“You’re too stupid and self-deceiving to just use Solomonoff induction” ~ “If you were less stupid and self deceiving you’d be able to just use Solomonoff induction” + “but since you are in fact stupid and self-deceiving, instead you have to use the less elegant approximation Science”
Actually, yes, because of the misleading signals in the inept writing. But thank you for clarifying.
Conclusion: The argument in written in a crazy fashion, but it really is merely stupid. There is no possible measure of simplicity that isn’t language relative. How could there be?
You seem to be confusing “language relative” with “non-mathematical.” Kolmogorov Complexity is “language-relative,” if I’m understanding you right; specifically, it’s relative (if I’m using the terminology right?) to a Turing Machine. This was not relevant to Eliezer’s point, so it was not addressed.
(Incidentally, this is a perfect example of you “hold{ing} views contrary to scientific consensus in specialized fields where {you} lack expert knowledge based on pure ratiocination,” since Kolmogorov Complexity is “one of the fundamental concepts of theoretical computer science”, you seemingly lack expert knowledge since you don’t recognize these terms, and your argument seems to be based on pure ratiocination.)
When I read that line for the first time, I understood it. Between our two cases, the writing was the same, but the reader was different. Thus, the writing cannot be the sole cause of our different outcomes.
Well, if a substantial fraction of readers read something differently or can’t parse it, it does potentially reflect a problem with the writing even if some of the readers, or even most readers, do read it correctly.
Absolutely. I intended to convey that if you don’t understand something, that the writing is misleading and inept is not the only possible reason. srdiamond is speaking with such confidence that I felt safe tabling further subtleties for now.
I can’t tell which way your sarcasm was supposed to cut.
The obvious interpretation is that you think rationality is somehow hindered by paying attention to form rather than substance, and the “exemplary rationality” was intended to be mocking.
But your comment being referenced was an argument that form has something very relevant to say about substance, so it could also be that you were actually praising gwern for practicing what you preach.
I read your three-part series. Your posts did not substantiate the claim “good thinking requires good writing.” Your second post slightly increased my belief in the converse claim, “good thinkers are better-than-average writers,” but because the only evidence you provided was a handful of historical examples, it’s not very strong evidence. And given how large the population of good thinkers, good writers, bad thinkers, and bad writers is relative to your sample, evidence for “good thinking implies good writing” is barely worth registering as evidence for “good writing implies good thinking.”
“You’re too stupid and self-deceiving to just use Solomonoff induction” ~ “If you were less stupid and self deceiving you’d be able to just use Solomonoff induction” + “but since you are in fact stupid and self-deceiving, instead you have to use the less elegant approximation Science”
That was hard to find out?
Actually, yes, because of the misleading signals in the inept writing. But thank you for clarifying.
Conclusion: The argument in written in a crazy fashion, but it really is merely stupid. There is no possible measure of simplicity that isn’t language relative. How could there be?
You seem to be confusing “language relative” with “non-mathematical.” Kolmogorov Complexity is “language-relative,” if I’m understanding you right; specifically, it’s relative (if I’m using the terminology right?) to a Turing Machine. This was not relevant to Eliezer’s point, so it was not addressed.
(Incidentally, this is a perfect example of you “hold{ing} views contrary to scientific consensus in specialized fields where {you} lack expert knowledge based on pure ratiocination,” since Kolmogorov Complexity is “one of the fundamental concepts of theoretical computer science”, you seemingly lack expert knowledge since you don’t recognize these terms, and your argument seems to be based on pure ratiocination.)
When I read that line for the first time, I understood it. Between our two cases, the writing was the same, but the reader was different. Thus, the writing cannot be the sole cause of our different outcomes.
Well, if a substantial fraction of readers read something differently or can’t parse it, it does potentially reflect a problem with the writing even if some of the readers, or even most readers, do read it correctly.
Absolutely. I intended to convey that if you don’t understand something, that the writing is misleading and inept is not the only possible reason. srdiamond is speaking with such confidence that I felt safe tabling further subtleties for now.
The philosophizing of inept, verbose writers like Yudkowsky can be safely dismissed based solely on their incompetence as writers. For a succinct defense of this contention, see my “Can bad writers be good thinkers? Part 1 of THE UNITY OF LANGUAGE AND THOUGHT” OR see the 3-part “Writing & Thought series” — all together, fewer than 3,000 words.
I believe what you wrote because you used so much bolding.
Way to deflect attention from substance to form. Exemplary rationality!
I can’t tell which way your sarcasm was supposed to cut.
The obvious interpretation is that you think rationality is somehow hindered by paying attention to form rather than substance, and the “exemplary rationality” was intended to be mocking.
But your comment being referenced was an argument that form has something very relevant to say about substance, so it could also be that you were actually praising gwern for practicing what you preach.
I choose to interpret it as praise, and receive a warm fuzzy feeling.
I read your three-part series. Your posts did not substantiate the claim “good thinking requires good writing.” Your second post slightly increased my belief in the converse claim, “good thinkers are better-than-average writers,” but because the only evidence you provided was a handful of historical examples, it’s not very strong evidence. And given how large the population of good thinkers, good writers, bad thinkers, and bad writers is relative to your sample, evidence for “good thinking implies good writing” is barely worth registering as evidence for “good writing implies good thinking.”