23

Kelly_Witwicki comments on Why & How to Make Progress on Diversity & Inclusion in EA - Effective Altruism Forum

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (229)

You are viewing a single comment's thread. Show more comments above.

Comment author: xccf 27 October 2017 03:04:41AM *  32 points [-]

Thanks for this post. There's a lot I agree with here. I'm in especially vigorous agreement with your points regarding hero worship and seeing newcomers as a source of fresh ideas/arguments instead of condescending them.

There are also some points I disagree with. And in the spirit of not considering any arguments above criticism, and disagreement being critical for finding the best answers, I hope you won't mind if I lay my disagreements out. To save time, I'll focus on the differences between your view and mine. So if I don't mention a point you made, you can default to assuming I agree with it.

First, I'm broadly skeptical of the social psychology research you cite. Whenever I read about a study that claims women are more analytical than men, or women are better leaders than men, I imagine whether I would hear about it if the experiment found the opposite result.

I recommend this blog post on the lack of ideological diversity in social psychology. Social psychologists are overwhelmingly liberal, and many openly admit to discriminating against conservatives in hiring. Here is a good post by a Mexican social psychologist that discusses how this plays out. There's also the issue of publication bias at the journal level. I know someone who served on the selection committee of a (minor & unimportant, so perhaps not representative) psychology journal. The committee had an explicit philosophy of only publishing papers they liked, and espousing "problematic" views was a strike against a paper. Anyway, I think to some degree the field functions as a liberal echo chamber on controversial issues.

There's really an entire can of worms here--social psychology is currently experiencing a major reproducibility crisis--but I don't want to get too deep in to it, because to defend my position fully, I'd want to share evidence for positions that make people uncomfortable. Suffice to say that there's a third layer of publication bias at the level of your Facebook feed, and I could show you a different set of research-backed thinkpieces that point to different conclusions. (Suggestion: if you wouldn't want someone on the EA Forum to make arguments for the position not X, maybe avoid making arguments for the position X. Otherwise you put commenters in an impossible bind.)

But for me this point is really the elephant in the room:

some people in broader society now respond to correctable offenses with a mob mentality and too much readiness for ostracization, but just because some people have swung too far past the mark doesn’t mean we should default to a status quo that falls so short of it.

I would like to see a much deeper examination here. Insofar as I feel resistant to diversity efforts, this feels like most of what I'm trying to resist. If I was confident that pro-diversity people in EA won't spiral towards this, I'd be much more supportive. Relevant fable.

All else equal, increased diversity sounds great, but my issue is I see a pattern of other pro-diversity movements sacrificing all other values in the name of trying to increase diversity. Take a statement like this one:

Some of the most talented and resolute people in this community are here because they are deeply emotionally compelled to help others as much as possible, and we’re currently missing out on many such people by being so cold and calculating. There are ways to be warm and calculating! I can think of a few people in the community who manage this well.

Being warm and calculating sounds great, but what if there's actually a tradeoff here? Just taking myself as an example, I know that as I've become aware of how much suffering exists in the grand scheme of things, I've begun to worry less about random homeless people I see and stuff like that. Even if there's some hack I can use to empathize with homeless people while retaining a global perspective, that hack would require effort on my part--effort I could put towards goals that seem more important.

this particular individual — who is probably a troll in general — was banned from the groups where he repeatedly and unrelentingly said such things, though it’s concerning there was any question about whether this was acceptable behavior.

Again, I think there's a real tradeoff between "free speech" and sensitivity. I view the moderation of online communities as an unsolved problem. I think we benefit from navigating moderation tradeoffs thoughtfully rather than reactively.

Reminding people off the forum to upvote this post, in order to deal with possible hostility, is also a minor red flag from my perspective. This resembles something Gleb Tsipursky once did.

None of this seems very bad in the grand scheme of things, especially not compared to what I've seen from other champions of diversity--I just thought it'd be useful to give concrete examples.

Anyway, here are some ideas of mine, if anyone cares:

  • Phrase guidelines as neutrally as possible, e.g. "don't be a jerk" instead of "don't be a sexist". The nice thing about "don't be a jerk" is it at admits the possibility that someone could violate the guideline by e.g. loudly calling out a minor instance of sexism in a way that generates a lot of drama and does more harm than good. Rules should exist to serve everyone, and they should be made difficult to weaponize. If most agree your rules are legitimate, that also makes them easier to enforce.

  • Team-building activities, icebreakers, group singalongs, synchronous movement, sports/group exercise, and so on. The ideal activity is easy for anyone to do and creates a shared EA tribal identity just strong enough to supersede the race/gender/etc. identities we have by default. Kinda like how students at the same university will all cheer for the same sports team.

  • Following the example of the animal-focused EAs: Work towards achieving critical mass of underrepresented groups. Especially if you can saturate particular venues (e.g. a specific EA meetup group). I know that as a white male, I sometimes get uncomfortable in situations where I am the only white person or the only man in a group, even though I know perfectly well that no one is discriminating against me. I think it's a natural response to have when you're in the minority, so in a certain sense there's just a chicken-and-egg problem. Furthermore, injecting high-caliber underrepresented people into EA will help dismantle stereotypes and increase the number of one-on-one conversations people have, which I think are critical for change.

  • Take a proactive, rather than reactive, approach to helping EA men with women. Again, I think having more women is playing a big role for animal-focused EAs. More women means the average man has more female friends, better understands how women think, and empathizes with the situations women encounter more readily. In this podcast, Christine Peterson discusses the value of finding a life partner for productivity and mental health. In the same way that CFAR makes EAs more productive through lifehacking, I could imagine someone working covertly to make EAs more productive through solving their dating problems.

  • Invite the best thinkers who have heterodox views on diversity to attend "diversity in EA" events, in order to get a diverse perspective on diversity and stay aware of tradeoffs. Understand their views in enough depth to market diversity initiatives to the movement at large without getting written off.

  • When hiring a Diversity & Inclusion Officer, find someone who's good at managing tradeoffs rather than the person who's most passionate about the role.

Again, I appreciate the effort you put in to this post, and I support you working towards these goals in a thoughtful way. Also, I welcome PMs from you or anyone else reading this comment--I spent several hours on it, but I'm sure there is stuff I could have put better and I'd love to get feedback.

Comment author: Kelly_Witwicki 27 October 2017 12:57:49PM *  2 points [-]

I can't address all of this but will say three quick things:

I'm broadly skeptical of the social psychology research you cite

I appreciate it's weakness, but it's at least some evidence against people's intuitions and in addition to the literature on how those intuitions are demonstrably false and discriminatory it should update people away from those discriminatory beliefs.

[Edit: I appreciate that I should generally behave as though my community will behave well, and as such I should not have requested that people upvote even if I just asked them to "upvote if [they] find the post useful." I want to be sure to flag in this response though the incredibly poor way in which people who disagree with claims and arguments in favor of diversity and inclusion are using their votes, in comments and on the whole post. It's worth explicitly observing that identity-driven voting here is not equal among opposers and supporters, but seems clearly dominated by opposers.]

I appreciate your suggestions a lot, but caution you to be careful of your own assumptions. For instance, I never suggested that a Diversity & Inclusion Officer should be the person most passionate about the role instead of most smart about it.

To emphasize though, so it doesn't get lost behind those critical thoughts: I thoroughly appreciate the suggestions you've contributed here.

[Edit: Apologies for some excessive editing. I readily acknowledge that in an already a hostile environment, my initial reaction to criticism regarding an important issue that is causing a lot of harm is too defensive.]

Comment author: xccf 28 October 2017 04:38:14AM *  9 points [-]

Another idea I had: add questions to the EA Survey to understand how people feel about the issues you are describing. This accomplishes a few things:

  • It allows us to track progress more effectively than observing our demographic breakdown. Measuring how people feel about EA movement culture gives us a shorter feedback loop, since changes in demographics lag behind culture changes. Furthermore, by attempting to measure the climate issue directly, we can zero in on factors under our control.

  • It helps fight selection effects that occur in online discussion of these issues. People on both sides can be reluctant to share their thoughts & ideas in a thread like this one. Online discussions in general can be wildly unrepresentative. I was surprised to learn about polls which found that most Native Americans aren't offended by the use of "Redskins" as a team name (criticism of this poll), and that a majority of black people are against affirmative action. And among the "anti-SJW" crowd, there's a perception that some folks are going to see racism/sexism in everything, and they will never be satisfied. So taking a representative poll of EAs, and perhaps comparing the results to some baseline, can help us come to agreement on the degree to which we have issues.

Comment author: Kelly_Witwicki 29 October 2017 03:02:20AM 2 points [-]

I like this idea. It will be skewed towards people who aren't turned off by the culture, as those who are will have less interest in, and in some or many cases may not even be exposed to, the survey, but getting more systematic info on people's feelings here would be very useful.

Comment author: xccf 30 October 2017 05:12:23AM *  9 points [-]

Some more thoughts:

  • I mentioned my concern that pro-diversity efforts in EA might "spiral" towards a mob mentality. I think one way in which this might happen is if the people working towards diversity in EA recruit people from underrepresented groups that they know through other pro-diversity groups, which, as you mention, frequently suffer from a mob mentality. If the pool of underrepresented people we draw from is not selected this way (e.g. if the majority of black people who are joining EA are against affirmative action, as is true for the majority of the black population in general), then I'm less worried.

  • I think some of your suggestions are not entirely consistent. For example, you mention that EA should not "throw around the term “AI” with no qualification or explanation". From my perspective, if I was hearing about EA for the first time and someone felt the need to explain what "AI" was an acronym for, I would feel condescended to. I imagine this effect might be especially acute if I was a member of a minority group ("How dumb do these people think I am?") Similarly, you suggest that we cut our use of jargon. In practice, I think useful jargon is going to continue getting used no matter what. So the way this suggestion may be interpreted in practice is: Don't use jargon around people who are members of underrepresented groups. I think people from underrepresented groups will soon figure out they are being condescended to. I think a better idea is to remember that we were once ignorant about jargon ourselves, and make an effort to explain jargon to newbies. Hopefully they feel like members of the ingroup after they've mastered the lingo.

  • Relatedly, there is a question which I think sometimes gets tied up with the diversity question, but perhaps should not get tied up, which is the question of whether EA should aim more to be a committed, elite core vs a broad church. My impression is lots of people privately favor the committed, elite core approach. I think we can have both diversity and a committed, elite core: consider institutions such as Harvard which are both elite and diverse. Furthermore, I think being more public about our elitism might actually help with diversity, because we'd be making our standards clearer and more transparent, and we could rely less heavily on subjective first impressions. (CC Askell on "buzz talk".) To put it another way: although "diversity" and "inclusion" are often treated as synonyms, it's actually possible to be both "diverse" and "exclusive" (and this seems likely ideal).

  • A benefit of diversity you didn't mention: Insofar as the EA movement has world peace and global cooperation as part of our goals, it's useful to have people from as many different groups as possible. This is also useful if we want to be able to speak authoritatively on topics like how AI should be used for the benefit of humanity and whatnot.

  • Unjustified hunch here, but I think maybe another failure mode that can come up when a movement tries to increase diversity is that people who are underrepresented start to receive more attention. Even if this attention is positive (e.g. "How can we cater to people like you better?"), I think this can result in an increased level of self-consciousness. (See my previous point about how people who look different may feel self-conscious by default even if they're not discriminated against.) Further unjustified conjecture: the sort of black person who supports affirmative action tends to enjoy the power they get from this, whereas the sort of black person who doesn't support affirmative action doesn't like it, thereby enhancing the "spiral" effect.

  • Another possible failure mode: Diversity advocates see something they don't like (e.g. a person suggesting that women do not contribute to society and are leeches if they don’t offer men sex), and they want to root the problem out. In order to rally support, they let everyone know about the problem (like you did in this post). But by letting everyone know about the problem, they've also made it in to a bigger problem: now every woman who reads this post knows that someone, at one point in an EA-related discussion somewhere, made this outrageous claim--which results in those women feeling less welcome and more on edge. The toxic echo of this person's post continues to reverberate as it is held up as part of a broader trend within EA, even though their post itself was long ago deleted. (This could contribute to the "spiral" effect I described, if the women who stick around after hearing about posts like these are disproportionately those that enjoy engaging in flame wars with people who make outrageous statements.)

  • I mentioned the EA Survey. One thing you could do is look at existing EA survey data and try to understand whether our issues with underrepresentation seem to be getting better or worse over the years. My impression is that gender thing, at least, has gotten much better since EA was founded. In any case, if things are already on a good path, I'm more skeptical about major diversity initiatives--"if it ain't broke, don't fix it".

Incidentally, I realized some of the points I'm making here are redundant with this essay which was already posted. (But I highly recommend reading it anyway, because it has some great points I hadn't thought of.)

Comment author: ateabug 30 October 2017 01:57:23PM *  1 point [-]

But by letting everyone know about the problem, they've also made it in to a bigger problem: now every woman who reads this post knows that someone, at one point in an EA-related discussion somewhere, made this outrageous claim--which results in those women feeling less welcome and more on edge. The toxic echo of this person's post continues to reverberate as it is held up as part of a broader trend within EA, even though their post itself was long ago deleted.

This can get very dangerous as it opens a door for trolls to negatively impact the community and potentially damage its reputation. Maybe these kinds of discussions need to be gated in some way, or be had offline or something.

Comment author: Kelly_Witwicki 30 October 2017 07:29:30PM 0 points [-]

Risk does come with greater publicity of such behavior, but that's part of the point of making it more public (in addition to the information value for people who want to avoid or address it). This is the first I've ever publicly said something about these issues in EA, after three years of many private conversations that seem to have resulted in limited or no impact. Greater publicity means greater accountability and motivation for action, both for the people who behave poorly and the people who let them do so without consequence.

Comment author: xccf 31 October 2017 11:30:05PM 0 points [-]

Out of curiosity, have you tried anything besides private conversations?

Comment author: Kelly_Witwicki 01 November 2017 12:54:51AM *  1 point [-]

Since I'm already working on inclusionary practices myself, there's not much else to do but private or public discussion.

The private discussions I have had explicitly around the issue have varied a lot in their content and purpose and can be characterized as any of the following or a combination thereof: Listening to people's experiences; sharing my own; discussing solutions; actively (beyond just listening) supporting people who were treated poorly; sharing information and concern about the issue with people in a better or still good position to do something about it; trying to discuss why this or more specific issues of exclusion are a problem with people who prefer the status quo; or endeavoring to show people why something they did was a problem and what they should do differently.

Dealing with a bewilderingly amateur situation myself and working to privately help the people responsible to understand the problem and improve took a month out of my life, and with a really important counterfactual, and that's strictly in time spent on the issue that I don't think I would have had to lose in e.g. the animal advocacy community, and not accounting for the emotional toll. I have good reason for (cautious) optimism that that was fruitful but also a red flag restraining that optimism and regardless only time will tell.

Basically I've spent a huge amount of time on those private and often solution-oriented conversations and have been hanging over the precipice of burnout with the community since day 1 several years ago. (The broader community at least, not the animal advocacy sub/intersected-community. And disclaimer that there are great individuals throughout the broader community who are my friends and/or whose presence in the community I am so happy for, etc.) And I'm definitely not alone in that.

I can do more to have private conversations with people in better positions than myself to make change here (such as people who are looked up to in the community by the people whose behavior could be more inclusionary, or donors to EA orgs), and I might if this post and the discussion here doesn't inspire other people to take more action on this issue, which is my hope.

Comment author: Buck 27 October 2017 11:32:29PM 2 points [-]

it's at least some evidence against people's intuitions

I don't think it is, at all, any more than Daryl Bem's research updates me towards thinking ESP is real. Like, who knows, the world is a crazy place, maybe the papers here are in the 36% of published psychology papers which hold up under replication. But I don't think that it makes sense to update against your beliefs about this stuff based on the published science--if you think that the scientists would have published these papers regardless of their truth, as I do, you shouldn't regard them as evidence.

Comment author: xccf 28 October 2017 12:51:26AM 4 points [-]

I think you're overstating your case.

I don't think it is, at all, any more than Daryl Bem's research updates me towards thinking ESP is real.

This strikes me as a misunderstanding of how Bayesian updates work. The reason you still don't believe in ESP is because your prior for ESP is very low. But I think hearing about Bem's research should still cause you to update your estimate in favor of ESP a tiny amount. In a world with ESP, Bem finds it easier to discover ESP effects.

if you think that the scientists would have published these papers regardless of their truth

I don't think social psychologists are that dishonest. Even 36% replicability suggests some relationship between paper-publishing and truth.

Furthermore, I think the fact that social psychologists are so liberal should cause some update in the direction that studying humans causes you to realize liberal views about human nature are correct.

Comment author: Buck 28 October 2017 01:16:45AM 3 points [-]

This strikes me as a misunderstanding of how Bayesian updates work. The reason you still don't believe in ESP is because your prior for ESP is very low. But I think hearing about Bem's research should still cause you to update your estimate in favor of ESP a tiny amount. In a world with ESP, Bem finds it easier to discover ESP effects.

I think you slightly misunderstand me. What I'm saying is that Bem's work isn't really a Bayesian update for me, because I think Bem is approximately as likely to publish papers in the world where (extremely weak) ESP works as the worlds where it doesn't. The strength of my prior doesn't feel relevant to me.

I think you're right that I slightly overstated my case.

Comment author: xccf 28 October 2017 01:58:36AM 1 point [-]

[Edit: I appreciate that I should generally behave as though my community will behave well, and as such I should not have requested that people upvote if they find the post helpful. I want to be sure to flag in this response though the incredibly poor way in which people who disagree with claims and arguments in favor of diversity and inclusion are using their votes, in comments and on the whole post.]

Thanks.

I'm also finding the voting in this thread frustrating.

I appreciate your suggestions a lot, but caution you to be careful of your own assumptions. For instance, I never suggested that a Diversity & Inclusion Officer should be the person most passionate about the role instead of most smart about it.

Sorry about that.

To emphasize though, so it doesn't get lost behind those critical thoughts: I thoroughly appreciate the suggestions you've contributed here.

Glad to hear it :)

[Edit: Apologies for some excessive editing. I readily acknowledge that in an already a hostile environment, my initial reaction to criticism regarding an important issue that is causing a lot of harm is too defensive.]

I'm an excessive editor too, I'm not sure it's something you need to apologize for :)

Comment author: Buck 28 October 2017 06:14:45AM 1 point [-]

xccf, I'd be interested to hear an examples of comments which you think were excessively downvoted.

Comment author: xccf 29 October 2017 12:28:50AM *  4 points [-]

If I recall correctly, this comment was at -2 when I first saw it, which frustrated me because I think people who publicly admit mistakes should get upvotes. Publicly admitting mistakes is really hard to do. I think we should take a moment to give people credit for this before demanding that they confess their sins even more thoroughly.