I've been thinking about a communication system that could allow faster than light communication. The basic principle that would make this system work come from the double-slit experiment and Wheleer's delayed choice experiment, taking advantage of the different patterns that would appear on a double-slit experiment if both slits are open (wave interference) or only one slit is open (particles). The double-slit experiment is closely related to wave-particle duality. Basically, it shows that the behaviour of quantum objects (photons, electrons, protons and even molecules) depends on the methods used to measure them. If we have a source of quantum objects, and put two very close slits on their way, they end creating an interference pattern when detected on a screen (even if the quanta arrive one at a time). This means they behaved as waves. But if we close one slit, then the pattern changes, interference dissapears, and they behave as particles. The Wheeler's delayed choice experiment states that we can not trick quantum objects "in flight" to be waves or particles before really detecting them, no matter what we do to the setup or how many times we change the number of slits until the final detection. So the presence of one or two slits at detection time will decide if we end up detecting a particle or a wave pattern. We could say that all the detection screen impacts are instantaneously correlated with the presence of one or two open slits on the system, irrespective of distance. The whole communication system (transmitter, channel and receiver) can be seen as a very big, elongated Mach-Zehnder interferometer. The transmitter (Tx) is composed of: A coherent light source that will work continuously, creating the communication channel. A beam splitter that splits the light from the source into two perpendicular paths. A mirror at 45º that redirects light from one path, so both paths get parallel. An optical switch that enables or disables one of the parallel light paths. The receiver (Rx) is composed of: A mirror at 45º that redirects the light on the path that was not redirected before, making both paths orthogonal and equal in length. A beam splitter that will receive the light coming from one or both paths. Two detectors at the other sides of the beam splitter, that will fire depending on light travelling one path (particles) or two paths (waves). The two long parallel light paths between transmitter and receiver comprise the communication channel. One of those light paths will be switched on and off using the Tx optical switch: When one light path gets disabled, light at the receiver is detected as having travelled just the path that is now available, so the photons received must behave as particles. As the Rx beam spliter can equally transmit or reflect particles, both Rx detectors will fire with 50% probability. When both light paths are enabled, light is detected as having followed two equal paths, so photons reach the Rx beam splitter and behave as a wave interfering with itself, thus, one Rx detector will fire all the time (100%) while the other won't fire at all (0%). Theoretically, there is no problem assuming a very long distance between transmitter and receiver, but practically, we must achieve a great level of accuracy placing all the elements for the system to work as intended. Also, when the transmitter light source is started for the first time we have to wait for the light to reach the receiver to stablish the communication channel, but once the channel is set, we have a way to know the transmitter state almost instantaneously at the receiver, irrespective of Tx/Rx distance. The correlation between Tx switch and Rx beam splitter is, in fact, instantaneous, but as we have to wait for some statistics to build up on the detectors (to distinguish the states correctly), communication between Tx and Rx can not be instantaneous, but it can be superluminal. This is so because the time to discriminate between the particle/wave pattern depends just on the Tx light source emission speed, not on the actual transmitter-receiver distance. This means the time we have to wait at Rx to discriminate Tx states can be less than the time light really takes going from Tx to Rx. Weird, isn't it? As an example, imagine the Tx light source fires 100 million photons per second (so each second you could have 100 million hits at the Rx detectors to distinguish between the particle/wave patterns). Suppose you only need 100 hits to decide between states. That means you could change the Tx switch at a microsecond rate and still be able to "decode" Tx switch states on the receiver. So you are able to know Tx states each microsecond... irrespective of Tx/Rx distance! If the transmitter is placed more than 300 meters away from the receiver, we get superluminal information. The actual distance (and time) it takes for the individual photons to reach the receiver is irrelevant, because the information the photons give when detected is encoded in their positions, in the instantaneous correlation between that photon impact and the actual state of the lightpaths present at detection time. I mean, photons take time at lightspeed to arrive to the receiver, but when they reach the detectors, they give information about the current state of the Tx switch at detection time, not the state the switch had when photons left! I think this experiment could be tested for real almost inexpensively with technology available today. What do you think?