First I want to say that I think your original comment and this one both express reasonable views, and do so in a civil manner. (Also, just in case you or anyone else was wondering, I’ve neither upvoted nor downvoted either comment.)
Also, while I think I disagree with you to some extent on some points, I think your comments have made me think more about things worth thinking about. I think they’ve also improved this post, via prompting me to add the following to the introductory section:
(Edit: I think that recommendations that aren’t commonly mentioned in EA are particularly valuable, holding general usefulness and EA-relevance constant. Same goes for recommendations of books by non-male, non-white, and/or non-WEIRD authors. See this comment thread.)
(I added part of that after your first comment, and the second sentence after reading your second comment.)
Also, I acknowledge that there are two separate (though related) points you’re highlighting, and that my reply didn’t explicitly address the gender diversity part.
You only list male authors and lists that only feature male authors: all of them are [also] WEIRD (Western, Educated, Industrialized, Rich, and Democratic).
I believe this is indeed true. (The part I’m slightly unsure of is whether all authors meet all WEIRD criteria, but in any case it definitely heavily skews WEIRD.) I also hadn’t specifically noticed that this was the case for all/almost all of these listed authors, so it does seem useful to me that you highlighted it. And I do agree that, all else equal, it’d be better to have more diversity on each of these dimensions on someone’s reading list.
One thing I’d say in response is that there’s more demographic diversity in the other forms of content I consume (in particular, papers, podcasts, and posts) than in these books. Though those other forms of content I consume do still skew somewhat towards male and WEIRD (and also white).
I think three ways to address that are for me to:
Get more recommendations for content to consume that’s by authors (or podcasters, or whatever) with other demographic characteristics
Make more effort to actively seek out ideas for content to consume that’s from authors with other demographic characteristics
Factor in the demographic characteristics of an author when deciding which of multiple specific pieces of content I should spend time consuming
I’d be very happy if (1) happened. And I think hopefully this post should contribute to it happening, especially now that I’ve added an edit prompted by your comments. I think your own first comment already helps with that, which I appreciate.
I think I should also do a bit of (2) and (3). (For a start, I’ve just now downloaded Obama’s A Promised Land, and made a note to maybe read Strangers Drowning.) But I’m unsure how much, because I’m unsure how much weight I should give to demographic diversity relative to other factors when making tradeoffs about how I spend my time. And it just does seem to be the case that, in many of the fields I want to learn about, most of the most prominent authors—meaning prominent among e.g. the relevant academics, not just EAs—are male, WEIRD, and white. And I think there’s value in reading things from the most prominent authors.
But I’m personally inclined not to debate here precisely how much of (2) and (3) people should do, and the precise extent to which the prominent authors in these fields skew male, WEIRD, and white. This is because I’m concerned that that might result in a long and tense thread, partly due to this being a written medium with an audience and a lot of people not personally knowing each other, rather than a face-to-face conversation.
(I’m not trying to silence such a debate; it can be had here; I’m just personally inclined for it to not happen here. Readers may also be interested in posts tagged diversity and inclusion.)
(I’d also like to pre-emptively ask readers to keep in mind that it’s easy to interpret things overly harshly when they’re written down on the internet by someone you don’t know personally. If you think I or Hauke are saying things that are stupid or horrible, please seriously consider the hypothesis that that’s not really what I or Hauke mean, or that we just phrased things poorly, or something like that.)
Thanks for the courteous reply. Agree with much of this!
To be clear, I didn’t mean to criticize you or anyone personally. Though judging by the downvotes I got, people might think that I’m EA’s wokest and hardest virtue-signalling SJW, but I actually only realized and was able to flag this issue because I’m guilty of recommending a very similar set of male authors too much myself. So this is something that should be improved more generally (in the community). Also, I agree that we shouldn’t spend much time on finding a precise ‘quota’ and I’m not saying that we should have 50% of women on AI safety syllabi (which would probably leave people scrambling and is more a society-wide issue) or cancel Toby Ord, but on current margin, we should probably err on the side of having a little more diversity in what we recommend. Not upvoting a list with 50 white males trending on the front page and implicitly endorse this as the EA cannon seems a really low bar. Hence the initial downvote, which I’ve now changed to an upvote, given that there’s a productive discussion in the comments, in particular thanks to Michael.
To be clear, I didn’t mean to criticize you or anyone personally. [...] Also, I agree that we shouldn’t spend much time on finding a precise ’quota
Yeah, to be clear, I didn’t get the impression of being criticised in a way that singles me out quite specifically, and my points about being inclined not to discuss the precise amount of (2) and (3) I should do was not me saying “You’ve said too much about this already!”, but rather “I’m a little concerned that this thread could become overly spicy and contentious” (and I primarily had in mind other people jumping in; I wasn’t worried about comments you’d write). I think the comments so far have been civil, as I mentioned.
on current margin, we should probably err on the side of having a little more diversity in what we recommend
Agreed.
Not upvoting a list with 50 white males trending on the front page and implicitly endorse this as the EA cannon seems a really low bar.
I’m not totally sure I agree, partly because every Forum post starts out on the front page, and I think it’d be really easy for EA to be flooded with a bunch more recommendation lists. So I think (a) I estimate a lower chance that this list ends up being extremely prominent than you do, and (b) if we’re worried about this list being too prominent, I think the best solution is just to vigorously encourage the posting of more lists (including ones with more demographically diverse authors).
As you noted, the Wiblin, Beckstead, and Muehlhauser lists are already quite prominent, and also skew towards male, white, WEIRD, etc. So I think it may be the case that “the only way out is through”—i.e., the best way to prevent there being too much focus on a small set of lists is to post more, not to avoid posting.
But, that of course wouldn’t fix the demographic diversity issue, unless those other lists either happen to include or are encouraged to include more demographic diversity. So you highlighting this with your comment seems useful.
(But I genuinely just mean “I’m not totally sure I agree”; I think your sentence is a reasonable claim.)
judging by the downvotes I got
Yeah, I don’t like that your comment is currently on net negative karma. I’m going to strong upvote it for balance’s sake, and make a separate comment about that.
First I want to say that I think your original comment and this one both express reasonable views, and do so in a civil manner. (Also, just in case you or anyone else was wondering, I’ve neither upvoted nor downvoted either comment.)
Also, while I think I disagree with you to some extent on some points, I think your comments have made me think more about things worth thinking about. I think they’ve also improved this post, via prompting me to add the following to the introductory section:
(I added part of that after your first comment, and the second sentence after reading your second comment.)
Also, I acknowledge that there are two separate (though related) points you’re highlighting, and that my reply didn’t explicitly address the gender diversity part.
I believe this is indeed true. (The part I’m slightly unsure of is whether all authors meet all WEIRD criteria, but in any case it definitely heavily skews WEIRD.) I also hadn’t specifically noticed that this was the case for all/almost all of these listed authors, so it does seem useful to me that you highlighted it. And I do agree that, all else equal, it’d be better to have more diversity on each of these dimensions on someone’s reading list.
One thing I’d say in response is that there’s more demographic diversity in the other forms of content I consume (in particular, papers, podcasts, and posts) than in these books. Though those other forms of content I consume do still skew somewhat towards male and WEIRD (and also white).
I think three ways to address that are for me to:
Get more recommendations for content to consume that’s by authors (or podcasters, or whatever) with other demographic characteristics
Make more effort to actively seek out ideas for content to consume that’s from authors with other demographic characteristics
Factor in the demographic characteristics of an author when deciding which of multiple specific pieces of content I should spend time consuming
I’d be very happy if (1) happened. And I think hopefully this post should contribute to it happening, especially now that I’ve added an edit prompted by your comments. I think your own first comment already helps with that, which I appreciate.
I think I should also do a bit of (2) and (3). (For a start, I’ve just now downloaded Obama’s A Promised Land, and made a note to maybe read Strangers Drowning.) But I’m unsure how much, because I’m unsure how much weight I should give to demographic diversity relative to other factors when making tradeoffs about how I spend my time. And it just does seem to be the case that, in many of the fields I want to learn about, most of the most prominent authors—meaning prominent among e.g. the relevant academics, not just EAs—are male, WEIRD, and white. And I think there’s value in reading things from the most prominent authors.
But I’m personally inclined not to debate here precisely how much of (2) and (3) people should do, and the precise extent to which the prominent authors in these fields skew male, WEIRD, and white. This is because I’m concerned that that might result in a long and tense thread, partly due to this being a written medium with an audience and a lot of people not personally knowing each other, rather than a face-to-face conversation.
(I’m not trying to silence such a debate; it can be had here; I’m just personally inclined for it to not happen here. Readers may also be interested in posts tagged diversity and inclusion.)
(I’d also like to pre-emptively ask readers to keep in mind that it’s easy to interpret things overly harshly when they’re written down on the internet by someone you don’t know personally. If you think I or Hauke are saying things that are stupid or horrible, please seriously consider the hypothesis that that’s not really what I or Hauke mean, or that we just phrased things poorly, or something like that.)
Thanks for the courteous reply. Agree with much of this!
To be clear, I didn’t mean to criticize you or anyone personally. Though judging by the downvotes I got, people might think that I’m EA’s wokest and hardest virtue-signalling SJW, but I actually only realized and was able to flag this issue because I’m guilty of recommending a very similar set of male authors too much myself. So this is something that should be improved more generally (in the community). Also, I agree that we shouldn’t spend much time on finding a precise ‘quota’ and I’m not saying that we should have 50% of women on AI safety syllabi (which would probably leave people scrambling and is more a society-wide issue) or cancel Toby Ord, but on current margin, we should probably err on the side of having a little more diversity in what we recommend. Not upvoting a list with 50 white males trending on the front page and implicitly endorse this as the EA cannon seems a really low bar. Hence the initial downvote, which I’ve now changed to an upvote, given that there’s a productive discussion in the comments, in particular thanks to Michael.
Yeah, to be clear, I didn’t get the impression of being criticised in a way that singles me out quite specifically, and my points about being inclined not to discuss the precise amount of (2) and (3) I should do was not me saying “You’ve said too much about this already!”, but rather “I’m a little concerned that this thread could become overly spicy and contentious” (and I primarily had in mind other people jumping in; I wasn’t worried about comments you’d write). I think the comments so far have been civil, as I mentioned.
Agreed.
I’m not totally sure I agree, partly because every Forum post starts out on the front page, and I think it’d be really easy for EA to be flooded with a bunch more recommendation lists. So I think (a) I estimate a lower chance that this list ends up being extremely prominent than you do, and (b) if we’re worried about this list being too prominent, I think the best solution is just to vigorously encourage the posting of more lists (including ones with more demographically diverse authors).
As you noted, the Wiblin, Beckstead, and Muehlhauser lists are already quite prominent, and also skew towards male, white, WEIRD, etc. So I think it may be the case that “the only way out is through”—i.e., the best way to prevent there being too much focus on a small set of lists is to post more, not to avoid posting.
But, that of course wouldn’t fix the demographic diversity issue, unless those other lists either happen to include or are encouraged to include more demographic diversity. So you highlighting this with your comment seems useful.
(But I genuinely just mean “I’m not totally sure I agree”; I think your sentence is a reasonable claim.)
Yeah, I don’t like that your comment is currently on net negative karma. I’m going to strong upvote it for balance’s sake, and make a separate comment about that.