# private_messaging comments on Problematic Problems for TDT - Less Wrong

34 29 May 2012 03:41PM

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

Sort By: Best

You are viewing a single comment's thread.

Comment author: 31 May 2012 11:28:35AM *  2 points [-]

I think we need a 'non-problematic problems for CDT' thread.

For example, it is not problematic for CDT-based robot controller to have the control values in the action A represent multiple servos in it's world model, as if you wired multiple robot arms to 1 controller in parallel. You may want to do this if you want the robot arms move in unison and pass along the balls in the real world imitation of http://blueballmachine2.ytmnd.com/

It is likewise not problematic if you ran out of wire and decided to make the '1 controller' be physically 2 controllers running identical code from above, or if you ran out of time machines and decided to control yesterday's servo with 1 controller yesterday, and today's servo with same controller in same state today. It's simply low level, irrelevant details.

Mathematical formalization of CDT (such as robot software) will one-box or two-box in newcomb depending to the world model within which CDT decides. If the world model has the 'prediction' as second servo represented by same variable, then it'll one-box.

Philosophical maxims like "act based on consequences of my actions", whenever they one box, or two box, depend in turn solely on philosophical questions like "what is self" . E.g. if "self" means the physical meat, then two-box, if "self" means the algorithm (a higher level concept), then one-box if you assume that the thing in predictor is "self" too.

edit: another thing. Stuff outside robot's senses is naturally uncertain. Upon hearing of the explanation in Newcomb's paradox, one has to update the estimates of what is outside the senses; outside might be that the money are fake, and there's some external logic and wiring and servos that will put real million into a box if you choose to 1-box. If the money are to pay for, I dunno, your child's education, clearly one got to 1-box. I'm pretty sure Causal Deciding General Thud can 1-box just fine, if he needs the money to buy the real weapons for the real army, and suspects that outside his senses there may be the predictor spying. General Thud knows that the best option is to 1-box inside predictor and 2-box outside. The goal is never to two box outside the predictor.