> There are plenty of good reasons not to be a rationalist. In fact, valuing the truth as a sort of Platonic ideal to which everything else must be subservient

> There are plenty of good reasons not to be a rationalist. In fact, valuing the truth as a sort of Platonic ideal to which everything else must be subservient is an arbitrary moral choice in a universe where moral facts are either non-existent or inaccessible. The rationalists' position is that believing true things is almost always strategically superior to believing untrue things. This is an empirical claim, not one about "Platonic ideals". You can choose to value whatever you want, but presumably you have some ends you want to achieve, which requires decision making, which is usually assisted more by true beliefs than by false beliefs. Now, there's some evidence about cases where this might not be true. The blog post raises this about happiness surveys, etc. This evidence seems to me to weakly establishes a negative correlation, but not necessarily that rationality causes less happiness. At any rate, for any particular subgoal (say, becoming better at programming, achieving material wealth, etc), believing true things is probably going to be good strategy. That's the rationalist claim, anyway. For happiness in general, it gets murkier. Yudkowsky also has a whole thing about how a rationalist can still win at Newcomb-like problems. The gist of it is, what's rational is whatever will meet your goals. This means if you need to preserve some false beliefs, you should go ahead and do that. For example, you may find false beliefs about how long a project will take motivate you to get started. If you really accepted that your project would take a long, long time to bear fruit, maybe you'd never have begun, and you'd have been worse for that. In this case you've got this other irrationality about how you're weighing future benefits that you're compensating for by lying to yourself about how far into the future you'll reap your rewards. Here, the two irrationalities together take you to a local maximum. If you only got rid of the false belief about the project times, you'd be worse off. But, if you could get rid of both irrationalities you'd be _better_ off. You wouldn't have the motivation problem AND you could plan more efficiently. But maybe you just can't get rid of your tendency to over-discount future benefits. If so, you should just accept the local maximum, rather than anti-strategically attacking your tendency to underestimate project costs and leaving yourself worse off. You could make a similar argument about religion and many other false beliefs. Which local maxima we can realistically step beyond is another empirical question. But I think it's clear that there are many false beliefs that are strictly negative: that is, you're better off without them, even if you don't change anything else about your thinking.