I found this to be a much better thought out, better explained, better reasoned and just plain more fun than all but the first few chapters of the actual book. In my mind these are just better examples, and examples that Eliezer understands better and for which he can present better and more accurate evidence.
It’s almost as if the book was written, using sufficiently modest examples and ludicrously charitable assumptions, so Eliezer would feel he has the right to say the things here, the things he actually means to say, and that this is the real point. That would help explain why (at least to me) the middle part of the book felt like it was flailing around and didn’t make sense. It wasn’t trying to. And so, at the right level of viewing, that too serves as an example of the problems at hand.
And of course, even then, this doesn’t go far enough. You don’t just need a hero liscence. You need an occupational liscence. You need a thinking liscence. You need a knowing anything at all liscence. And you need to not be caught not enforcing the liscensing agreements, at an arbitrary meta level, or to be caught being motivated by something other than such enforcement, lest you be smacked down to avoid someone else being caught thus, or even worse, scapegoated. And you don’t know what would be used as evidence or trigger such action, including your failure to identify that there was evidence on which you should have taken such action. Thus, the imitation game and its recursive, backward-chaining justifactions, like modesty arguments that mysteriously always have the same answer written at the bottom of the page before they start.
Or you could realize that the result of such a process seems unlikely to give efficient or reasonable answers in general, decide that’s more important than these hits to your status, and not do that.
Additional predition: it was more fun to write this than the book, and the writing involved an initial long contiguous chunk.
(Where it’s coming from: my enjoyment of reading the above was mostly a little bit of thrill, of the kind I get from watching someone break rules that I always wished I could break. If that makes sense.)
Counterpoint: I got bored reading this about halfway through, and I my guess is that the sort of person who actually needs the concepts is more likely to close the browser indignantly.
I think Zvi’s description of what’s going on sounds accurate, but I think it was actually necessary to write InEq in (roughly) the manner it was written.
((Counter-counter-point, I assume dramatically less time was put into this, and that among other things it wasn’t broken into chapter segments that made it more easily consumable, which is most of the reason I got bored))
There is a need for a certain amount of repetition to make the point, but no question this has some amount of apology for writing a long letter because there wasn’t time to write a shorter one. That extra length probably was worse because you’d already read the book, so you mostly knew where it was going.
I’m sad and somewhat scared that chapter segments make that much difference, but I can believe it, and I’m also happy since we can just use them. Good job being self-aware and pointing it out explicitly. Weird that it might be a big win to literally spend five minutes adding “I”, “II”, “III” etc as line breaks.
It is now the future, and I was rereading the post without remembering that there was ever a time when there were no chapter breaks, and I found the chapter breaks quite reassuring, giving me bits of “ah, I have made some progress!”
Huh, the chapters thing doesn’t feel especially sad to me. Or, it’s sad in the sense that it’d be better if all humans were immune to all disease and could fly and twice as smart, but given the general shape of the world and our brains, it seems reasonable to take into consideration…
1) working memory
2) how much time people have to read a thing at a time
3) expectations about how long a given thing will take
Like, if I start reading a blogpost, I expect it to be blogpost length, and that’s how much time I’m budgeting for it. If it’s longer than blogpost length, then once I run out of time I start thinking “okay, is this worth putting in the effort to remember to read later, and to remember where I was when I left off, or to postpone whatever I was going to do next instead of finishing it?”
Designing a post to accomodate seems quite reasonable to me.
((Again, assuming this is all stuff Eliezer knows and generally takes into account, and that the issue here was that this post was coming out of the “off hours bucket”))
A thing that feels sad to me (in the vein of chapter sections being necessary): the fact that the majority of all long form content that I and anybody I know takes in, is SSC, WaitButWhy, and John Oliver—the only stuff where the hedonic hit rate is at least once per paragraph/breath. I can’t remember the last time I read a book (yes I can, it was 4 months ago when I read InEq in a single sitting. But I really can’t remember before that.).
I get the sense that in the past many people read books, and now I know very few people who read whole books. I remember as a teenager I used to read many non-fiction books. I have maybe read one a year for the past three years.
Quick Googling says that your experience was below average as of a few years ago, with the median person reading six books, 72 percent reading at least one, 60 percent reading at least one fiction and 60 percent reading at least one non-fiction, so one per year is well below the implied mean. So this seems like a ‘local’ problem. I’d definitely call it a problem, and consider myself to be reading far fewer full books than I should versus too much other stuff (and to not be writing enough reviews even of the ones I do read).
Reviews are definitely worthwhile, as are quote blogs like this one. The more books you review and quote, the more books we can add to common knowledge (because the reviews/quotes recommend them to enough people that the contents of the book catch on with them and they use them and recommend them to other people and so on) and the fewer books we have to read (because they either can be distilled to a review/series of quotes or are found to be not worth reading).
As for actually reading, take a book with you on your commute if your commute doesn’t involve driving, and leave another in the bathroom.
(The people who read the most of anyone I know share a certain health issue with Martin Luther, and that’s why they read so much.)
Having it be about providing bookmarks makes me feel a little better, I guess? Although ‘keep the tab open until later even on a phone’ seems perfectly reasonable. And you can use the scroll bar’s size to quickly estimate length and get an alert that it might be long (comments can mess with this a little, but it’s quick to check for rough length when you get worried).
I agree that expecting a 10k word thing to be harder to get through than a 2k thing, and therefore read less, isn’t sad, but that we’re clearly doing too little long-form due to the incentives on the modern internet (and the karma system here makes this worse).
Although ‘keep the tab open until later even on a phone’ seems perfectly reasonable
Dunno. I regularly just close my browser completely.
I’m super pro long-form stuff, but longform stuff still should be optimized for being longform. A 10k blogpost without sections suffers from even being able to find/trust where you left off, even if you’re just looking for comments)
Agreed with Raemon that this was kinda boring. Chapters/sections weren’t part of it for me, either. Just seemed to beat a dead horse a bit, especially after the rest of InEq.
I wouldn’t have bothered with this criticism, except that I find the divided reaction interesting.
I agree with Ray—the chapter was too long and spent too many words saying what it was trying to say. I read it in several sittings due to lack of adequate time block nd couldn’t find my place, which lead to me losing time and rereading portions and feeling generally frustrated. think the impact would be improved by reducing by a considerable margin.
I feel like this was doing a fairly different thing than InEq itself (or at least, the effect it had on me was pretty different)
This post mostly made me think about status dynamics, both within myself, Eliezer, and other people. I think this was useful to think about, but as Eliezer notes somewhere in the middle of this: it feels like this is focusing our attention disproportionately on the wrong half of the equation. (I don’t think we should think zero about the status thing, but it seems like it should be something like a 70⁄30 ratio of “thinking about system behaviors we can change” and “thinking about how status in particular works”.)
Whereas InEq focused my attention on “How do I actually notice when situations are likely to be inadequate? When can I reasonably expect outsized success for my effort?”
I think of this as a call to temporarily focus on that part of the equation so one can realize that this is what most modesty arguments are actually motivated by, and that such arguments are not useful in dealing with the object level so you should ignore them in favor of examining the object level directly. It’s a call to stop paying attention to status arguments (including when they take the form of modesty arguments) so you can see the object level.
Then there’s the mostly seperate question of using status and social dynamics as tools to analyze the failures of systems, which I agree is a useful tool that shouldn’t take up that much of our time.
Hmm. The way this feels to me is the way you once described the phrase “You shouldn’t feel bad about X”, or “If you did X, I wouldn’t blame you.” Technically the phrase is saying not to blame people, but it’s creating the implication that maybe we should blame people.
This feels similarly—since the whole thing focuses on the status frame, it encourages thinking about the status frame even if it’s saying not to.
(I think the net impact of InEq + This Post probably ends up pointing collectively in the right direction, with the right balance, I’m mostly disagreeing with (what I assumed to be) your initial point that this is an all-around better post, at least for the audience that needed to most hear it)
Datapoint: Reading InEq changed my thinking habits when evaluating projects much more than this essay did. This essay mostly reminded me to be surprised at the success of HPMOR, and think about all the useful cognitive updates Pat Modesto would make if he were to fully update on the datapoint of HPMOR.
I found this to be a much better thought out, better explained, better reasoned and just plain more fun than all but the first few chapters of the actual book. In my mind these are just better examples, and examples that Eliezer understands better and for which he can present better and more accurate evidence.
It’s almost as if the book was written, using sufficiently modest examples and ludicrously charitable assumptions, so Eliezer would feel he has the right to say the things here, the things he actually means to say, and that this is the real point. That would help explain why (at least to me) the middle part of the book felt like it was flailing around and didn’t make sense. It wasn’t trying to. And so, at the right level of viewing, that too serves as an example of the problems at hand.
And of course, even then, this doesn’t go far enough. You don’t just need a hero liscence. You need an occupational liscence. You need a thinking liscence. You need a knowing anything at all liscence. And you need to not be caught not enforcing the liscensing agreements, at an arbitrary meta level, or to be caught being motivated by something other than such enforcement, lest you be smacked down to avoid someone else being caught thus, or even worse, scapegoated. And you don’t know what would be used as evidence or trigger such action, including your failure to identify that there was evidence on which you should have taken such action. Thus, the imitation game and its recursive, backward-chaining justifactions, like modesty arguments that mysteriously always have the same answer written at the bottom of the page before they start.
Or you could realize that the result of such a process seems unlikely to give efficient or reasonable answers in general, decide that’s more important than these hits to your status, and not do that.
Zvi’s probably right.
Additional predition: it was more fun to write this than the book, and the writing involved an initial long contiguous chunk.
(Where it’s coming from: my enjoyment of reading the above was mostly a little bit of thrill, of the kind I get from watching someone break rules that I always wished I could break. If that makes sense.)
Counterpoint: I got bored reading this about halfway through, and I my guess is that the sort of person who actually needs the concepts is more likely to close the browser indignantly.
I think Zvi’s description of what’s going on sounds accurate, but I think it was actually necessary to write InEq in (roughly) the manner it was written.
((Counter-counter-point, I assume dramatically less time was put into this, and that among other things it wasn’t broken into chapter segments that made it more easily consumable, which is most of the reason
I got bored))
There is a need for a certain amount of repetition to make the point, but no question this has some amount of apology for writing a long letter because there wasn’t time to write a shorter one. That extra length probably was worse because you’d already read the book, so you mostly knew where it was going.
I’m sad and somewhat scared that chapter segments make that much difference, but I can believe it, and I’m also happy since we can just use them. Good job being self-aware and pointing it out explicitly. Weird that it might be a big win to literally spend five minutes adding “I”, “II”, “III” etc as line breaks.
Yeah, this is cheap to try. There are a few section divisions in the dialogue now.
It is now the future, and I was rereading the post without remembering that there was ever a time when there were no chapter breaks, and I found the chapter breaks quite reassuring, giving me bits of “ah, I have made some progress!”
Huh, the chapters thing doesn’t feel especially sad to me. Or, it’s sad in the sense that it’d be better if all humans were immune to all disease and could fly and twice as smart, but given the general shape of the world and our brains, it seems reasonable to take into consideration…
1) working memory
2) how much time people have to read a thing at a time
3) expectations about how long a given thing will take
Like, if I start reading a blogpost, I expect it to be blogpost length, and that’s how much time I’m budgeting for it. If it’s longer than blogpost length, then once I run out of time I start thinking “okay, is this worth putting in the effort to remember to read later, and to remember where I was when I left off, or to postpone whatever I was going to do next instead of finishing it?”
Designing a post to accomodate seems quite reasonable to me.
(See also Scott’s post on writing nonfiction)
http://slatestarcodex.com/2016/02/20/writing-advice/
((Again, assuming this is all stuff Eliezer knows and generally takes into account, and that the issue here was that this post was coming out of the “off hours bucket”))
A thing that feels sad to me (in the vein of chapter sections being necessary): the fact that the majority of all long form content that I and anybody I know takes in, is SSC, WaitButWhy, and John Oliver—the only stuff where the hedonic hit rate is at least once per paragraph/breath. I can’t remember the last time I read a book (yes I can, it was 4 months ago when I read InEq in a single sitting. But I really can’t remember before that.).
I get the sense that in the past many people read books, and now I know very few people who read whole books. I remember as a teenager I used to read many non-fiction books. I have maybe read one a year for the past three years.
Quick Googling says that your experience was below average as of a few years ago, with the median person reading six books, 72 percent reading at least one, 60 percent reading at least one fiction and 60 percent reading at least one non-fiction, so one per year is well below the implied mean. So this seems like a ‘local’ problem. I’d definitely call it a problem, and consider myself to be reading far fewer full books than I should versus too much other stuff (and to not be writing enough reviews even of the ones I do read).
Reviews are definitely worthwhile, as are quote blogs like this one. The more books you review and quote, the more books we can add to common knowledge (because the reviews/quotes recommend them to enough people that the contents of the book catch on with them and they use them and recommend them to other people and so on) and the fewer books we have to read (because they either can be distilled to a review/series of quotes or are found to be not worth reading).
As for actually reading, take a book with you on your commute if your commute doesn’t involve driving, and leave another in the bathroom.
(The people who read the most of anyone I know share a certain health issue with Martin Luther, and that’s why they read so much.)
Having it be about providing bookmarks makes me feel a little better, I guess? Although ‘keep the tab open until later even on a phone’ seems perfectly reasonable. And you can use the scroll bar’s size to quickly estimate length and get an alert that it might be long (comments can mess with this a little, but it’s quick to check for rough length when you get worried).
I agree that expecting a 10k word thing to be harder to get through than a 2k thing, and therefore read less, isn’t sad, but that we’re clearly doing too little long-form due to the incentives on the modern internet (and the karma system here makes this worse).
Dunno. I regularly just close my browser completely.
I’m super pro long-form stuff, but longform stuff still should be optimized for being longform. A 10k blogpost without sections suffers from even being able to find/trust where you left off, even if you’re just looking for comments)
Agreed with Raemon that this was kinda boring. Chapters/sections weren’t part of it for me, either. Just seemed to beat a dead horse a bit, especially after the rest of InEq.
I wouldn’t have bothered with this criticism, except that I find the divided reaction interesting.
Seconding this. I was unable to come even close to finishing this chapter, I’m afraid.
I honestly felt like it got better as the post went on. The middle was the most boring part, though, even as someone who enjoted it.
I agree with Ray—the chapter was too long and spent too many words saying what it was trying to say. I read it in several sittings due to lack of adequate time block nd couldn’t find my place, which lead to me losing time and rereading portions and feeling generally frustrated. think the impact would be improved by reducing by a considerable margin.
Having actually read the whole thing now:
I feel like this was doing a fairly different thing than InEq itself (or at least, the effect it had on me was pretty different)
This post mostly made me think about status dynamics, both within myself, Eliezer, and other people. I think this was useful to think about, but as Eliezer notes somewhere in the middle of this: it feels like this is focusing our attention disproportionately on the wrong half of the equation. (I don’t think we should think zero about the status thing, but it seems like it should be something like a 70⁄30 ratio of “thinking about system behaviors we can change” and “thinking about how status in particular works”.)
Whereas InEq focused my attention on “How do I actually notice when situations are likely to be inadequate? When can I reasonably expect outsized success for my effort?”
I think of this as a call to temporarily focus on that part of the equation so one can realize that this is what most modesty arguments are actually motivated by, and that such arguments are not useful in dealing with the object level so you should ignore them in favor of examining the object level directly. It’s a call to stop paying attention to status arguments (including when they take the form of modesty arguments) so you can see the object level.
Then there’s the mostly seperate question of using status and social dynamics as tools to analyze the failures of systems, which I agree is a useful tool that shouldn’t take up that much of our time.
Hmm. The way this feels to me is the way you once described the phrase “You shouldn’t feel bad about X”, or “If you did X, I wouldn’t blame you.” Technically the phrase is saying not to blame people, but it’s creating the implication that maybe we should blame people.
This feels similarly—since the whole thing focuses on the status frame, it encourages thinking about the status frame even if it’s saying not to.
(I think the net impact of InEq + This Post probably ends up pointing collectively in the right direction, with the right balance, I’m mostly disagreeing with (what I assumed to be) your initial point that this is an all-around better post, at least for the audience that needed to most hear it)
Datapoint: Reading InEq changed my thinking habits when evaluating projects much more than this essay did. This essay mostly reminded me to be surprised at the success of HPMOR, and think about all the useful cognitive updates Pat Modesto would make if he were to fully update on the datapoint of HPMOR.