TODO describe problem (people disagree about important things, don’t change their minds often, aren’t good at crossing inferential distances e.g. see The Hanson-Yudkowsky AI-Foom Debate)

May be a good complement to Forecasting methods so that forecasters can update.

Check out argument mapping software.

See Wikipedia page on argument mapping

Debate tools

Ought is a new company with a team of EA or EA-adjacent people.

“Research on the underlying models and methods he uses to resolve disagreements between individuals with large inferential gaps. He will develop lessons to transfer these to test students, and see what skills are transferred and what skills still lack.”¹ I haven’t seen anything public about this so not sure what’s going on.

List people who are interested in this?

See also

• Better explanations

External links

• Wei Dai has some ideas