>Which tells me I have not properly understood what's going on. What I had thought is that someone creates a pair of entangled particles, call them E and O. The second particle is moved to the Chinese facility in orbit about 500 miles above the earth. Someone on earth changes the state of E, and an observer notices that the state of O changes to match at the same time.
>Wednesday night I met an old church friend for dinner and had a long debate over this. At the time we were focused on simultaneity, and it was my thesis that in order to know that O's change instantly followed E's, they had to have synchronized clocks*. We also discussed the implications for communication by ansible: For that to work, you'd have to ship millions of entangled bits from E to O, and each bit would have to be addressable, which we're certainly not ready to do yet.
>But in this case apparently they did have a million bits at each end, and they were addressable...if my picture was accurate. So I must be wrong about their method. Start back at the beginning: Can anyone tell me how you a) create two entangled bits and then b) separate them by distance? I don't have a really scientific forum to hang out at, but it's been remarked before what a broad range of knowledge we have here so maybe...?
>[Some hours later:] I forgot to add that, depending on your answer to the above question, the next step may be to wonder why 0.1% success is considered, well, successful. If you change the state of particle E and the state of particle O changes only every thousandth time, what are we to believe is proven?
>* About synchronized clocks: Yes, I understand that relativity denies the concept of simultaneity. I maintain the possibility of synchronicity, nevertheless, on two grounds: 1) If it's true that particles O and E changed simultaneously, that doesn't violate the speed of light because nothing has traveled from E to O. No violation of c, therefore no violation of the ban on simultaneity. 2) If you insist on forbidding simultaneity (as my friend did) despite the fact that this case doesn't affect c, it's still alright: We can synchronize clocks practically speaking (close enough for government work) as follows: Calibrate clock E on earth to run accurately under 1g. Calibrate clock O in orbit to run accurately in free fall. Calculate td as the time it takes light to travel from E to O. At time t, send a tick from E to O. Set clock O to t minus td. From then on, clocks E and O are synchronized, practically speaking. If td is, say, 5e-2s and particle O changed its step 1e-12s after particle E, then simultaneity notwithstanding we know the effect is essentially simultaneous—at any rate it certainly exceeds c.