	I read a paper once, by David Chalmers, which argues that "functionally equivalent systems" will have the same experiences, whether the systems are two brains, a brain and a computer, two computers, or anything. That is, a system of neurons with connections between them (i.e. a brain) and for example a system of microchips with connections between analogous to the brain's connections, and analogous results when those connections are activated, will have the exact same experiences as the brain it models. The thought experiment he constructs to reach this conclusion is very complicated to describe but sheds much light on the problem of whether two people can have different color experiences. To summarize his argument, imagine an advanced neuroscience by which you can isolate the system of a person's brain that is responsible for color perception, and that you can construct a computer which has the exact same behavior as that brain system. That is, when certain input is given to that section of the brain (in the form of the patterns of activation of incoming neurons), it will return a certain output (in the form of the patterns of activation of neurons going out of the system), and that the computer you construct will have these exact same input/output pairs. Now you set it up so that the computer is attached to the person's brain, but it is not always turned on. You have a switch, and when you flip the switch, all the brain connections going into or coming out of the person's natural color processing area of his brain, are instead suddenly going into and coming out of the computer. When you flip the switch back, the computer is turned off and the person's natural color processing areas are doing the work of perceiving colors again. You can flip the switch as much as you want, with no damage to the brain or the computer, and by flipping the switch to "computer", you are basically replacing a portion of the person's brain with the computer system, and the replacement is immediate. Furthermore, if the person is using one of the systems, and in the process of perceiving a color, that system undergoes a change, the other system that is not being used will undergo an analogous change. So, for example if the switch is flipped to "brain" and the person sees a red apple, and synaptic changes occur to record the memory of the red apple, then the idle computer will undergo similar physical changes at the same time. The most important feature of this setup is that the two systems be functionally equivalent, their input/output pairs as well as the changes they undergo must be exactly the same. As a consequent of this functional equivalency, the person's behavior (including linguistic behavior) will be exactly the same regardless of which system is switched on. This leads to the following bizarre consequence: suppose that although the systems are functionally the same, the use of one system makes the person see red cards as red while the use of the other system makes him see red cards as green. Therefore, as you sit flipping the switch back and forth once a second, the person who is hooked up to the system is looking at a red card, experiencing it switching back and forth between red and green, and saying "no, I don't see anything unusual at all, it's just a plain red card" (because the systems are functionally equivalent). This incongruity between real experience and reported experience leads Chalmers to conclude that functionally equivalent systems must have the same experiences as eachother. It doesn't matter that our current technology doesn't allow for such a system to be constructed, because as far as we know there is nothing physically impossible about the setup of such an experiment, and if it was set up, then the switch back and forth between experiences while reporting the same experience would be a highly improbable result. 
