Well, I am not into the AI stuff, so maybe my take on it will be far more ameboid than what you had in mind. I rather view the broadly-LW corpus as something like TV Tropes and try to find the “individual articles” (not always, or even consciously) in the things I interact with. Maybe other people do it, too. And if I have to interact with something regularly, like for work, it’s far easier to bear when it’s fun.
It’s like, “I will have to sell this book, which means I gotta be ready to say something about it” and “But this translation of “Wyrd Sisters” flattens witches’ individuality—smoothes out prickliness, adds a veneer of experience to something which can only be awkward bumbling, turns a salacious remark into a good-natured explanation, etc. The text is readable and well-built, but has denotational and connotational issues. Through what lense did the translator view the original to arrive at this version?” So “Maybe he decided to translate from the bird’s eye view, where the text as a whole must have internal logic and structure, but not from the characters’ views, where what they think/do is what matters and actually exists at all” and then “But characters’ agency is important, in a meta-text about theatre” which leads to “Who of our translators does consistently preserve 1) author’s intonation, 2) author’s view of characters, 3) general readability, 4) characters’ view of the world, 5) characters’ view of themselves? And which publishing house can be counted on to have an editor who gives a damn?”
Which gives me, in the end, two lists—one longer, of quality translators, and one short—of publishing houses. Both rather subjective, but they will do in a pinch.
I’m not sure if this is what you had in mind when you asked your question, or even if this can be called “intellectual activity”, it just feels like curiosity. But that’s my answer.
I think a primary, maybe the primary, effect that the sequences have on a reader’s thinking is through this kind of pattern-matching. It is shallow, as rationality techniques go, but it can have a large effect nonetheless. It’s like the only rationality technique you have is TAPs, and you only set up taps of the form “resemblance to rationality concept” → “think of rationality concept”. But, those taps can still be quite useful, since thinking of a relevant concept may lead to something.
Concerning the rest—not sure what to comment on, but it is a datapoint.
Hmm, this also reminds me of the thing that we discussed at our last meetup (there were two of us, so I wouldn’t say it was about Solstice, exactly): how history of plant morphology is a handle for some fairly unrelated fields (social sciences), if you grok it. Not the best possible one, not capable of turning around many axes, and not easy to turn, but—a handle.
The lj post that started it described a monograph on the history of research of inflorescence structure. I won’t link it, it’s in Russian anyway. It talked how plants are systems of “little integrity”—when you look at them, you see they have only a few easy-to-recognize building blocks, but the blocks themselves are very different case to case. And if one wants to build a system of kinship between cases, suddenly the space of block names explodes.
(The other person present at the meetup actually could not cope with it, he tried to imagine “a world without morphology”, an evolution culling out the diversifying misfits—it was freakin’ fascinating to watch. I saw the abyss between his engineering background and my observational one, *and I actually think now that “curiosity” means different things to the two of us*.)
It turns out that what people had used to describe as a block, falls apart into several neat categories that only superficially resemble each other, due to convergent evolution of plants or microscopy milestones or *something*. History of morphology reconstructs the gradual focusing of thought on how plants are really built on the inside and outside, and how we kind of feel where current names don’t fit already. According to that lj post, social sciences have yet to reach this point, but clearly they, too, deal with systems of little integrity, where one has to invent names for the many real, and not the visible blocks. The lj post advised social scientists to read up on the already covered grounds of plant morphology, to gauge the depth of what they would have to do (my paraphrase).
And this is how I think about “rationality techniques”, too; that they are going to fall apart into different clusters, and the engineering-inclined people would want to try to glue them back. Intellectual activity, if it employs specific techniques, should be able to destruct-test them, to arrive at new and better blocks, which is easier to do if I am not at the same moment building something bigger with the old ones.
Well, for techniques to be more than TAPs, they have to kind of branch, don’t they? In which case there has to be an (internally natural) hierarchy of concepts, which I am afraid to build, because for me “rationality techniques” as presented here are phenomenological observations. Or stop-signs.
-------------
I don’t like the concept of “fully general counter-argument”, for example, and I try to make do without it. If we have “fully general counter-arguments”, then we have “fully general supporting arguments” and “fully general misses” etc. I always try to treat someone’s counter-argument as not fully-general unless we both understand it so; because for some reason they view it as the thing to say. It might be an irrelevant reason, but very many are, and the world keeps spinning.
Curiosity is just that—if you are asking whether being told about rationality helps develop curiosity under some conditions, then maybe we shouldn’t talk about all “rationality techniques”, because they as a whole are not aimed at developing curiosity. Choose some.
Well, I am not into the AI stuff, so maybe my take on it will be far more ameboid than what you had in mind. I rather view the broadly-LW corpus as something like TV Tropes and try to find the “individual articles” (not always, or even consciously) in the things I interact with. Maybe other people do it, too. And if I have to interact with something regularly, like for work, it’s far easier to bear when it’s fun.
It’s like, “I will have to sell this book, which means I gotta be ready to say something about it” and “But this translation of “Wyrd Sisters” flattens witches’ individuality—smoothes out prickliness, adds a veneer of experience to something which can only be awkward bumbling, turns a salacious remark into a good-natured explanation, etc. The text is readable and well-built, but has denotational and connotational issues. Through what lense did the translator view the original to arrive at this version?” So “Maybe he decided to translate from the bird’s eye view, where the text as a whole must have internal logic and structure, but not from the characters’ views, where what they think/do is what matters and actually exists at all” and then “But characters’ agency is important, in a meta-text about theatre” which leads to “Who of our translators does consistently preserve 1) author’s intonation, 2) author’s view of characters, 3) general readability, 4) characters’ view of the world, 5) characters’ view of themselves? And which publishing house can be counted on to have an editor who gives a damn?”
Which gives me, in the end, two lists—one longer, of quality translators, and one short—of publishing houses. Both rather subjective, but they will do in a pinch.
I’m not sure if this is what you had in mind when you asked your question, or even if this can be called “intellectual activity”, it just feels like curiosity. But that’s my answer.
Concerning TV Tropes --
I think a primary, maybe the primary, effect that the sequences have on a reader’s thinking is through this kind of pattern-matching. It is shallow, as rationality techniques go, but it can have a large effect nonetheless. It’s like the only rationality technique you have is TAPs, and you only set up taps of the form “resemblance to rationality concept” → “think of rationality concept”. But, those taps can still be quite useful, since thinking of a relevant concept may lead to something.
Concerning the rest—not sure what to comment on, but it is a datapoint.
Hmm, this also reminds me of the thing that we discussed at our last meetup (there were two of us, so I wouldn’t say it was about Solstice, exactly): how history of plant morphology is a handle for some fairly unrelated fields (social sciences), if you grok it. Not the best possible one, not capable of turning around many axes, and not easy to turn, but—a handle.
The lj post that started it described a monograph on the history of research of inflorescence structure. I won’t link it, it’s in Russian anyway. It talked how plants are systems of “little integrity”—when you look at them, you see they have only a few easy-to-recognize building blocks, but the blocks themselves are very different case to case. And if one wants to build a system of kinship between cases, suddenly the space of block names explodes.
(The other person present at the meetup actually could not cope with it, he tried to imagine “a world without morphology”, an evolution culling out the diversifying misfits—it was freakin’ fascinating to watch. I saw the abyss between his engineering background and my observational one, *and I actually think now that “curiosity” means different things to the two of us*.)
It turns out that what people had used to describe as a block, falls apart into several neat categories that only superficially resemble each other, due to convergent evolution of plants or microscopy milestones or *something*. History of morphology reconstructs the gradual focusing of thought on how plants are really built on the inside and outside, and how we kind of feel where current names don’t fit already. According to that lj post, social sciences have yet to reach this point, but clearly they, too, deal with systems of little integrity, where one has to invent names for the many real, and not the visible blocks. The lj post advised social scientists to read up on the already covered grounds of plant morphology, to gauge the depth of what they would have to do (my paraphrase).
And this is how I think about “rationality techniques”, too; that they are going to fall apart into different clusters, and the engineering-inclined people would want to try to glue them back. Intellectual activity, if it employs specific techniques, should be able to destruct-test them, to arrive at new and better blocks, which is easier to do if I am not at the same moment building something bigger with the old ones.
Well, for techniques to be more than TAPs, they have to kind of branch, don’t they? In which case there has to be an (internally natural) hierarchy of concepts, which I am afraid to build, because for me “rationality techniques” as presented here are phenomenological observations. Or stop-signs.
-------------
I don’t like the concept of “fully general counter-argument”, for example, and I try to make do without it. If we have “fully general counter-arguments”, then we have “fully general supporting arguments” and “fully general misses” etc. I always try to treat someone’s counter-argument as not fully-general unless we both understand it so; because for some reason they view it as the thing to say. It might be an irrelevant reason, but very many are, and the world keeps spinning.
Curiosity is just that—if you are asking whether being told about rationality helps develop curiosity under some conditions, then maybe we shouldn’t talk about all “rationality techniques”, because they as a whole are not aimed at developing curiosity. Choose some.