[microjim]

‘Horrible’ epistemology is a bit harsh (signal = approximate measure rather than direct), but this is indeed a valid loophole. Would be interesting to explore if third or fourth groups that do not benefit from the mutually agreed lie could feasibly counter this.

[guerrilla]

I don't think it's harsh at all. Such an epistemology contributes nothing but false positives because parties with conflicting interests agreeing has zero bearing on truth. It's not even a good rule of thumb; it's just completely useless and irrelevant to knowledge production. Most people believe in some kind of religion or other nonsense like astrology; that they disagree on other things yet agree on that has absolute no bearing on the truth of any of those things. Think about it.

[SilasX]

"Nothing but false positives" is obviously over-the-top. Just yesterday they caught a pretty clear one.

https://news.ycombinator.com/item?id=33472161

[rad88]

It's not an epistemic problem. Twitter has to address their users' toxic behavior turning the whole place/world into a dump. Rather that the confrontation taking place in replies and quote tweets and spreading all over the network, it happens in birdwatch (in theory).

[microjim]

I agree that solid epistemology has better specificity and sensitivity for what's true, and that's the idea with independent fact-checker institutions, but it appears none have managed to gain or retain widespread trust.

An approximate correlation is progress here. If you know of a better heuristic I'd be very interested in hearing it.

Also, can we agree on what problem is most important to be solving right now? Is it 1) we haven't reached complete truth across all dimensions of knowledge or that 2) we need to build bridges crossing the massive societal trust schisms?