You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

linkhyrule5 comments on Harry Potter and the Methods of Rationality discussion thread, part 23, chapter 94 - Less Wrong Discussion

8 Post author: elharo 08 July 2013 12:04PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (343)

You are viewing a single comment's thread. Show more comments above.

Comment author: linkhyrule5 10 July 2013 01:39:57AM 3 points [-]

This isn't quite Newcomb's Problem, though.

Consider this: You have essentially the same set up, and you're a one-boxer, so you walk up and take your million dollars out of box B.

And then Omega comes back and says, "Oh, by the way, want to open Box A too?"

And you say, "Um. Okay?" And open Box A, and get a thousand dollars. Or not. Doesn't really matter - you already have your million dollars, so you have nothing to lose by opening Box A.

This differs from the traditional two-boxing argument in one very important respect: you get new information in the middle of the experiment. Your single Omega-predictable algorithm doesn't have to "change its mind" in the middle, it gets interrupted.

This is essentially what (might be) happening here. Harry has opened Box B and found a dead Hermione inside. That's set and done. Assuming that he has reason to be believe that Box A will help him more than it will hurt him on average (and won't contain, say, a dead/mindless/insane Fred and George), he has no reason not to open Box A.

Comment author: Kindly 10 July 2013 04:41:39AM 2 points [-]

A more perfectly isomorphic variant of Newcomb's problem is the following. Both boxes are transparent, and Omega acts according to the following rule: if you two-box when box B is empty, then box B is always empty, while if you one-box, box B is empty with a 50% chance.

If you one-box in this variant, you win half a million dollars in expectation. If you intend to two-box should you see that box B is empty, then you only win a thousand dollars.

Comment author: linkhyrule5 10 July 2013 05:02:43AM *  0 points [-]

... Perhaps, but that's no longer isomorphic to the actual problem. We are past the point of Omega's influence; in Newcomb's problem, it'd be as if Omega grabbed your mind state, ran it forward until it confirmed that you would not initially two-box, and then stopped. Omega itself has been removed from the problem, and you're left with one empty box that you've already claimed a million dollars from (or negative one million, in this case) and one closed box.

Near as we can tell, history can't change in the MoRverse: what's done is done, so Harry might as well exploit it.