sd — LessWrong

LESSWRONG
LW

sd — LessWrong

Sorry I can't tell if you meant to link to actual articles or just the definitions. If it is just the definitions, I can't tell how they relate to the statements that follow the links.

Overall unclear what you mean by "incremental move towards rationality is not always ideal" and how it relates to the post.

This is a bizarre comment. Aren't most examples hypothetical? When you start a math question with "Jack buys 3 melons", does Jack need to be a real guy who actually bought 3 melons?...

The true question is whether the example is consistent with real empirical evidence. It appears to me perfectly possible and in fact anecdotally true that two people might have very similar beliefs with the exception of a small subset - which has little decision value but strong intrinsic value, in favour of the apparently less accurate beliefs. Which follows that it seems reasonable and realistic that holding a small and non-decision relevant set of "false" beliefs might turn out to be beneficial.

Although as I mentioned, the degree to which one can make this a conscious strategy is very much arguable.

Yes indeed.

There is still - almost always - an (arbitrarily) small chance that even "ridiculous" statements turn out to be true.

E.g., maybe the Earth IS flat, and we have just been brainwashed by an advanced alien race to perceive physics such that it appears spherical. Very very very very unlikely, but not impossible.

Note that from Bayes' Theorem, if your prior is either 0 or 1, then no amount of evidence can change it. Which for a rationalist is a horrible position to be in, since it basically means that you have stopped learning and your beliefs are now disentangled from reality.

Thanks for the feedback :)

The twelve virtues are already linked in the article, but I hadn't come across the Basics of Rationalist Discourse yet. It is indeed very interesting and aligned with the objective of this essay. I very much intend to incorporate its content where applicable and give it credit (I will link it at the beginning of the post).

I wonder if the Basics of Rationalist Discourse doesn't represent a more detailed and tactical view of the topic, as opposed to the more generic top-of-mind principles that this essay tried to cover. The former is a manual for controlling specific behaviours, the latter a memento of the general mindset to hold... (read more)

My argument was never that abduction is a subset of induction, but that it can always be replaced by a combination of deduction and induction.

The effect of simplicity and consistency on probabilities can both be classified as deductions:

the former as a correct application of probabilistic logic (every additional assumption reduces the probability of the conclusion being true due to probability product)
the latter as a correct application of Bayes' theorem or other types of logic (when [belief A] implies [not belief B])

The underlying issue is that what we are trying to do with abduction is find the hidden mechanism behind the directly observable, the force of gravity that makes the apple fall. Since

... (read more)

I am not sure what your definition of Bayesian reasoning is, but I personally think of:

Deduction as somewhat equivalent to mathematical logic (including probabilistic logic and other sub-types of logic).
Induction as somewhat equivalent to inferential statistics.

I am not sure where most people would place Bayes' theorem itself, but Bayesian statistics is usually considered a subfield of (or rather an approach to) inferential statistics.

Apologies, I am not sure I understood.

Is it that:
- Induction/Deduction fit data to existing models
- Abduction is about proposing new models?

I think "best" is a bit of a generic term, but in regard to beliefs, it seems safe to say that the "quality" of a belief should be judged exclusively on its probability of being true.

Simplicity and consistency are then factors that positively affect the probability of a belief being true, but they are not in themselves determinants of the "best" explanation.

Given:
- Belief A, which is 51% likely to be true but very complex and somewhat discordant with existing beliefs (this is included in the probability estimation)
- Belief B, which is 49% likely to be true but very simple and elegant

Assuming the probabilities are correct, I think Belief A should be considered the "best". Would you agree with this?

P.S. Quantum field theory is an example of a very complex theory and inconsistent with other accepted theories (e.g. General Relativity), but still the "best" explanation for empirical evidence.