This is a particular kind of error I think needs more attention but am having trouble articulating. https://twitter.com/LinchZhang/status/1290526701848420358

Another example would be "we need to get people in a habit of donating now, and we'll teach them about effectiveness later." There are circumstances under which that works, but I do not like it as a plan.

It's in the neighborhood of cargo culting. Step 1 is generate something that resembles what you want, and step 2 is introducing the key thing that actually makes it good.

Or people who emphasize Just Vote over figuring out who to vote for.

My classic example is Plague Inc, in which you add features to a plague over time (it was more fun in 2019). When you add a feature, every infection in the world gets it at the same time.

So obviously you crank up infectivity at first, and only once you've infected literally everyone do you crank up lethality so you can achieve the win condition of "everyone dies" (again, more fun in 2019).

Another example: https://theunitofcaring.tumblr.com/post/160988890191/a-couple-weeks-ago-i-went-to-a-debate-at-berkeley. "We have to do useless, costly things to build our capacity to do useful things later. This is tough because you do typically have to start out learning on less than the most useful thing.

But this trains/selects for people willing to violate general norms to keep subculture norms, and that feels really creepy.