= 27d567bc2d48d9f40e9ccc20ea370b15)
Eliezer_Yudkowsky comments on Open thread, Nov. 24 - Nov. 30, 2014 - 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 (317)
Weird fictional theoritical scenario. Comments solicited.
In the future, mankind has become super successful. We have overcome our base instincts and have basically got our shit together. We are no longer in thrall to Azathoth (Evolution) or Mammon (Capitalism).
We meet an alien race, who are way more powerful than us and they show their values and see ours. We seek to cooperate on the prisoner's dilemma, but they defect. In our dying gasps, one of us asks them "We thought you were rational. WHY?..."
They reply " We follow a version of your meta-golden rule. Treat your inferiors as you would like to be treated by your superiors. In your treatment of super intelligences that were alive amongst you, the ones you call Azathoth and Mammon, we see that you really crushed them. I mean, you smashed them to the ground and then ran a road roller, twice. I am pretty certain you cooperated with us only because you were afraid. We do to you what you did to them"
What do we do if we could anticipate this scenario? Is it too absurd? Is the idea of extending our "empathy" to the impersonal forces that govern our life too much? What if the aliens simply don't see it that way?
That's not how TDT works.
Is TDT accurately described by "CDT + acausal comunication through mutual emulation"?
Communication isn't enough. CDT agents can't cooperate in a prisoner's dilemma if you put them in the same room and let them talk to each other. They aren't going to be able to cooperate in analogous trades across time no matter how much acausal 'communicaiton' they have.
I view TDT as a bit unnatural, UDT is more natural to me (after people explained TDT and UDT to me).
I think of UDT as a decision theory of 'counterfactually equitable rational precommitment' (?controversial phrasing?).
So you (or all counterfactual "you"s) precommit in advance to do the [optimal thing], and this [optimal thing] is defined in such a way as to not give preferential treatment to any specific counterfactual version of you. This is vague. Unfortunately the project to make this less vague is of paper length.
:)
Folks working on UDT, feel free to chime in to correct me if any of above is false.
But isn't UDT relying on perfect information about the problem at hand?
If this is so, could it be seen as the limit of TDT with complete information?