Anonymous asked: The new decision theory paper exists because the old TDT paper did not have any formalism at all. The not really new FDT is the obvious way to formalize the hand waving in the original TDT paper but otherwise there is no new content. I think the paper exists because they decided they should at least write down a formalism somewhere.
Huh, okay, that makes some sense.
In the comment thread on MIRI’s blog post announcing the paper, I found the following comment from Rob Bensinger, answering the question “How does this relate to TDT?”:
FDT is an improvement over TDT that adds updatelessness. FDT is similar to what people usually mean by “UDT” (though “UDT” has come to refer to a lot of different things over the years and is less well-defined). E.g., Wei Dai’s UDT 1.1 ( http://lesswrong.com/lw/1s5… ) is a particular version of FDT, with FDT as the umbrella category.
UDT 1.1 was a big improvement over TDT, but it wasn’t perfect (though it may be the best we’ve got), it built in some debatable assumptions (over and above what’s needed to show why TDT, CDT, and EDT don’t work), and it was somewhat harder to explain than TDT. FDT as presented here is agnostic about the weirdest parts of UDT 1.1, and is about as easy to explain as TDT.
I never really looked into UDT, so I don’t actually know what “updatelessness” means here, but I guess it distinguishes FDT from the original TDT. Based on this comment, it looks like they are trying to specify a minimal set of desiderata that their new decision theory must satisfy, even if they don’t have the theory itself (i.e. don’t have a working algorithm completely free of hand-waving).
