I’m in the same boat. I don’t have anything against music, but never derived pleasure from it like other people seemed to. I’ve enjoyed particular songs, e.g. stuff from Dr. Horrible’s Sing-Along Blog, because I can enjoy cleverness and humor and a lot of songs have content that is clever or funny. But the music itself is background at best.

What I feel when listening to Bach isn't what someone else feels about a song they dislike; more like what you feel about an overheard conversation with nothing to do with you. Or a speech on an issue you don't care about. I have tried to change this, because of utilitarianism, but it turns out it's hard.

I think the obvious solution is basically this, with a Time-Turning involved. The troll could be real, or not (probably is). The hardest part about changing the past is faking the evidence including memories, but with a False-Memory Charm that becomes trivial. Memory charm Harry and possibly Dumbledore as well, depending on whether he objects "but I remember feeling a student die." They won't do it this way because it's too finger snap-ish and not dramatic enough, but if it's not at least addressed then I will allege a holding of the Idiot Ball.

Harry will need the help of Dumbledore or Quirrel to unlock the time-turner and cast the memory charms. Quirrel wouldn't help but I'd be interested to see his excuse; Dumbledore should be possible to convince but might not be. McGonagall or someone might be capable of it but wouldn't obliviate Albus without him asking for it. Unless this is what was foreshadowed with the question about her first loyalty?

I said not receiving a CD from the future is the most likely because that's what usually happens. But I do have a pretty huge sampling bias of mainly talking to people who don't have time machines.

i would expect "no CD" to be the most common even if you do have one, just because I feel like a closed time loop should take some effort to start. But this is probably a generalization from fiction, since if they happen in the real universe they do "just happen" with no previous cause. So I guess I can't support it well enough to justify my intuition. I will say that if I'm wrong about this, any time traveller should be prepared for these to happen all the time on totally trivial things.

The Novikov Self-Consistency Principle can help answer that. It is one of my favorite things. I don't think it was named in the post, but the concept was there.

The idea is that contradictions have probability zero. So the first scenario, the one with the paradox, doesn't happen. It's like the Outcome Pump if you hit the Emergency Regret Button. Instead of saying "do the following," it should say "attempt the following." If it is one self-consistent timeline, then you will fail. I don't know why you'll fail, probably just whatever reason is least unlikely, but the probability of success is zero. The probability distribution is virtually all at "you send the same number you received." (With other probability mass for "you misread" and "transcription error" and stuff).

If your experiment succeeds, then you are not dealing with a single, self-consistent universe. The Novikov principle has been falsified. The distribution of X depends on how many "previous" iterations there were, which depends on the likelihood that you do this sequence given that you receive such a CD. I think it would be a geometric distribution?

The second one is also interesting. Any number is self-consistent. So (back to Novikov) none of them are vetoed. If a CD arrives, the distribution is whatever distribution you would get if you were asked "Write a number." More likely, you don't receive a CD from the future. That's what happened today. And yesterday. And the day before. If you resolve to send the CD to yourself the previous day, then you will fail if self-consistency applies

Have you read HPMoR yet? I also highly recommend this short story.

Not just because I can. Maybe for other reasons, like the fact that I still care about the punier humans and want to make it better for them. That depends on preferences that an AI might or might not have.

It's not really about what I would do; it's the fact that we don't know what an arbitrary superintelligence will or won't decide to do.

(I'm thinking of "superintelligence" as "smart enough to do more or less whatever it wants by sheer thinkism," which I've already said I agree is possible. Is this nonstandard?)