Because it reads like an undergraduate essay that regurgitates the ideas from the lectures and the required reading, without any sense that the writer himself is actually thinking about the ideas and trying to decide the truth of anything (which undergraduates in philosophy aren’t supposed to be doing anyway — see Pirsig’s concept of “philosophology”). All that such an essay accomplishes is to show that the student (or AI assistant) has learned the names of the important people and has at least some idea of what they said.
It does not reach much of a conclusion and does not make much of an argument for it. The conclusion seems (because it comes at the end) to be that “We should treat morality as a compelling fiction”. The fictionalism section is suffused with the idea of moral progress, yet the concept is not queried. Whence the “should”, whence the compulsion, and whence the claim of progress? The writer questions all the concepts of morality previously discussed and finds them wanting, but this time the questions go unasked and unanswered.
You could reorder the sections, making any of them the conclusion. All it takes is to not argue against whichever one you put last. Maybe the lecturer is in the fictionalist camp.
Not the one who downvoted it (although I am also tempted to), but here is my objection: The text is written in a needlessly complicated way, which may be a required ritual when publishing in academic sources, but not necessary when communicating on a blog, and it only makes it more difficult for the readers. It’s like writing in Latin.
For example:
J.L. Mackie’s “argument from queerness” identifies the core problem.
Yeah, I guess I am supposed to know what that means. Trying to find some answers online, I think it means the following: “Moral properties are strange, different from other properties of physical objects in this universe. Therefore we must have a special way of evaluating them. Further research is needed.”
This is academic writing, where the clarity of text is low priority, and making proper references to high-status sources is high priority. If instead you tried to explain the while situation to a smart teenager, you would probably say something like “When you look at physical objects, you see nothing inherently good or bad. Even looking at humans from the perspective of ‘they are huge systems of atoms’ does not provide a foundation for what ‘good’ and ‘bad’ might mean. Intuitively, we care about things like love or suffering, but there is no obvious reason why a robot or a space alien (possibly incapable of any of these feelings) should care about that. If we want to keep using these words—and yes, they seem useful in many important situations—we need a definition, and will probably have no simple relation to the physical properties of things.”
The obvious question is: why is it better to write for smart teenagers? Shouldn’t we optimize for higher-status people instead?
My answer is that this is related to the problem of “distillation”. The world is already difficult. If we make things more difficult by not communicating as clearly as we could, the tower of knowledge will collapse under its own weight. (Well, at some point it will collapse anyway, but sooner if we keep adding extra weight.)
It’s like badly written code. It may still compile and execute correctly, but at some moment you should refactor it.
Almost everything here is written in a needlessly complicated way. for instance, you dont have to write “utility function” when you mean “preferences”.
There are situations that may appear the same to most voters (sometimes possibly to all).
it is a good article, but requires expert audience
it is correct but relatively banal, i.e. it requires expert audience to read it, but such audience would find no new insights in the article
it is a crackpot article, but only expect audience can reliably recognize it
For each of those, only the expert audience can make an opinion other than “over my head”.
There is a risk that when everyone clicks “over my head”, all three kinds of posts will pass. Especially if they get an upvote or two, because people are more likely to upvote than downvote in general.
The voting system can’t fix the people doing the voting. From behaviour , what they want is at least one of the (usual content, usual style)..the same old same old. What they don’t want is to be challenged or stretched.
My answer is that this is related to the problem of “distillation”. The world is already difficult. If we make things more difficult by not communicating as clearly as we could
That’s a great idea, and you, collectively, should try it. Just look at how badly you are communicating the central claim of AI threat.
Why on earth has this been downvoted?
Because it reads like an undergraduate essay that regurgitates the ideas from the lectures and the required reading, without any sense that the writer himself is actually thinking about the ideas and trying to decide the truth of anything (which undergraduates in philosophy aren’t supposed to be doing anyway — see Pirsig’s concept of “philosophology”). All that such an essay accomplishes is to show that the student (or AI assistant) has learned the names of the important people and has at least some idea of what they said.
It does not reach much of a conclusion and does not make much of an argument for it. The conclusion seems (because it comes at the end) to be that “We should treat morality as a compelling fiction”. The fictionalism section is suffused with the idea of moral progress, yet the concept is not queried. Whence the “should”, whence the compulsion, and whence the claim of progress? The writer questions all the concepts of morality previously discussed and finds them wanting, but this time the questions go unasked and unanswered.
You could reorder the sections, making any of them the conclusion. All it takes is to not argue against whichever one you put last. Maybe the lecturer is in the fictionalist camp.
You know thats also true of many highly upvoted postings in The Sequences?
I do not know this, having no more interest in Sequenceology than I do in philosophology.
Not the one who downvoted it (although I am also tempted to), but here is my objection: The text is written in a needlessly complicated way, which may be a required ritual when publishing in academic sources, but not necessary when communicating on a blog, and it only makes it more difficult for the readers. It’s like writing in Latin.
For example:
Yeah, I guess I am supposed to know what that means. Trying to find some answers online, I think it means the following: “Moral properties are strange, different from other properties of physical objects in this universe. Therefore we must have a special way of evaluating them. Further research is needed.”
This is academic writing, where the clarity of text is low priority, and making proper references to high-status sources is high priority. If instead you tried to explain the while situation to a smart teenager, you would probably say something like “When you look at physical objects, you see nothing inherently good or bad. Even looking at humans from the perspective of ‘they are huge systems of atoms’ does not provide a foundation for what ‘good’ and ‘bad’ might mean. Intuitively, we care about things like love or suffering, but there is no obvious reason why a robot or a space alien (possibly incapable of any of these feelings) should care about that. If we want to keep using these words—and yes, they seem useful in many important situations—we need a definition, and will probably have no simple relation to the physical properties of things.”
The obvious question is: why is it better to write for smart teenagers? Shouldn’t we optimize for higher-status people instead?
My answer is that this is related to the problem of “distillation”. The world is already difficult. If we make things more difficult by not communicating as clearly as we could, the tower of knowledge will collapse under its own weight. (Well, at some point it will collapse anyway, but sooner if we keep adding extra weight.)
It’s like badly written code. It may still compile and execute correctly, but at some moment you should refactor it.
Almost everything here is written in a needlessly complicated way. for instance, you dont have to write “utility function” when you mean “preferences”.
Is moral philosophy 101.
Then I am clearly not the intended audience.
And the question his, how is the voting system supposed to work in such case.
Maybe be a third pair of buttons , for “over my head” is needed.
There are situations that may appear the same to most voters (sometimes possibly to all).
it is a good article, but requires expert audience
it is correct but relatively banal, i.e. it requires expert audience to read it, but such audience would find no new insights in the article
it is a crackpot article, but only expect audience can reliably recognize it
For each of those, only the expert audience can make an opinion other than “over my head”.
There is a risk that when everyone clicks “over my head”, all three kinds of posts will pass. Especially if they get an upvote or two, because people are more likely to upvote than downvote in general.
The voting system can’t fix the people doing the voting. From behaviour , what they want is at least one of the (usual content, usual style)..the same old same old. What they don’t want is to be challenged or stretched.
That’s a great idea, and you, collectively, should try it. Just look at how badly you are communicating the central claim of AI threat.