I call BS. If that's true, then Victor can not only send information across a distance, but back in time. Victor is changing the correlation of Alice and Bobs measurements after they have already been made. That's going to need serious verification for me to accept.
Physicists are increasingly convinced that time is an emergent property that arises due to low level quantum effects, principally entanglement [1].
The experiment in the Ars Technica article is the equivalent of the universe using lazy evaluation. It's really not so mysterious if you think outside the box of classical physics.
Regarding sending information back in time, from what I've read physicists are divided on the issue. Most agree that sending matter back in time is impossible, but information is still up for debate. In this particular experiment, however, any past observer who measured the information sent back in time would have collapsed the entangled quantum state and prevented the experiment from being conducted successfully. So being able to "successfully" send information back in time doesn't appear too useful if you can't read it.
>> Also, do you have some explanation for violations of Bell's inequality that don't rely on spooky action?
The article that started this thread is one. In some ways it's not even important that the model be a 100 percent match with reality. It behaves very similarly to reality and it matches the Bell inequalities without spooky action at a distance. Bells statement pretty much says this can not exist, yet there it is. So now we can stop talking about models that can not be - because they do - and see if any of them is actually a good fit for reality. Of course I'm assuming the math in the paper turns out to be correct upon review.
