Sometimes you need someone to give the naive view, but doing so hurts the reputation of the person stating it.
For example suppose X is the naive view and Y is a more sophisticated view of the same subject. For sake of argument suppose X is correct and contradicts Y.
Given 6 people, maybe 1 of them starts off believing Y. 2 people are uncertain, and 3 people think X. In the world where people have their usernames attached. The 3 people who believe X now have a coordination problem. They each face a local disincentive to state the case for X, although they definitely want _someone_ to say it. The equilibrium here is that no one makes the case for X and the two uncertain people get persuaded to view Y.
However if someone is anonymous and doesn’t care that much about their reputation, they may just go ahead and state the case for X, providing much better information to the undecided people.
This makes me happy there are some smart people posting under pseudonyms. I claim it is a positive factor for the epistemics of LessWrong.
Sometimes you need someone to give the naive view, but doing so hurts the reputation of the person stating it.
For example suppose X is the naive view and Y is a more sophisticated view of the same subject. For sake of argument suppose X is correct and contradicts Y.
Given 6 people, maybe 1 of them starts off believing Y. 2 people are uncertain, and 3 people think X. In the world where people have their usernames attached. The 3 people who believe X now have a coordination problem. They each face a local disincentive to state the case for X, although they definitely want _someone_ to say it. The equilibrium here is that no one makes the case for X and the two uncertain people get persuaded to view Y.
However if someone is anonymous and doesn’t care that much about their reputation, they may just go ahead and state the case for X, providing much better information to the undecided people.
This makes me happy there are some smart people posting under pseudonyms. I claim it is a positive factor for the epistemics of LessWrong.