but as just a bit of mood it's lovely. The design, the music, everything is perfect. It reminds me of my youth, actually. An aethestic and lifestyle that's no longer fashionable.

Adam's song - Spooky Action At A Distance - is fantastic.

BUT - there is no fucking spooky action at a distance!

Ooo, you say, Einstein was confused about it. It must be all mysterious and nobody understands it. NO!

Quantum Mechanics is not some crazy mysterious metaphysical thing that science can't explain and therefore allows all sorts of quackery.

First of all, the average person shouldn't read thoughts about Quantum Mechanics that were written in the 20's and 30's when it was still new and not very well understood yet. Hey, obviously they were a little confused back then. Read something modern!

Second of all, I think the way QM is taught in schools is still largely terrible. I don't know how it's done now, but when I went through it it was still being taught in a "historical" way. That is, going through the progression in the same way it was discovered, roughly. First you learn quantization of photons, then you get a wave function, you get copenhagen interpretation and measurement collapse, and then you finally start to get bra-ket and hilbert spaces and only much later (grad school) do you get decoherence theory.

The right way to teach QM is to jump straight to the modern understanding without confusing people with all that other gunk :

1. The universe is a linear vector space of *amplitudes* (this is the source of all the weirdness and is the fundamental postulate) (the universe seems to work in the square roots of things, and double-covers, and this leads to a lot of the non-classical weirdness)

2. The entire universe is QM, there is no classical "observer" that lies outside the QM system.

3. The universe takes all possible outcomes all the time. However, an observer in any given state is not in all of those outcomes.

4. Measurement is the process of entangling the QM state of the measurement device with the QM state of the observed process.

In a very brief nutshell measurement decoherence (entanglement of the measurement device / outside world with a quantum system) goes like this :


You have a measurement device which is initially in an undetermined state |M>

You have something to measure.  Let's take a quantum state which is either 0 or 1 
(eg. a single particle that's either spin up or down) 

|0> + |1>

Initially they are not entangled :

|M> ( |0> + |1> )

And measurement changes them to be entangled :

( |M0>|0> + |M1>|1> )

M0 means the measurement device has observed the state 0

Note that there is no "collapse". Both possibilities still exist, but if the measurement device observes a 0 then the particle must be in state 0.

Now, in the book teaching QM we'd have to go into more detail about what "measurement decoherence" actually is. Because the entire universe is governed by QM, we'd have to show that that process happens by QM as well. In fact that has been done and the details worked out in the past 40 years, so it's relatively new.

Measurement decoherence doesn't happen at the single-particle level. Rather it's a large-scale phenomenon that happens due to entropy, much like thermodynamics. This is why "measurement collapse" could be seen as a connection to an outside classical system for large objects, because it only happens at large system scales. It only happens one way - you can entangle the measurement device with the quantum particles, but you can never de-entangle them (this is why in the old copenhagen interpretation you could get away with talking about "collapse", because if the measurement device is classical-scale then it can never go backwards). The basic laws of QM are time-reversible, but decoherence is not. It's exactly like if you took a bucket of red water and a bucket of blue water and mixed them in one larger bucket - it is now effectively impossible to ever separate out all the particles and restore the two different buckets.

So. The so-called "EPR Paradox" that is not a paradox at all. (well it is if you assume a deterministic physics with hidden variables, which is just wrong; it should be called the EPR Proof that Einstein was Wrong Sometimes). The EPR thought experiment goes like this :

You have two particles that are either in state 0 or 1, and are entangled so that they are either both 0 or both 1 :


|0>|0> + |1>|1>

You separate them to opposite ends of the universe, where lie Alice and Bob :

|A>|B>( |0>|0> + |1>|1> )

Bob then measures his particle and becomes entangled with it, in state B0 or B1 :

|A>( |B0>|0>|0> + |B1>|1>|1> )

If Alice now measures her particle, she will get either a 0 or 1, and see the same thing as Bob. However, nothing has moved faster than light. Her entanglement happens entirely on her local particle :

|A>|0> -> |A0>|0>

it has nothing to do with the state of Bob far away. The initial entangled state of the particle simply gaurantees they will get the same result. But the physical interaction is still local and slower than light, and the information about the result must travel slower than light.

There is never any "action at a distance" or any "paradox" or anything "spooky".

Sheesh. Come on vampires, you've been alive long enough to get your physics right.

2 comments:

Jason Schulz said...

Relativistic effects were probably spooky to everybody who already knew Newtonian physics.

johnb said...

Have you watched this Google Talk? If so, what do you think of it?
"The Quantum Conspiracy: What Popularizers of QM Don't Want You to Know"
https://www.youtube.com/watch?v=dEaecUuEqfc
(The title is tongue-in-cheek, of course)

Unfortunately I've forgotten almost all of the things I learned in the Quantum Computing module I took as an undergrad, despite it having possibly the best lecturer I've had (Richard Jozsa). It was very interesting while I felt like I understood it though.

old rants