It is getting spookier with experiments where each photon, say, hits a semi-opaque mirror and chooses a path randomly. The Geneva experiment was doing precisely

It is getting spookier with experiments where each photon, say, hits a semi-opaque mirror and chooses a path randomly. The Geneva experiment was doing precisely that: You start with an ultraviolet photon and split it into two photons. One goes one way and the other goes another way, both to identical interferometers. Entering its own interferometer, each photon must make a random decision as to whether it will travel a long pathway through the device or a short one. Then you look for a correlation between the pathways taken by the photons in their respective interferometers. If the timing between the photons is exactly adjusted, each twin seems to know what the other is doing and matches its choice of pathway to coincide with that of its distant partner. (upd) From http://www.cebaf.gov/news/internet/1997/spooky.html http://www.cebaf.gov/news/internet/1997/spooky.html