Using Intellectual Processes to Combat Bias

If you're a prominent intellectual, then a lot of people want your time, and most of them will predictably waste it. You need a filtering/prioritization strategy. This article says that your filtering/prioritization strategy should be public and legible. There are three fairly fundamental problems with this.

The first problem is that most good filtering strategies involve inputs that can't be made legible. If your strategy is to read the first paragraph and predict whether the rest will be worth it, then "predict whether the rest will be worth it" is a complicated black box with internals that are difficult to discuss.

The second problem is that, if you create a legible set of rules which says you'll talk to anyone who's worthy, then any time someone who isn't worth your time tries to talk to you, you're forced to insult them. That will end up, in aggregate, wasting more time than you have.

The third problem is that if you're transparent about how your filters work, then people can optimize for getting through your filters. There's a Goodhart's Law issue where people who optimize for looking like good discussion partners will end up outcompeting people who optimize for being good discussion partners.

As far as your own signaling is concerned: you've made yourself look like someone who's very costly to try and engage with. Specifically, people will anticipate that, if they try to reply to you in a time-efficient manner, you'll be unsatisfied with a short response, demand a longer one, and leverage the lack of a longer response into a social attack. This makes it unsafe to write you even a short response. You've set that up by demanding responses from specific people to your long essays, and then using public channels to try to shame them into giving you an answer.

This seems to be the same person (or a person from the same organization) as users "curi [LW · GW]" and "elliot_temple [LW · GW]". I apologize for using an ad-hominem argument, but there is something about "past behavior is the best predictor of future behavior".

In a nutshell, the debating strategy of this person?/group? is to draw people to an unending debate, until the other side decides it is no longer worth their time, in which case this person?/group? declares a moral victory. For everything you say, they will provide a reply, so the ball is always in your court (and they will publicly document this fact). It's just that after some time you will realize they are neither listening to you, nor updating about anything you said. They just play the game of replying to everything and having the last word in the debate, and then using it as a proof that you are unable to engage in a critical discussion (as apparently is anyone other than themselves).

I take issue with paths forward. It's flawed by an underlying entitlement. People don't owe you an answer. Only from within a premise of agreeing with PF can you demand of other people in the way that you demand answers.

For whatever reason that we cannot know, the person doesn't want to give you an answer. That doesn't mean they are irrational, they are an external actor, blaming an external actor on your internal ability to not get an answer is just externalising your problems and then not solving them. You don't actually create rationality, you wall off irrationality and say "we won't touch that".

I disagree. I believe that if you want to get an answer, do as PET suggests and own the problem. If you have an issue, it's your problem, you are the one responsible for solving the problem. Blaming someone else and not solving the problem yourself just wipes out your own agency.

Is there an alternative solution to the same problems this essay is concerned with, written down somewhere, which you think is successful? If not – if you regard this issue as a major, unsolved problem – why respond to this attempted solution with such initial bias and hostility, instead of making a more serious attempt to figure out how it could work (you looked only for negative things to say, not for positives, and not solutions to the negatives you thought of)? Alternatively, if you regard the whole issue as a non-problem requiring no solution at all, you haven’t explained that.