I’m pretty sure it is the same post I wrote my comment in response to.
It is hard for someone like Pressman to appreciate the challenges in engaging with average people
And similar comments seem like a level of snark and condescension that the other descriptions don’t have.
It’s also wrong, I do appreciate the challenges with that and think they are in a sense the fundamental difficulty of writing a book like this. What I was originally going to say in my review , before deciding it was going to get lambasted by 50 parties anyway and I didn’t need to be one of them when there’s praiseworthy elements I can focus on instead, is that the book as written is barely structured like an argument at all. It’s more like a series of disconnected vignettes arranged in roughly chronological order. This is because the authors don’t actually feel they can make an argument. They don’t feel they can just make an argument because normally when they try doing that their audience will interject with some ridiculous cope thing, and the ridiculous cope is different for different people and it’s difficult to know in advance which ridiculous cope thing the audience will want you to respond to. So what the book does is interrupt itself constantly to try and head off some aside or point that the imagined reader might make there, and I guess I can only hope that the chosen structure is actually the product of constant beta readers objecting at various points and then deciding to put the interruptions at the most common points of objection. I hope this but do not really expect it because the book honestly comes across as the product of an incestuous editing process where it wasn’t really shown to any critical perspectives who would point out flaws as basic as “if I skim the beginning of each part of the book I will pick up on it always starting with an example sufficiently whimsical and not-real that the phrase ‘once upon a time’ is warranted”.
Nevertheless my review was primarily written from the perspective of “You are an AI (alignment) researcher and you want to know if there’s anything important in this book for you to spend your limited time reading.” so I didn’t go into that much detail about the style flaws besides noting that they exist and what I feel the biggest flaws are. Part of why I didn’t want to do an extended takedown of the style is that ultimately it’s an empirical question, the book will either be successful or it won’t be and going into a long dissection of what I think is wrong stylistically in a book that isn’t even written with me as the intended audience doesn’t seem like a very self aware thing to do.
They don’t feel they can just make an argument because normally when they try doing that their audience will interject with some ridiculous cope thing, and the ridiculous cope is different for different people and it’s difficult to know in advance which ridiculous cope thing the audience will want you to respond to. So what the book does is interrupt itself constantly to try and head off some aside or point that the imagined reader might make there
I dont see why that’s a bad thing, or fundamentally different to making an argument.: making a case and refuting objections as you go along, is a stronger form of argumentation than just making a case.
Well it’s a bad thing because it makes it harder to follow what they’re trying to say. It’s also a necessary thing for this subject and kind of the core question that the book has to answer is “how do we anticipate people’s objections and get past them?” and the book represents one hypothesis for how to do that. I am doubtful that it is a correct hypothesis, the reception to it doesn’t seem very positive BUT the reviews we’re getting are not from the target audience and as Scott Alexander said in his review Yudkowsky is a genius who has a history of seeing memetic opportunities that other people do not. So I feel comfortable noting I am doubtful and otherwise letting the reaction tell the story.
Well it’s a bad thing because it makes it harder to follow what they’re trying to say
If you don’t dont, you end up with in a position where the audience can read a short book, and know what your claim is, and disagree with it because of one of the unanswered objections.
and the book represents one hypothesis for how to do that. I am doubtful that it is a correct hypothesis
I’m pretty sure it is the same post I wrote my comment in response to.
And similar comments seem like a level of snark and condescension that the other descriptions don’t have.
It’s also wrong, I do appreciate the challenges with that and think they are in a sense the fundamental difficulty of writing a book like this. What I was originally going to say in my review , before deciding it was going to get lambasted by 50 parties anyway and I didn’t need to be one of them when there’s praiseworthy elements I can focus on instead, is that the book as written is barely structured like an argument at all. It’s more like a series of disconnected vignettes arranged in roughly chronological order. This is because the authors don’t actually feel they can make an argument. They don’t feel they can just make an argument because normally when they try doing that their audience will interject with some ridiculous cope thing, and the ridiculous cope is different for different people and it’s difficult to know in advance which ridiculous cope thing the audience will want you to respond to. So what the book does is interrupt itself constantly to try and head off some aside or point that the imagined reader might make there, and I guess I can only hope that the chosen structure is actually the product of constant beta readers objecting at various points and then deciding to put the interruptions at the most common points of objection. I hope this but do not really expect it because the book honestly comes across as the product of an incestuous editing process where it wasn’t really shown to any critical perspectives who would point out flaws as basic as “if I skim the beginning of each part of the book I will pick up on it always starting with an example sufficiently whimsical and not-real that the phrase ‘once upon a time’ is warranted”.
Nevertheless my review was primarily written from the perspective of “You are an AI (alignment) researcher and you want to know if there’s anything important in this book for you to spend your limited time reading.” so I didn’t go into that much detail about the style flaws besides noting that they exist and what I feel the biggest flaws are. Part of why I didn’t want to do an extended takedown of the style is that ultimately it’s an empirical question, the book will either be successful or it won’t be and going into a long dissection of what I think is wrong stylistically in a book that isn’t even written with me as the intended audience doesn’t seem like a very self aware thing to do.
I dont see why that’s a bad thing, or fundamentally different to making an argument.: making a case and refuting objections as you go along, is a stronger form of argumentation than just making a case.
Well it’s a bad thing because it makes it harder to follow what they’re trying to say. It’s also a necessary thing for this subject and kind of the core question that the book has to answer is “how do we anticipate people’s objections and get past them?” and the book represents one hypothesis for how to do that. I am doubtful that it is a correct hypothesis, the reception to it doesn’t seem very positive BUT the reviews we’re getting are not from the target audience and as Scott Alexander said in his review Yudkowsky is a genius who has a history of seeing memetic opportunities that other people do not. So I feel comfortable noting I am doubtful and otherwise letting the reaction tell the story.
If you don’t dont, you end up with in a position where the audience can read a short book, and know what your claim is, and disagree with it because of one of the unanswered objections.
What’s the right way?