= 27d567bc2d48d9f40e9ccc20ea370b15)
pragmatist comments on Open Thread, September 1-15, 2012 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (353)
I've decided I should educate myself about LW-specific decision theories. I've downloaded Eliezer's paper on timeless decision theory and I'm reading through it. I'm wondering if there are similar consolidated presentations of updateless and ambient decision theory. Has anyone attempted to write these theories up for academic publication? Or is the best place to learn about them still the blog posts linked on the wiki?
I'm currently researching TDT, UDT, and ADT. So far as I am aware, there have been no comprehensive presentations of UDT and ADT. Eliezer's paper itself is a step in the right direction, but is unfinished and has some major flaws.
SI has contracted the philosopher Rachael Briggs to write a paper on TDT for a peer-reviewed, academic journal. Last time I spoke to Luke about it, he said that the pre-print will be done sometime this winter. I don't know whether the pre-print will available to the general public, or just to internal researchers.
Edit: According to Nisan, the information in the second paragraph is out-of-date.
Rachael Briggs is no longer working on that project. It's been taken over by SI Research Fellow Alex Altair.
Ah, I was unaware. Thank you for the update. Was there any explanation given as to why she is no longer working on the project? Do we have a revised timeline for the paper's completion?
I think Briggs wanted to stop; I don't know why. And I don't know when the project will be completed.
Is there a place on this website (or elsewhere) where the major flaws in Eliezer's paper are pointed out and discussed?
There likely is, but I don't have any links off-hand. IMO, the major flaw is that some passages are dense and unclear. It's difficult to understand some explanations and examples as a result. Don't be discouraged if you have to re-re-re-read a part of a paper in order to decipher the meaning. I certainly had to.
Beyond that, people disagree about TDT itself and have tried to make revisions, as well as revisions to those revisions. (Hence UDT and ADT.) Those flaws are discussed in the blog posts on decision theory, as well in in comment sections. Even still, that information is dispersed and unorganized. So far as I can tell, most of it just exists within the minds of individuals and hasn't been formally written up yet.