Saturday, 3 January 2015

Does an entangled particle “know” the state of the other particle?

No, it doesn’t.  Here is an excellent explanation by  Qiaochu Yuan, University of California, Berkeley. 

People think entanglement is a quantum phenomenon, but it's not. It is a probabilistic phenomenon, and it has both classical and quantum instantiations. Classical entanglement is entirely possible provided that you're uncertain about something in the universe (and who isn't?), and people usually call it correlation.

Here's a simple example. I'm going to draw a card from a deck, but I'm not going to tell you what it is. I am going to write the name of the card on a piece of paper. Then I'm going to send the card to Mars and the piece of paper to Jupiter. If you're on Mars and you look at the card, you now know - instantaneously- what the piece of paper on Jupiter says. Spooky action at a distance indeed!

So that was classical entanglement. As you can see, the "mechanism" causing the piece of paper and the card to be "entangled" is just that I did something that made your uncertainty about the piece of paper and your uncertainty about the card correlated. It's not like there was some kind of mysterious physical process transmitting information from Jupiter to Mars. There was some not-at-all-mysterious physical process transmitting information from the card to the paper on Earth, and it was me looking at the card and then writing on the paper.

The thing that's actually mysterious here is a more general mysterious feature of quantum mechanics, which is that in a classical setting we can convince ourselves that working probabilistically just reflects our ignorance of a deterministic and in principle fully knowable universe, but in a quantum setting we can't: quantum mechanics seems to force an inherent kind of uncertainty on us. 

In particular, while a classical Bayesian update (the thing you're doing when you look at the card on Mars) looks like something that happens in our brains, a quantum Bayesian update (the thing you're doing when you "collapse a wave function") looks like something that happens in the universe. I admit I'm confused about this as well. Maybe everyone is?

No comments:

Post a Comment