Ben Pace comments on Disagreement Comes From the Dark World

[Ben Pace]

Most real-world disagreements of interest don’t look like the bus arrival or math problem examples—qualitatively, not as a matter of trying to prove quantitatively harder theorems. Real-world disagreements tend to persist; they’re predictable—in flagrant contradiction of how the beliefs of Bayesian reasoners would follow a random walk. From this we can infer that typical human disagreements aren’t “honest”, in the sense that at least one of the participants is behaving as if they have some other goal than getting to the truth.

Importantly, this characterization of dishonesty is using a functionalist criterion: when I say that people are behaving as if they have some other goal than getting to the truth, that need not imply that anyone is consciously lying; “mere” bias is sufficient to carry the argument.

Dishonest disagreements end up looking like conflicts because they are disguised conflicts.

I like a lot of the content in this post but this is where I get off the bus. I do not believe that Aumann’s Agreement Theorem shows that all disagreements are dishonest; this is because Aumann’s Agreement Theorem shows an end state and is not an argument that coming to consensus is an efficient algorithm.

I only know of one paper attempting to argue that Aumann Agreement is efficient, and I don’t believe it succeeds. It argues that you only need to communicate a relatively low number of bits between two agents, but the computational limit it puts on how much thinking you have to do is utterly impractical—to get a 50% chance at arriving within 50 percentile points of one anotherm, you must execute $2^{36,864}$ subroutine calls, which is not remotely efficient.

So while I believe Aumann Agreement applies to unbounded agents, I don’t believe it has practical implications for bounded agents like us.