What is required for people to reach consensus in public?

public-consensus.md

Public Consensus

What is required for people to reach consensus in public?

To be specific, I'm talking about written statements in public that individuals will attest to.

It seems necessary that participants are sufficiently:

1. grounded in shared reality; a common foundation
2. motivated; inclined to participate
3. comfortable; they feel safe to participate
4. communicative; skilled at interactions
5. intelligent; able to comprehend the ideas being discussed
6. disinterested; without conflicts of interest
7. self-confident; making a mistake or being wrong is ok
8. authentic; willing and able to share their ideas
9. truth-seeking; striving to advance our understanding

Probably Required

It seems likely that correct consensus can only be reached unless people are trained and practiced in:

10. logic
11. critical thinking
12. rationality

Nice to Have

13. calibration: until people have a track record of making predictions, it is usually premature to give their predictions any particular degree of credibility.

Suggestion from a friend:

6. disinterested; without conflicts of interest

This is arguably not necessary.

Another suggestion from a friend:

willingness to compromise

openness to other's ideas

Clarification: by calibration I'm referring to the degree to which one's predicted confidence ranges correspond with the predictor's statistical correctness.

e.g. “One component of the Brier score is calibration, AKA reliability. Calibration refers to the agreement between observed probabilities of outcomes and predictions of probabilities of outcomes.” from https://scicast.org/blog_subdomain/2014/09/10/market-accuracy-and-calibration/

To many eyes, the following three concepts have a very high overlap: (a) logic; (b) critical thinking; (c) rationality -- but to me, after reading The Sequences*, they have very different meanings in terms of people's particular skills.

For example, one doesn't learn about Aumann's Agreement Theorem in (a) or (b), at least not in my experience.

When I think of "logic", I typically think of deductive logic or sometimes other variations, but I rarely think about how Bayes Theorem generalizes over logic and makes it quantitative even under uncertainty. Claude 3.5 Sonnet wrote this about E.T. Jaynes who wrote "Probability Theory: The Logic of Science", and I agree with it completely:

Jaynes demonstrated that Boolean algebra (classical logic) emerges as a special case of probability theory when we restrict ourselves to propositions that are known with certainty (probabilities of 0 or 1). His work showed how the rules of probability theory are not merely ad hoc tools for handling uncertainty, but rather represent the unique consistent extension of classical logic into the domain of partial knowledge.

* There are other places too, not least of which is LessWrong.

I should also mention that many of the tools of probability that modern rationalists take for granted are often less than 50 years old! For some reason, rigorous probability took a shockingly long time to reach us!