Thread by @seanmcarroll

Okay, entanglement and spooky action at a distance. Let me give you my version of the best way to think about it. One's mileage, as ever, is permitted to vary. (1/n)

The fundamental idea is that we shouldn't think of "two separate particles" as two separate particles. They are just one thing. But in a precise way that departs from our classical intuition. (2/n)

Say we have two spinning electrons. Four important things about quantum mechanics: (1) When we *measure* the spin of one particle, we will only get "up" or "down" with respect to our chosen axis, never something in between. (3/n)

(2) When we're not measuring, the spin can be a superposition of both up and down. (3) On the basis of the particular superposition it is in, we can predict the probability of a measurement outcome, but can't be certain. (4/n)

(4) Two particles may be (don't have to be) entangled: we don't know what answer we'll get if we measure either spin, but if we measure one we can instantly be 100% confident what the other will be, even if it's light-years away. (5/n)

What's up with that? How do they know? (6/n)

The formalism of quantum mechanics is perfectly clear about the answer, but sometimes it's hard to face up to it.

We are attached to our observations, and are reluctant to admit a gap between them and the underlying reality. (7/n)

We describe the two particles using a wave function, or "quantum state." It's a *single* wave function, even if it's two particles. Or any other number of particles. Just one wave function, not a separate one for each particle. (8/n)

So: there aren't two separate things. There is just one thing - the quantum state - with two parts to it.

So of course the measurement outcomes will be related. We're measuring different aspects of the same thing. (9/n)

The mystery is not "how can these outcomes be correlated even though they're far away?", it's "why is it ever useful to think of different parts of the wave function as representing `things far away from each other'?" (10/n)

I.e., there's not just a little bit of non-locality we have to learn to live with. QM is by its nature entirely, hilariously non-local - or "alocal," even better. Why are we so tempted by locality at all?

This question has an easy aspect and a hard one. (11/n)

Easy aspect: In the real world, the way that quantum states evolve is naturally thought of in terms of local variables. Particles have positions, fields have values at points in space. Those give very convenient ways of representing the quantum state. (12/n)

Almost too convenient. There are "no-signaling" theorems that say we cannot use quantum entanglement to send information faster than light. These theorems depend on the *specific* way that quantum states evolve -- the "laws of physics." (13/n)

Hard aspect: Why do the laws of physics take just the right form so as to allow us to describe the world (approximately, but really really well) in terms of stuff happening at "points in space"? (14/n)

Hope you don't expect me to answer that one. We don't know, but some of us have ideas! Science isn't finished yet, but we're working on it. Stay tuned. (15/15)