DaveInNYC comments on Raising the Sanity Waterline - Less Wrong

112 Post author: Eliezer_Yudkowsky 12 March 2009 04:28AM

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

Comments (207)

You are viewing a single comment's thread. Show more comments above.

Comment author: DaveInNYC 23 January 2010 10:58:50PM 1 point [-]

We think that religions are false, and a shared priority of Less Wrong denizens is to >believe things that are true instead

You'll get no argument from me that religions are false. You will get practically no argument from me that it makes sense to want to believe things that are true. What I question is, is it always rational to make others believe things that are true? If I leave my lights on when I leave the house so that would-be robbers think I am home when I am not, I am making a rational decision to make others believe something that is false.

If I am playing the Prisoner's Dilemma with someone (just once, so no tit-for-tat or anything), and I have the choice of making my opponent either act rationally or irrationally, the rational thing for me to do is make him act irrationally.

Comment author: Vladimir_Nesov 23 January 2010 11:05:01PM 3 points [-]
Comment author: DaveInNYC 23 January 2010 11:37:04PM 1 point [-]

Thanks; the Bayesians vs. Barbarians post is exactly the kind of thing I was looking for. I'll have to read some of the posts that it links to (as well as re-read the background posts you referred to; haven't read them in a while), as the way it stands I still think the Barbarians would win.

Comment author: MichaelGR 24 January 2010 05:22:25AM 1 point [-]

If I am playing the Prisoner's Dilemma with someone (just once, so no tit-for-tat or anything), and I have the choice of making my opponent either act rationally or irrationally, the rational thing for me to do is make him act irrationally.

Rationalists should win.

Comment author: Alicorn 23 January 2010 11:13:51PM 1 point [-]

is it always rational to make others believe things that are true?

This depends on your values. If chief among them is "honesty", and you caveat the "make others believe things that are true" with a "for the right reasons" clause, then probably, yeah. If honesty has to compete with things like keeping your property, maybe not.

If I am playing the Prisoner's Dilemma with someone (just once, so no tit-for-tat or anything), and I have the choice of making my opponent either act rationally or irrationally, the rational thing for me to do is make him act irrationally.

I'm not sure what the content of "making your opponent behave (ir)rationally" is supposed to be. It's certainly not an uncontroversial tidbit of received wisdom that the rational thing to do in the Prisoner's Dilemma is to defect, which is what you seem to imply.

Comment author: DaveInNYC 23 January 2010 11:20:14PM -1 points [-]

I'm not sure what the content of "making your opponent behave (ir)rationally" is >supposed to be. It's certainly not an uncontroversial tidbit of received wisdom that the >rational thing to do in the Prisoner's Dilemma is to defect, which is what you seem to >imply.

Exactly, if I was able to make him act irrationally, he would not defect, whereas I would. And if the definition of rationality is that it makes you win, then it can be perfectly rational to have others act irrationally (i.e. believe wrong things).

Comment author: Vladimir_Nesov 23 January 2010 11:25:30PM *  3 points [-]

If you both cooperate, instead of you both defecting, you'd both be better off, which is a more rational (more winning) outcome. Thus, making "cooperation" a synonym for "irrational" will irk people around here. (Of course if you defect and the other player cooperates, you'd have the best possible payoff.)