CarlShulman comments on UDT agents as deontologists - Less Wrong

8 Post author: Tyrrell_McAllister 10 June 2010 05:01AM

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

Comments (109)

You are viewing a single comment's thread.

Comment deleted 10 June 2010 02:08:07AM *  [-]
Comment author: Vladimir_Nesov 10 June 2010 02:13:12AM *  1 point [-]

With respect to (1), UDT maximizes over worlds where the zillionth digit of pi is 1, 2, 3...8, 9, 10.

These are not different worlds for UDT, but a single world that can have different possible execution histories that state zillionth digit of pi to be 0,1,...,9. Mathematical intuition establishes a probability distribution over these execution histories for the fixed world program that defines the subject matter.

Comment author: Tyrrell_McAllister 10 June 2010 04:14:46AM 0 points [-]

It seems this post could benefit from distinguishing between possible and impossible possible worlds.

That may be so, but I need to think about how to do it. I said that the possible worlds are whatever the agent's builders thought was possible. That is, "possibility" refers to the builders' ignorance, including their ignorance about the zillionth digit of pi.