The actual scenario is full of distractions, but I'll try to ignore them (1).
The thing is, I think the pain in this scenario is a distraction as well. The relevant property in this sort of scenario, which drives my impulse to prevent it and causes me to experience guilt if I don't, is my inference of suffering (2).
So the question becomes, how do I characterize the nature of suffering?
Which is perhaps a mere semantic substitution, but it certainly doesn't feel that way from the inside. I can feel pain without suffering, and suffering without pain, which strongly suggests that there are two different things under discussion, even if I don't clearly understand either of them.
I'll probably play a second round against the GLUT, since if there is any suffering involved there it has already happened and I might as well get some benefit from it. (3)
The others, I am less certain about.
Thinking about it more, I lean towards saying that my intuitions about guilt and shame and moral obligation to reduce suffering are all kind of worthless in this scenario, and I do better to frame the question differently.
For example, given #3, perhaps the right question is not "are those uploads experiencing suffering I ought to alleviate" but rather "ought I cooperate with those uploads, or ought I defect?"
Not that that helps: I'm still left with the question of how to calibrate their cost/benefit equation against my own, which is to say of how significant a term their utility is in my utility function. And sure, I can dodge the question by saying I need more data to be certain, but one can fairly ask what data I'd want... which is really the same question we started with, though stated in a more general way.
So... dunno. I'm spinning my wheels, here.
==
(1) For example, I suspect that my actual response in that scenario is either to keep all the money, under the expectation that what I'm seeing is an actor or otherwise something not experiencing pain (in a non-exotic way, as in the Milgram experiments), or to immediately quit the experiment and leave the room, under the expectation that to do anything else is to reinforce the sadistic monster running this "experiment."
I cannot imagine why I'd ever actually press the button.
But of course that's beside the point here.
(2) That is, if the person wired up to the chair informs me that yes, they are experiencing pain, but it's no big deal, then I don't feel the same impulse to spend $100 to prevent it.
Conversely, if the neurologists monitoring the person's condition assures me credibly that there's no pain, but they are intensely suffering for some other reason, I feel the same impulse to prevent it. E.g., if they will be separated from their family, whom they love, unless I return the $100, I will feel the same impulse to spend $100 to prevent it.
The pain is neither necessary nor sufficient for my reaction.
Note that I'm expressing all this in terms of what I perceive and what that impels me to do, rather than in terms of the moral superiority of one condition over another, because I have a clearer understanding of what I'm talking about with the former. I don't mean to suggest that the latter doesn't exist, nor that the two are equivalent, nor that they aren't. I'm just not talking about the latter yet.
(3) I say "probably" because there are acausal decision issues that arise here that might make me decide otherwise, but I think those issues are also beside your point.
Also, incidentally, if there is any suffering involved, the creation of the GLUT was an act of monstrous cruelty on a scale I can't begin to conceive.
Related To: Eliezer's Zombies Sequence, Alicorn's Pain
Today you volunteered for what was billed as an experiment in moral psychology. You enter into a small room with a video monitor, a red light, and a button. Before you entered, you were told that you'll be paid $100 for participating in the experiment, but for every time you hit that button, $10 will be deducted. On the monitor, you see a person sitting in another room, and you appear to have a two-way audio connection with him. That person is tied down to his chair, with what appears to be electrical leads attached to him. He now explains to you that your red light will soon turn on, which means he will be feeling excruciating pain. But if you press the button in front of you, his pain will stop for a minute, after which the red light will turn on again. The experiment will end in ten minutes.
You're not sure whether to believe him, but pretty soon the red light does turn on, and the person in the monitor cries out in pain, and starts struggling against his restraints. You hesitate for a second, but it looks and sounds very convincing to you, so you quickly hit the button. The person in the monitor breaths a big sigh of relief and thanks you profusely. You make some small talk with him, and soon the red light turns on again. You repeat this ten times and then are released from the room. As you're about to leave, the experimenter tells you that there was no actual person behind the video monitor. Instead, the audio/video stream you experienced was generated by one of the following ECPs (exotic computational processes).
Then she asks, would you like to repeat this experiment for another chance at earning $100?
Presumably, you answer "yes", because you think that despite appearances, none of these ECPs actually do feel pain when the red light turns on. (To some of these ECPs, your button presses would constitute positive reinforcement or lack of negative reinforcement, but mere negative reinforcement, when happening to others, doesn't seem to be a strong moral disvalue.) Intuitively this seems to be the obvious correct answer, but how to describe the difference between actual pain and the appearance of pain or mere negative reinforcement, at the level of bits or atoms, if we were specifying the utility function of a potentially super-intelligent AI? (If we cannot even clearly define what seems to be one of the simplest values, then the approach of trying to manually specify such a utility function would appear completely hopeless.)
One idea to try to understand the nature of pain is to sample the space of possible minds, look for those that seem to be feeling pain, and check if the underlying computations have anything in common. But as in the above thought experiment, there are minds that can convincingly simulate the appearance of pain without really feeling it.
Another idea is that perhaps what is bad about pain is that it is a strong negative reinforcement as experienced by a conscious mind. This would be compatible with the thought experiment above, since (intuitively) ECPs 1, 2, and 4 are not conscious, and 3 does not experience strong negative reinforcements. Unfortunately it also implies that fully defining pain as a moral disvalue is at least as hard as the problem of consciousness, so this line of investigation seems to be at an immediate impasse, at least for the moment. (But does anyone see an argument that this is clearly not the right approach?)
What other approaches might work, hopefully without running into one or more problems already known to be hard?