The Best Kind of Correct

Technically correct. The Best Kind of Correct. Or is it?

There’s a fundamental tension between allowing speech as valid even when it’s not perfectly precise. Listening with generosity. Steelmanning. And precluding speech that’s ultimately wrong because of its flaws.

You want to allow a lot of speech, but you also need to be able to get to ground. A healthy discourse is full of clarifications. But there is a spectrum of clarification ranging from clarifications which truly are nitpicks and corrections that completely reverse the conclusion.

It’s virtuous to listen generously. It’s also virtuous to be right.

One attempt to name the bind that society is in here is Brandolini’s Law:

“The amount of energy needed to refute bullshit is an order of magnitude bigger than that needed to produce it.”

How do we address this? I’ve felt inspired recently by the ACX post If It's Worth Your Time To Lie, It's Worth My Time To Correct It.

People don’t like nitpickers. “He literally did the WELL AKTUALLY!” If you say Joe Criminal committed ten murders and five rapes, and I object that it was actually only six murders and two rapes, then why am I “defending” Joe Criminal?

Because if it’s worth your time to lie, it’s worth my time to correct it.

If one side lies to make all of their arguments sound 5% stronger, then over long enough it adds up.

My hope comes in the form of some clever mechanism design solution. Think Prediction Markets, Quadratic Voting, or Liquid Democracy. However, I haven’t cracked this particular problem yet.

The rough idea is to allow debaters to "pressure" or force fact checking to get resolved. Kinda like an escalatory double-or-nothing until one side retracts their false claim.

The Slippery Weasel Problem

You know how people are sometimes like slippery weasels? You try to correct them about one thing, but they shift the frame, or reply about something else, or just give you too much material to address. The conversation moves on. This is one reason that long form writing is so valuable. Writing forces people to commit enough to permit a detailed response.

This works great if you have a society of pedants and nitpickers. Several smart people have tried to cultivate a thriving culture with these values and have succeeded to varying degrees.

Unfortunately, you can’t easily convince the weasels that they should become any less slippery. This makes the mechanism design problem a lot harder.

I love trust networks / deferral networks more than most. It seems better to me than deciding on one single source of truth like a trusted judge, or wikipedia, or Snopes.

I want to allow the pluralism of letting people play their own game. You don’t have to find the same things persuasive as me. But I want to be clear that one of the rules of the game that I'm interested in playing is the one where, when your opponent lays out 10 arguments. You get to say, “okay, I'm gonna pick one. And I'm gonna take it apart.” And instead of that leaving me nine points behind, that's going to leave me in the lead with a score of one-to-zero.