This post strikes me as fairly pedantic. Is there a live confusion it’s intending to solve?
The Wittgensteinian / Eliezerian view (something like “words are labels pointing to conceptual clusters that have fuzzy boundaries”) seems to fully dissolve the need to precisely specify definitions of words.
Longtermism’ is a new term, which may well become quite common and influential. The aim in giving the term a precise meaning while we still have the chance is to prevent confusions before they arise. This is particularly important if you’re hoping that a research field will develop around the idea. I think that this is really crucial.
Some confusions that happened in part because we were slow to give ‘Effective Altruism’ a precise definition: people unsure on how much EA required sacrifice and sometimes seeing it as extremely demanding; people unsure on whether you could be focused on preserving nature for its own sake and count as an EA; people seeing it as no different from applied utilitarianism.
Some confusions that are apt to arise with respect to longtermism:
are we talking about the strong version or the minimal version? The former is a lot more unintuitive, do we want to push that?
How long is long term? Are you in the longtermist club if you’re focused on the next hundred years? What if you’re focused on climate change? (that’s a bit of important pedantry I didn’t get into in the post!)
Are you committed to a particular epistemology? Ben Kuhn seemed to think so. But my next post is on what I call ‘boring longtermism’, which separates out longtermism from some other claims that long-term oriented EAs tend to endorse.
Is this just a thing for sci if nerds? Is this intellectual movement just focused on existential risk or something broader? Etc
Thanks – I agree that confusions are likely to arise somewhere as a new term permeates the zeitgeist.
I don’t think longtermism is a new term within EA or on the EA Forum, and I haven’t seen any recent debates over its definition.
[Edited: the Forum doesn’t seem like a well-targeted place for clarification efforts intending to address potential confusions around this (which seem likely to arise elsewhere)]. Encyclopedia entries, journal articles, and mainstream opinion pieces all seem better targeted to where confusion is likely to arise.
Even if the Forum isn’t a “well-targeted place” for a certain piece of EA content, it still seems good for things to end up here, because “getting feedback from people who are sympathetic to your goals and have useful background knowledge” is generally a really good thing no matter where you aim to publish something eventually.
Perhaps there will come a time in the future when “longtermism” becomes enough of a buzzword to justify clarification in a mainstream opinion piece or journal article. At that point, it seems good to have a history of discussion behind the term, and ideally one meaning that people in EA already broadly agree upon. (“This hasn’t been debated recently” =/= “we all have roughly the same definition that we are happy with”.)
The definitions of many words are fuzzy in practice, but that doesn’t mean it’s ideal for things to be that way. And I seriously doubt it’s ideal in technical research fields like philosophy or engineering.
In those cases shared and precise meanings can speed up research progress, and avoid terrible mistakes, by preventing misunderstandings.
Establishing some consistency in our terminology, like a shared definition of longtermism, strikes me as highly worthwhile.
I didn’t downvote as that would have caused the comment to be invisible, but do want to note that:
I find having to read comments that are written in an unnecessarily fight-y or dismissive style to be quite a significant tax on publishing blogposts, increasing the cost of publishing a blog post by 20% or more.
There are almost always easy alternatives to that style of language. Like, I am (clearly!) pretty anal about grammar issues in a way that’s ripe for gentle ribbing—seems like an emoticon-laden joke would have conveyed the same sentiment in a nicer way. Or the direct approach could have been simply saying, “I think the paragraphs on hyphenation could have been relegated to a footnote or appendix” (which seems very reasonable).
I downvoted the above comment, because I think it is more critical than helpful in a way that mildly frustrates me (because I’m not sure quite what you meant, or how to update my views of the post in response to your critique) and seems likely to frustrate the author (for similar reasons).
What is your goal in trying to make points about whether this information is “important to stay up-to-date about” or worth being “six paragraphs” long?
Do you think this post shouldn’t have been published? That it should have been shorter? That it would have been good to include more justification of the content before getting into detail about these definitions?
Raemon thought that it seems good for leaders to keep people updated on how they are conceptualizing things.
I argued that this doesn’t seem true in all cases, pointing out that six paragraphs on whether to hyphenate “longetermism” isn’t important to stay updated on, even when it comes from a leader.
---
For stuff like this, my ideal goal is something like “converge on the truth.”
I usually settle for consolation prizes like “get more clarity about where & how I disagree with other folks in EA” and/or “note my disagreements as they arise.”
… or how to update my views of the post in response to your critique
For what it’s worth, I suspect there’s enough inferential distance between us on fundamental stuff such that I wouldn’t expect either of us to be able to easily update while discussing topics on this level of abstraction.
“Update my views of the post” probably wasn’t the right phrase to use—better would be “update my views of whether the post is a good thing to have on the Forum in something like its current form”.
In general, I have a strong inclination that people should post content on the Forum if it is related to effective altruism and might be valuable to read for even a small fraction of users. I’m not concerned about too many posts being made (at least at the current level of activity).
I might be concerned if people seem to be putting more time/effort into posts than is warranted by the expected impact of those posts, but I have a high bar to drawing that conclusion; generally, I’d trust the author’s judgment of how valuable a thing is for them to publish over my own, especially if they are an expert writing about something on which I am a non-expert.
Even if the information in this post wasn’t especially new (I’m not sure which bits have and haven’t been discussed elsewhere), I expect it to be helpful to EA-aligned people who find themselves trying to communicate about longtermism with people outside of EA, for some of the reasons Will outlined. I can imagine referring to it as I work on an edition of the EA Newsletter or prepare for an interview with a journalist.
--
Finally, on hyphenation:
a. There are at least two occasions in the last two months that I, personally, have had to decide how to spell “longtermism” in something written for a public audience. And while I am an unusual case...
b. …hyphenation matters! Movements look less professional when they can’t decide how to express terms they often use in writing (hyphenation, capitalization, etc.) Something like this makes me say “huh?” before I even start reading the article (written by a critic of a movement in that case, but the general point stands).
These are far from the most important paragraphs ever published on the Forum, but they do take a stand on a point with two reasonable sides and argue convincingly for one of them, in a way that could change how many readers refer to a common term.
I downvoted your comments as well, Milan, because I think this is exactly the kind of thing that should go on the EA Forum. The emergence of this term “longtermism” to describe a vaguer philosophy that was already there has been a huge, perhaps the main EA topic for like 2 years. I don’t even subscribe to longtermism (well, at least not to strong longtermism, which I considered to be the definition before reading this post) but the question of whether to hyphenate has come up many times for me. This was all useful information that I’m glad was put up for engagement within EA.
And the objection that words can never be precise is pretty silly. Splitting hairs can be annoying but this was an important consideration of meaningfully different definitions of longtermism. It’s very smart for EA to figure this out now to avoid all the problems that Will mentioned, like vagueness, when the term has become more widely known.
It sounded like your objection was that this post was about words and strategy instead of about the concepts. I for one am glad that EA is not just about thinking but about doing what needs to be done, including reaching agreement about how to talk about ideas and what kind of pitches we should be making.
This post strikes me as fairly pedantic. Is there a live confusion it’s intending to solve?
The Wittgensteinian / Eliezerian view (something like “words are labels pointing to conceptual clusters that have fuzzy boundaries”) seems to fully dissolve the need to precisely specify definitions of words.
Longtermism’ is a new term, which may well become quite common and influential. The aim in giving the term a precise meaning while we still have the chance is to prevent confusions before they arise. This is particularly important if you’re hoping that a research field will develop around the idea. I think that this is really crucial.
Some confusions that happened in part because we were slow to give ‘Effective Altruism’ a precise definition: people unsure on how much EA required sacrifice and sometimes seeing it as extremely demanding; people unsure on whether you could be focused on preserving nature for its own sake and count as an EA; people seeing it as no different from applied utilitarianism.
Some confusions that are apt to arise with respect to longtermism:
are we talking about the strong version or the minimal version? The former is a lot more unintuitive, do we want to push that?
How long is long term? Are you in the longtermist club if you’re focused on the next hundred years? What if you’re focused on climate change? (that’s a bit of important pedantry I didn’t get into in the post!)
Are you committed to a particular epistemology? Ben Kuhn seemed to think so. But my next post is on what I call ‘boring longtermism’, which separates out longtermism from some other claims that long-term oriented EAs tend to endorse.
Is this just a thing for sci if nerds? Is this intellectual movement just focused on existential risk or something broader? Etc
Thanks – I agree that confusions are likely to arise somewhere as a new term permeates the zeitgeist.
I don’t think longtermism is a new term within EA or on the EA Forum, and I haven’t seen any recent debates over its definition.
[Edited: the Forum doesn’t seem like a well-targeted place for clarification efforts intending to address potential confusions around this (which seem likely to arise elsewhere)]. Encyclopedia entries, journal articles, and mainstream opinion pieces all seem better targeted to where confusion is likely to arise.
Even if the Forum isn’t a “well-targeted place” for a certain piece of EA content, it still seems good for things to end up here, because “getting feedback from people who are sympathetic to your goals and have useful background knowledge” is generally a really good thing no matter where you aim to publish something eventually.
Perhaps there will come a time in the future when “longtermism” becomes enough of a buzzword to justify clarification in a mainstream opinion piece or journal article. At that point, it seems good to have a history of discussion behind the term, and ideally one meaning that people in EA already broadly agree upon. (“This hasn’t been debated recently” =/= “we all have roughly the same definition that we are happy with”.)
The definitions of many words are fuzzy in practice, but that doesn’t mean it’s ideal for things to be that way. And I seriously doubt it’s ideal in technical research fields like philosophy or engineering.
In those cases shared and precise meanings can speed up research progress, and avoid terrible mistakes, by preventing misunderstandings.
Establishing some consistency in our terminology, like a shared definition of longtermism, strikes me as highly worthwhile.
I suspect the goal here is less to deconfuse current EAs and more to make it easier to explain things to newcomers who don’t have any context.
(It also seems like good practice to me for people in leadership positions to keep people up to date about how they’re conceptualizing their thinking)
Basically agree about the first claim, though the Forum isn’t really aimed at EA newcomers.
Eh, some conceptualizations are more valuable than others.
I don’t see how six paragraphs of Will’s latest thinking on whether to hyphenate “longtermism” could be important to stay up-to-date about.
I downvoted this comment because:
i) The hyphenation segment clearly isn’t the central argument of the post. This is a straw man.
ii) It’s generally a bit dismissive and unkind.
ii) If you don’t think something’s important to stay up-to-date on, you don’t have to read it or engage with it.
I didn’t downvote as that would have caused the comment to be invisible, but do want to note that:
I find having to read comments that are written in an unnecessarily fight-y or dismissive style to be quite a significant tax on publishing blogposts, increasing the cost of publishing a blog post by 20% or more.
There are almost always easy alternatives to that style of language. Like, I am (clearly!) pretty anal about grammar issues in a way that’s ripe for gentle ribbing—seems like an emoticon-laden joke would have conveyed the same sentiment in a nicer way. Or the direct approach could have been simply saying, “I think the paragraphs on hyphenation could have been relegated to a footnote or appendix” (which seems very reasonable).
I downvoted the above comment, because I think it is more critical than helpful in a way that mildly frustrates me (because I’m not sure quite what you meant, or how to update my views of the post in response to your critique) and seems likely to frustrate the author (for similar reasons).
What is your goal in trying to make points about whether this information is “important to stay up-to-date about” or worth being “six paragraphs” long?
Do you think this post shouldn’t have been published? That it should have been shorter? That it would have been good to include more justification of the content before getting into detail about these definitions?
Raemon thought that it seems good for leaders to keep people updated on how they are conceptualizing things.
I argued that this doesn’t seem true in all cases, pointing out that six paragraphs on whether to hyphenate “longetermism” isn’t important to stay updated on, even when it comes from a leader.
---
For stuff like this, my ideal goal is something like “converge on the truth.”
I usually settle for consolation prizes like “get more clarity about where & how I disagree with other folks in EA” and/or “note my disagreements as they arise.”
For what it’s worth, I suspect there’s enough inferential distance between us on fundamental stuff such that I wouldn’t expect either of us to be able to easily update while discussing topics on this level of abstraction.
“Update my views of the post” probably wasn’t the right phrase to use—better would be “update my views of whether the post is a good thing to have on the Forum in something like its current form”.
In general, I have a strong inclination that people should post content on the Forum if it is related to effective altruism and might be valuable to read for even a small fraction of users. I’m not concerned about too many posts being made (at least at the current level of activity).
I might be concerned if people seem to be putting more time/effort into posts than is warranted by the expected impact of those posts, but I have a high bar to drawing that conclusion; generally, I’d trust the author’s judgment of how valuable a thing is for them to publish over my own, especially if they are an expert writing about something on which I am a non-expert.
Even if the information in this post wasn’t especially new (I’m not sure which bits have and haven’t been discussed elsewhere), I expect it to be helpful to EA-aligned people who find themselves trying to communicate about longtermism with people outside of EA, for some of the reasons Will outlined. I can imagine referring to it as I work on an edition of the EA Newsletter or prepare for an interview with a journalist.
--
Finally, on hyphenation:
a. There are at least two occasions in the last two months that I, personally, have had to decide how to spell “longtermism” in something written for a public audience. And while I am an unusual case...
b. …hyphenation matters! Movements look less professional when they can’t decide how to express terms they often use in writing (hyphenation, capitalization, etc.) Something like this makes me say “huh?” before I even start reading the article (written by a critic of a movement in that case, but the general point stands).
These are far from the most important paragraphs ever published on the Forum, but they do take a stand on a point with two reasonable sides and argue convincingly for one of them, in a way that could change how many readers refer to a common term.
I downvoted your comments as well, Milan, because I think this is exactly the kind of thing that should go on the EA Forum. The emergence of this term “longtermism” to describe a vaguer philosophy that was already there has been a huge, perhaps the main EA topic for like 2 years. I don’t even subscribe to longtermism (well, at least not to strong longtermism, which I considered to be the definition before reading this post) but the question of whether to hyphenate has come up many times for me. This was all useful information that I’m glad was put up for engagement within EA.
And the objection that words can never be precise is pretty silly. Splitting hairs can be annoying but this was an important consideration of meaningfully different definitions of longtermism. It’s very smart for EA to figure this out now to avoid all the problems that Will mentioned, like vagueness, when the term has become more widely known.
It sounded like your objection was that this post was about words and strategy instead of about the concepts. I for one am glad that EA is not just about thinking but about doing what needs to be done, including reaching agreement about how to talk about ideas and what kind of pitches we should be making.