Stefan Schubert points out a source of bias: putting yourself in the shoes of only one party, rather than all parties concerned. To avoid bias, you should look at things from all relevant angles. 1/10 https://twitter.com/StefanFSchubert/status/1351660903478009860

This strategy will indeed lead to less bias on your part. Yet it is exactly what I'm trying to do less.

In private life, you'll often find yourself in a situation in which A tells you what an asshole B was, usually to A's disadvantage. 2/

A is looking to have his emotionally-driven assessment of the situation confirmed. Hence you should usually (not always!) agree that B was indeed being an asshole. You may want to expand on that. You'll rarely go wrong by calling B a wanker and a moron. 3/

Biased? Probably. But here's the thing. Rationality is an important value. So is friendship. By calling B a moron, you're doing your friend a favour & at the same time signaling loyalty, both of which are good. 4/

More generally, there's always a meta-decision to be made whether rationality is the only thing you want to maximize. Often the best answer to that question is "no". 5/

I guess the kinds of people who think about these questions are also the most likely to be biased in favour of going for maximizing rationality only. This is suboptimal, especially when you're dealing with normies, who, as the name suggests, are the majority of people. 6/

For example, suppose you want to move people to behave altruistically in a more effective manner. Don't hit them with a meta-analysis of randomized controlled trials! Instead, go in through the back door. 7/

*First* establish that people who are *really* caring will care to cause the most good with their resources. Those are the really caring people. Are you one of the really caring people?

Think up a slogan. "Big hearts, big minds" or something. 8/

Once that's established, you can tell them how to do it. 9/

And if you want me to like you, you had better agree with this analysis ;-) 10/10