One thing I've been doing is limiting my 'true discussions'¹ to ones where I can reasonably say that both parties are acting in the specific discussion like the

One thing I've been doing is limiting my 'true discussions'¹ to ones where I can reasonably say that both parties are acting in the specific discussion like they would approximate the conditions required for Aumann's agreement theorem². What follows is that every disagreement should be followed by new information. At the point where disagreements are close to "agree to disagree", I know that at least one of the two participants (either me or the other person) is not participating in an honest rational manner allowing me to note that the conversation should not be a means to inform myself. This is not an indictment of character. It's merely the nature of our minds that we are not always capable of usefully transmitting information to others. Repeating the discussion with a sufficient number of people will allow me to consider whether I'm the poor partner or whether it's the other people. Of course, the practice of this is somewhat removed from the description above, but I believe it's improved my ability to truth-find (judging by the degree to which my new opinions match reality) to even attempt this approach, even in a flawed manner. I _am_ doing better on the market now, but naturally it's impossible to tell if that's because I'm more well-informed or if that's luck. ¹ discussions where I'm attempting to actually expand my knowledge, as opposed to ones I'm participating in for social reasons ² people familiar with this will argue that you can't find people with the same priors as you, but is this really true? Shouldn't all rational agents arrive fully formed into the world void of any knowledge? If so, all you have to accept is fully conditioning everyone's statements on their entire experience from birth. Fortunately, this shouldn't be terribly hard if you both otherwise meet these conditions: http://www.scottaaronson.com/papers/agree-econ.pdf http://www.scottaaronson.com/papers/agree-econ.pdf