xouper wrote:ceptimus wrote:As an analogy, if I were able to send you 100 coins that had been entangled with 100 coins I keep here, then we could both flip them at any time (including simultaneously) and record 100 instances of heads or tails. We'd both get a random string of heads and tails - just as with regular coins - but when we meet up to compare notes we'd find that whenever you flipped 'heads' then I will have flipped the corresponding 'tails'.

Pardon my failure to fully understand your explanation, but I'm not sure that is a completely valid analogy. Are you suggesting that an entangled coin is not flipped until it is "measured"? How do you know the entangled pair of coins were not both flipped (simultaneously as a pair) at the moment they were created as an entangled pair but you simply do not know which one is heads until you look at one of them at a later time? What am I missing in your analogy?

You're taking the analogy too far. I was just saying that if it were possible to entangle two coins so that in one sense they behaved like entangled particles, then when one of them is flipped and the result recorded, the outcome of flipping the other coin is then predestined.

So when the first coin is flipped the outcome is just the usual random 50:50 distribution of heads/tails. The person that flipped it now knows in advance which way the other coin will land when it is flipped. He could phone the other person up and say, 'when you flip it I know it will land heads!' but of course that information is just travelling at regular telephone line speed. If you were trying to communicate faster than light, then it's no use - the second person just sees the usual random result of flipping a coin. It only becomes spooky if and when the first coin flipper is able to communicate with the second (by conventional light-speed-or-slower means) and is then able to predict the coin toss if it's not yet been flipped - or if it has been flipped the result is already known without being told.

It's the fact that the two flippers both know that they have the same coin toss result without speaking about it that makes the technique useful for cryptography - the secret message can be sent in a way that requires knowledge of the coin toss results to decode. The message still has to be sent by regular means but the key is sent by the spooky entangled means. It's impossible for an eavesdropper to 'see' the entangled part of the information as when he makes any attempt to measure it the quantum entanglement is broken and the two people who are trying to send the message would know and then not transmit the actual (encoded) secret message. The procedure would be: 1. send the entangled particles and measure them. 2. communicate by normal unencrypted means and make sure they both agree that various 'checksums' for the entangled stream match. Then they can be sure that they both hold the same "random" stream of results and the sender can use the random stream as a one-time key to encrypt the secret message and send it via normal open communication. A one-use random key is universally acknowledged to be a 100% secure way of encrypting and transmitting a message, and the entangled particles allow a one-time key to be shared by two people in a way that they can be sure (assuming quantum physics theory is correct) is immune to eavesdropping.