As I understood it, the paradox was that by the rules of the thread, “This comment will be massively upvoted. 100%” is something I should upvote if I believe it’s unlikely to be true. But if I upvote it on that basis, I should expect others to upvote it as well. But if I expect others to upvote it, then I should expect it to be upvoted, and therefore I should consider it likely to be true. But if I consider it likely to be true, then by the rules of the thread, I should downvote it. But if I downvote on that basis, I should expect others to downvote it as well, and therefore I should consider it unlikely to be true. But...
Everyone should agree that 100% certainty of something is infinitely overconfident. Then, everyone should upvote. Knowing this, I’m completely certain that I’ll get lots of upvotes, and so absurdly large amounts of certainty seem justified. And as a kicker, everyone said I was overconfident of something that turned out to be correct.
Obviously, there are other possibilities (like me retracting the comment before it can be massively upvoted), so (like usual) 100% certainty really isn’t justified. And unforseen consequences like that are exactly why you don’t play with outcome pumps, as the time turner story reminds us.
What paradox? There wasn’t even a paradox.
As I understood it, the paradox was that by the rules of the thread, “This comment will be massively upvoted. 100%” is something I should upvote if I believe it’s unlikely to be true. But if I upvote it on that basis, I should expect others to upvote it as well. But if I expect others to upvote it, then I should expect it to be upvoted, and therefore I should consider it likely to be true. But if I consider it likely to be true, then by the rules of the thread, I should downvote it. But if I downvote on that basis, I should expect others to downvote it as well, and therefore I should consider it unlikely to be true. But...
Naively:
Everyone should agree that 100% certainty of something is infinitely overconfident. Then, everyone should upvote. Knowing this, I’m completely certain that I’ll get lots of upvotes, and so absurdly large amounts of certainty seem justified. And as a kicker, everyone said I was overconfident of something that turned out to be correct.
Obviously, there are other possibilities (like me retracting the comment before it can be massively upvoted), so (like usual) 100% certainty really isn’t justified. And unforseen consequences like that are exactly why you don’t play with outcome pumps, as the time turner story reminds us.