Death Note, Anonymity, and Information Theory

gwern

Death Note, Anonymity, and Information Theory — LessWrong

Comment Permalink

The cat-and-mouse games between L and Light could be turned into decision theory problems; the rules related to ownership of the notebook could be used to explore timeless reasoning (much like the film Memento).

Eliezer's Timless Decision Theory is interesting, but I don't yet understand what real problem it is solving. He worked it into his Harry Potter fanfic when Harry was dealing with Azkaban (sp?), and given that he wrote the paper I can see why he made use of it, but here's someone else saying it's useful.

Does timeless reasoning have any application in situations where your opponents can't read your mind?

(I haven't watched Death Note, so my apologies if the answer to this question is obvious to people who have.)

Showing 3 of 5 replies (Click to show all)

Vladimir_Nesov15y120

Does timeless reasoning have any application in situations where your opponents can't read your mind?

We all read each other's minds to some extent, and to the extent this happens, TDT will give better advice than CDT. See section 7 of the TDT paper:

"Modeling agents as influenced to some greater or lesser degree by "the sort of decision you make, being the person that you are", realistically describes present-day human existence."

0[anonymous]15y

Even if they can't read your mind with 100% accuracy, if they have some ability to predict your cognition, CDT will go astray - for example, in Parfit's Hitchhiker. Also, TDT allows co-operation in prisoners' dilemmas with copies of yourself.

3benelliott15y

One reason reason is that it seems like it might be helpful with friendliness proofs, particularly the part where you have to prove the AI's goal will remain stable over millions of self-modifications (the harder, and all too frequently ignored, side of the problem). Basically, it takes dilemma's which might otherwise tempt an AI to self-modify, and shows that it need not have to. I think with CDT you can prove an AI won't need to modify its goal system on action-determined problem, while with TDT you can prove the same for the broader class of decision-determined problems. This leaves many issues, but its a step in the right direction. Disclaimer: The above post should not be taken to speak for Eliezer Yudkowsky, SIAI, or anyone other than me. I am not in any way a member of SIAI or any other similar organization. There is a good chance that I am talking out of my arse.

See in context

60

Death Note, Anonymity, and Information Theory

60

60

60

Death Note, Anonymity, and Information Theory

60

60