Hide table of contents

The EA community has both a professional aspect and a social aspect; sometimes, these overlap. In particular, there are networks of EAs in major hubs who variously date each other, are friends with each other, live together, work at the same organizations, and sometimes make grantmaking or funding decisions related to each other. (I’ll call this the ‘work/social overlap’ throughout).

Recently, lots of people have criticised the work/social overlap in EA, blaming it for things like the misogynistic, abusive dynamics described in the recent Time article, or for the FTX collapse. In this post, I  talk a bit about why I’m wary of calls to “deal with” or “address” the work/social overlap in some way (this includes, but isn’t limited to, calls to “deal with” polyamory in EA specifically).

This is partly because I think that the work/social overlap has some strong benefits, which is why it exists, and the benefits outweigh the drawbacks (though we should certainly try to mitigate the drawbacks as best we can). If you want to get rid of something, you should first try to make sure that it’s not importantly load-bearing first; I think in fact that EA work/social overlap probably is importantly load-bearing in some ways. (I never thought that I’d be invoking a Chesterton’s fence argument in support of polyamorously dating all your coworkers :p but there we go).

And it’s partly because I think it’s immoral and harmful to try to prevent people from consensually forming relationships with whom they want, and it’s only a little better if the mechanisms you use to attempt this are supposedly “soft” or “non-coercive”.

In Forum tradition, disclosures and caveats first:

I’m fully embedded in the EA work/social overlap. I work as a freelancer, mostly with EA clients, some of whom I know socially; I live in London and am friends with many other London EAs; and I’m dating two other EAs.

I also want to acknowledge that there are downsides to the work/social overlap - the critics aren’t wrong about their points, they are just missing important other parts of the story. In particular:

1. The work/social overlap gives rise to conflicts-of-interest (and this is more of a problem in highly poly communities, because there are just more conflict-of-interest-ish relationships)

2. The work/social overlap means that people who are engaged with EA professionally, but not part of the social community, may miss out on opportunities. The recent ‘come hang out in the Bay Area’ push seems to be a tacit acknowledgement that this is, in fact, how it works. It’s good that Bay Area EAs tried to mitigate this by inviting non-locals to visit the Bay and get more plugged in to the in-person social community, but it would be better, in my opinion, if organizations in hubs were less clique-y and more willing to hire people who weren’t part of the social community (and indeed, people who don’t identify as EA at all).

3. You’re putting all your eggs in one basket. If you live with EAs, date EAs, are friends with EAs, and work with EAs, it makes the thought of leaving the EA community really hard. This is bad for epistemics: for a while, I found it really hard to think clearly about where I stood on EA philosophical/ideological positions, because I felt so ‘locked in’ to the community in all parts of my life.

(I got past this thanks to a fruitful coaching session with Tee Barnett, where he pointed out that my social relationships formed through EA are now much deeper than our EA connection, and when I actually thought about it, I realised that most of my EA friends wouldn’t ditch me or disown me if I came to disagree with EA, so leaving EA wouldn’t really mean losing access to my social community. So, thanks Tee!)

I would really support efforts to mitigate these downsides to the work/social overlap. But I don’t think the answer is ‘get rid of the work/social overlap’. I think working with one's friends and friendly fellow community members is really motivating and meaningful, and to the extent that EAs do that healthily, it’s actually a real benefit to the community.

People choose whom they date and befriend - no-one is forcing EAs to date each other, live together, or be friends. EAs associate socially because they share values and character traits. If you show me a random EA, I’ll be quite confident (though not 100%) that they’ll be altruistic, interested in discussing unusual ideas, and invested in thinking clearly about deep issues. These things are really important to me, and I really want to date and be friends with people who share these traits. Of course, it’s not only EAs who have these traits (and not all EAs have the traits); but the EA community concentrates this sort of people. I'll probably find more compatible potential friends/partners at an EA gathering than at gathering of randomly-selected people. So I find it unsurprising that many EAs living in hubs have, like me, ended up dating and befriending each other.

Even though the work/social overlap is inevitable, it might still not be good. But I think there are advantages. First, you spend a lot of time at work. Would you rather work with people you like and care about, or people you don’t like and don’t vibe with? Ordinary workplaces do team-building exercises because they recognise that there are productivity and morale benefits when colleagues trust each other, respect each other, and have fun together. If you work with friends - or befriend your coworkers, e.g. by regularly hanging out with them outside of work, or sharing a group house together - you get these benefits for free. 

(Of course, a dark side to this is that people might hire their friends at the expense of better-qualified non-friends. This is a problem, to the extent that it happens). 

I’ve worked with friends, and people I know socially, before. There are some challenges; for example, one person didn’t really like what I wrote for them, and that felt a little awkward, perhaps more awkward because we were friends (they were very nice about it though!) I can also imagine that it might be especially awkward if I had to nag friends to pay me. But there are also benefits: when working with friends, I feel like I can trust them more, I can be more honest with them, and I feel especially motivated to do good work for them. 

I basically think that while the degree of work/social overlap in EA is unusually high by modern Western standards, it’s not unusual historically. Like, I’m guessing that if you were a peasant living in a medieval village, you didn’t have separate ‘work colleagues’ and ‘personal friends’. You’d farm your land alongside your family. You’d bring in the harvest with your fellow villagers, who you’d also go to church with on Sundays, drink with, marry and have children with, etc.

Obviously, ‘default’ doesn’t mean ‘good’. But more importantly, most plausible utopian futures I can imagine also involve little work/social separation. I imagine utopia as a close-knit group of mutually chosen people, who care about each other and enjoy each other's company, spending most of their time together, either working together to do things that the community needs, or not working (because that’s done by robots). Inasmuch as some pockets of the EA community seem to have achieved this, that seems kind of utopian to me.

I also think that power dynamics are the source of the biggest problems in the work/social overlap, so a flatter power structure might be a good way of avoiding some of the pitfalls and abuses of the work/social overlap.

On nudges and norms

Above, I’ve been arguing that there are some advantages to the work/social overlap. I now want to make a slightly different point: even if I thought that the work/social overlap was net bad, I’d feel extremely wary about attempts to intervene to change this. I think that autonomy - particularly in the matter of relationships - is incredibly important. It's bad to infringe on people's free choice to do things that would improve their lives, and saying that you'd do it supposedly "non-coercively" - e.g., using nudges or norms - doesn't make the situation much better for me. 

I'm sometimes frustrated by these conversations, because people are often vague about what changing the culture might entail, concretely - a vagueness that cloaks serious harms, as vague language often does. In an exchange about polyamory in the community, I wrote:

To get a sense of why poly people are upset about this, imagine if someone was like 'there are better outcomes if people are celibate - you save so much time and emotional energy that can be spent on research! So you should break up with your partner'.

…which my interlocutor Jeff Kaufman fairly called out as unreasonably hyperbolic. But what I was getting at was… I don’t see any way you could meaningfully “address” the work/social overlap without trying to get people not to date, live with or befriend people they otherwise would have dated, lived with, or befriended. And if you put it in those terms, it seems messed up, right?

I think most people, in response, would say ‘of course I wouldn’t support telling people to break up with EA partners or never see EA friends! We should just change community norms, or implement nudges to make the work and social spheres less intertwined.’ 

Inasmuch as the punishments meted out by norms or nudges are lesser than punishments meted out by, e.g., systems of rules or laws, they are less bad and less ‘harsh’. But in my opinion, a lot of the harm that comes from (many) laws, rules and norms is just that they prevent people from doing harmless things that they want to do.

For example, there are some drugs, like modafinil or Wellbutrin or MDMA, that I have reason to believe might improve my wellbeing a lot, but which I either can’t legally access at all, or can only legally access if I successfully bypass medical gatekeeping. Why are these laws harmful? Part of it is that dealers might go to prison. If dealing and taking these drugs was not illegal, but there was merely a strong social stigma against it, that harm would be less. But a major reason these laws are harmful is that they work as a deterrant and lots of people don't take the drugs - drugs which would maybe make their lives better! And inasmuch as a social norm against taking these drugs actually worked, and dissuaded some people from taking the drugs who otherwise would have benefited - it would be harmful. ( And inasmuch as it didn't work and didn't meaningfully change people’s behaviour, it would be pointless, of course).

I feel the same way about the idea of ‘nudges’ to make the community have less of a work/social overlap, ‘be less polyamorous’, or anything else that aims to alter consensual social dynamics in the community. The harm isn’t in the harshness or softness of the punishment - it’s friendships nipped in the bud, beautiful relationships that can never even get started. I think people should think harder about the sacrifices that in practice people would have to make, and about how important social relationships are. 





More posts like this

Sorted by Click to highlight new comments since:

I don’t see any way you could meaningfully “address” the work/social overlap without trying to get people not to date, live with or befriend people they otherwise would have dated, lived with, or befriended. And if you put it in those terms, it seems messed up, right?

I don't actually think that's necessarily messed up? That sometimes your role conflicts with a relationship you'd like to have is unfortunate, but not really avoidable:

  • A company telling its managers that they can't date their reports .

  • A person telling their partner that they can't date other people.

  • A person telling their partner that they can't date a specific other person.

  • A school telling professors they can't date their students.

  • A charity telling their donor services staff that they can't date major donors.

The person has the option of giving up their role (the manager and report can work with HR to see if either can change roles to remove the conflict, the poly partner can dump the mono one, etc.) but the role's gatekeeper saying you both can't keep the role and date the person seems fine in many cases?

The harm isn’t in the harshness or softness of the punishment - it’s friendships nipped in the bud, beautiful relationships that can never even get started. I think people should think harder about the sacrifices that in practice people would have to make, and about how important social relationships are.

In agreement there!

Most of my comments have been trying to say "we should evaluate both sides of these tradeoffs". Too much of the discussion has been "X has downsides" or "X has upsides" as if these are decisive.

I also think the word "punishment" in the original post is too loaded. Taking on certain social roles has an effect of limiting the person's freedom in various ways; it disqualifies them from taking certain inconsistent roles. But we wouldn't, for instance, say that a charity is "punishing" its donor services staff by stating that they can't date major donors.

I don't actually think that's necessarily messed up? That sometimes your role conflicts with a relationship you'd like to have is unfortunate, but not really avoidable:

  • A company telling its managers that they can't date their reports .
  • ...
  • A school telling professors they can't date their students.
  • A charity telling their donor services staff that they can't date major donors.


In theory, I think it makes a lot of sense  to have some clear hard lines related to some power dynamics, but even when I'm trying to write those red lines I notice myself writing guidelines because human dynamics are subtle and context specific. For instance:

  • because you shouldn't have power dynamics potentially seep into a romantic relationship or influence your work behaviour: don't date direct reports; and most likely same for anyone in your workplace hierarchy; or where you have any inkling this negative dynamic could arise 
  • Someone at a certain level of seniority and / or power within an organisation will find others feel less able to speak up if their behaviour doesn't chime with them, so you should be extra careful in non-work social settings, especially if there's banter which could be flirtatious or a bit too close to the bone
  • Someone very experienced / well regarded in a company - even if very junior - can wield a lot of power over someone more senior, so they too need to keep themselves in check in terms of how they affect the other person; including how they challenge them

Basically, I don't think there's a feasible checklist for dealing effectively with the range of issues that might occur: affection and loyalty, and power, influence and control are all so subtle. To ensure you're not doing wrong to others or being done wrong to, it's more a constant process of checking in with  yourself and empowering others to speak up.

Moreover, I think it's difficult to separate out what counts as more / less ok workplace relationships. You could say we need to see fewer people working in EA orgs who were friends outside of work (as opposed to friends you make at work) or romantic relationships starting in EA orgs, but then there's just the people who you get along with and see eye-to-eye with within the work place and sometimes develop more impactful and / or toxic relationships with. For example: 

  • having political loyalty to a colleague, leading to factions 
  • nepotism between close friends / former work colleagues
  • simply the more junior person feels their career is still dependent on their mentor / friend[1]

Which is to say implement all these rules targeting 'out of work' friendships / relationships, but human power dynamic issues will still prevail in adjacent domains.

Reflecting that tighter human alliances and power dynamics is somewhat inevitable (I postulate), it's worth noting that a lot of the time companies - big or small - deal with relationships in flexible ways, and sometimes this is worth considering. Things like: 

  • if two people want to start dating - whether they're in the same team or in a line management hierarchy - give them the option of changing teams where that's appropriate for the business and still maintains sufficient separation to avoid power imbalances / 
  • in some cases, just turning a blind eye because people have their shit together and it isn't interfering in their work-lives
  • stipulating that a married couple cannot be in the same senior leadership team (as they'll be a unit)



  1. ^

    Side note: most of the above examples are why I'm often banging on about diversity within organisations - just newbies full stop - because they can break much of these dynamics up both through behaviours and new ideas, but I won't get on that hobby-horse just now.

I do agree that it's reasonable/inevitable that sometimes roles conflict with who you want to date. But in all your examples, I wouldn't necessarily frame this as 'telling someone who they couldn't date' (more like saying 'you can't date x while one or both if you is in y circumstance'). Like, if I ran an organization and a manager came to me and said 'uh, my report and I have kind of fallen for each other, and we want to date', I wouldn't be like 'well you can't date them', I'd be like 'congratulations! But yeah, you probably shouldn't manage them anymore - I'll find them another manager'. When potential romantic relationships arise in workplace settings where there's a power dynamic, I think the best move is to let the relationship play out and move around the working relationships so there's no longer a power dynamic between the two. The reason I think this is that romantic relationships are very precious for people, and not that easy to find, whereas manager/report relationships (or professor/student or whatever) are generally less meaningful and require less compatibility, so it makes sense to prioritize the romantic relationship over the professional one.

 Similarly, if a monogamous person says 'I won't date you if you date other people', that seems like them (reasonably) expressing a condition on who they will date - similar to if they said 'I won't date you if you eat meat/are a social conservative/are a smoker/want kids'. This feels different to external people trying to stop me from dating another person who wants to date me. 

Maybe a lot of this is semantic, but the substantive thing might be 'in cases where there are clashing social and work relationships, I'd be in favour of prioritizing the social one over the work one, and in workplaces accommodating social relationships that arise within in them, rather than trying to prevent them from happening' - similarly to how good workplaces should accommodate people having kids (by e.g. offering parental leave), rather than telling them to just quit if they want a kid. 

whereas manager/report relationships (or professor/student or whatever) are generally less meaningful and require less compatibility, so it makes sense to prioritize the romantic relationship over the professional one.

While I do think this is generally true for managers at large organizations, there's also an issue where there could have been some amount of abuse of power around the manager and report getting together. I think that's even clearer in the professor and student case, where I would be extremely surprised to see a "congratulations" from the Dean.

You're right that in some circumstances people would choose to handle a conflict by giving up their existing role, and "you can't date X" phrasing from role gatekeepers assumes that someone strongly values their specific existing role. But I do think it's reasonably common that people do value their existing roles more strongly, especially when we're talking about casual dating and not "I think I've found my life partner". So I do think this is responsive to the point you made in your original post, which is that having norms against combining certain roles with certain relationships does have consequences in terms of people not getting to enjoy some otherwise positive and fulfilling relationships.

congratulations! But yeah, you probably shouldn't manage them anymore - I'll find them another manager

In lots of small orgs this is tremendously costly or impossible. It might be a nonstarter at a lot of large orgs too, I just wouldn't know about it because I've never worked at a large org. 

I think the word "probably" in this quotation is quite concerning - you should 100%, definitely, in every case and without question not let someone manage someone they are dating. It's an unresolvable conflict of interest and totally unprofessional.

But also, to Quinn's point, if it's a small org, even making this change might not really mitigate the problem. Imagine a 5 person team, where the CEO and one of the staff are dating, so then you change the reporting line for the junior person in the relationship. It seems highly probable that the new manager is going to be influenced by the fact that their boss is dating their subordinate.

I'm going to write a longer comment on how I think you can manage this below.

Thanks for writing this Amber. I pretty firmly disagree, but I'm upvoting it anyway because I think we need to discuss these issues in the open, and you've put across your point of view in a measured, reasonable way. I hope to draft a response soon with some alternative suggestions.

(Also, as usual, @Peter Wildeford has made most of my points in his comment.)

I think my main disagreement is that this is taking on a straw man argument. I'm not aware of anyone suggesting that we should "prevent people from forming relationships with whom they want", at least not in the strong sense of "banning" relationships. Indeed, I've spoken recently to a couple of people with experience of managing these issues, and they all say "people are free to date whomever they want" (provided the usual caveats about it being fully consensual etc.).

What is being suggested is that there should be clear policies for how these relationships will be handled, to ensure a) the safety of those involved, especially the partner(s) with less power; and b) the safety of the community as whole. Those then give all parties the chance to make informed decisions about whom they form relationships with, and to give informed consent.

Accordingly, I think all EA orgs should adopt a clear "relationships at work" policy (and I hope to release a template for this soon). This wouldn't say "you can't date anyone from work". But it would say:

1. If you have a nonprofessional relationship with someone at/directly related to work, we will actively manage any conflicts of interest, by doing X, Y and Z
2. If you have a nonprofessional relationship with someone at/directly related to work, you have to tell the relevant organisation(s) promptly so they can achieve (1)
3. You can't use work time to advance your romantic interests, in particular by e.g. propositioning people
4. If you date someone at/related to work and there is a power imbalance, you have a particular duty to think very carefully about this, per @Peter Wildeford 's comment
5. Even when you aren't on work time, we reserve the right for your actions to have professional consequences if they are plausibly harmful (e.g. if someone comes into work on a Monday and says you sexually harassed them on a Friday night, that's not 'off limits' for professional consequences) 

Part 1 would involve things like recusing anyone who has a nonprofessional relationship with someone else from any decision about their pay, promotions, disciplinary processes etc., to avoid the obvious conflict of interest here (I commented elsewhere that I am concerned by you saying you would only "probably" find someone a new manager if their current manager starts dating them, although I don't want to focus in too hard on a single word).

It also involves removing them from decisions about funding, hiring etc. for indirect work relationships.

Part 5 might seem like the most draconian but is based on real examples of people who repeatedly sexually harassed colleagues, but because it happened 'after hours' it was never acknowledged or dealt with.

If orgs adopted a policy like this from the get-go, it would give everyone involved the chance to give informed consent - that is, they can understand exactly the consequences of their actions in advance and decide what works best for them. This isn't "preventing people from forming relationships" - it's allowing them to form relationships in an informed way.

You are right, of course, that this might lead to people not starting a relationship that they otherwise would have, but that's just tough, I'm afraid - as other commentators have pointed out, this happens all the time in professional settings. Alternatively, they might choose to start the relationship anyway, and then they can choose whether to live with how that is managed or one or other person can seek a job elsewhere, in a different team etc..

Anyway, in sum, I think this is the steelman position, and I think its harms (maybe people have fewer relationships/less sex with people they work with) are greatly outweighed by its benefits (a healthier and safer workplace and community, which conforms to wider workplace norms).

Thanks Jack, I found this to be a very pragmatic and fair approach. Maybe the community health team could develop a suggested policy document for personal relationships within the professional context using this as a starting basis.

And some version of point 5 is the employer's legal obligation, at least under some circumstances in the US. E.g., this discussion:


Yeah this is basically what I was trying to say in my comment.

Your answer here may just be that it is organization specific, but something I'm keenly interested in is what the "X, Y and Z" from point 1 might look like. Questions come up here of a sort of federal vs state way of legislating how to navigate these, where a problem I see with your potential proposition is that it leaves the decisions in the hands of those who direct these various companies, potentially leading EA to (what I perceive to be) the predominant method of "X, Y and Z" being "romantic relationships are not allowed with colleagues". But I think the spirit of your comment is otherwise equally amenable to a process I think more, which is the federal equivalent here of discussing the issue as a community and putting forth ideals based on that discussion for entities relating to EA, guiding principles, if you will, that can be deviated from but that set a standard. This also carries the benefit of helping us to create standards for non-work contexts like EAGs.

I see the next, most productive, outcropping of your comment to be figuring out what "X, Y and Z" should be, and not being experienced in this, would love to hear your thoughts about what might be reasonable. 

Thanks Tristan. I outlined these briefly above but I think they are things like:

  • anyone in a relationship with anyone else is recused from all professional decision-making affecting that person. They can't hire or fire them, they can't conduct performance reviews, they can't promote them, they can't set their pay. They definitely shouldn't be the decision-maker on these things, but ideally shouldn't have input either - they just are not able to be impartial, and any process they fed into could easily be challenged as unfair (by their partner if they don't get the outcome they want; or by someone else e.g. Person A got promoted and I didn't, and it's because Person A is sleeping with the CEO)
  • If a management relationship exists, it will be changed, so that no one is managed by their partner
  • If a funding relationship exists (e.g. a grantmaker is sleeping with their grantee), it will be changed, and that person/org's applications will be assessed by someone else

Finally, I think there are some relationships that are highly likely to be de facto inappropriate (e.g. if I as CEO started a sexual relationship with a OFTW student volunteer), in which case X, Y and Z would include an investigation and disciplinary action. However, I am a little less certain of this, as I can see the 'but if they are both consenting adults...' line of argument.

Personally, I would definitely choose not to have any romantic involvement with a student, but I could imagine circumstances where this might happen with someone else and it might not be considered wrong (or not considered wrong by everybody). But in general I'm pretty unsympathetic to powerful people who sleep with people over whom they have power, so I tend to take a dim view of this sort of thing, and I'd be fairly comfortable saying this was inappropriate at OFTW.

I would like to emphasize that when we discuss community norms in EA, we should remember the ultimate goal of this community is to improve the world / humanity's future as much as possible, not to make our lives as enjoyable as possible. Increasing the wellbeing of EAs is instrumentally useful for increased productivity and attracting more people to make sacrifices like "donate tens of thousands of dollars" or "change your career plan to work on this problem", but ultimately the point isn't to create a jolly in-group of ambitious nerds. For example, if the meshing of polyamorous and professional relationships causes less qualified candidates to earn positions in EA organizations, this may be net negative, even if the polyamorous relationships make people really happy.

I think it is possible that things that make the community happy will end up being net negative for the world. But I do think that creating a happy, thriving social community, that people feel comfortable in, is going to be really important for the longterm success of this movement (as you acknowledge). And there's a kind of tricky thing where... like, if I felt like 'oh, the movement tolerates you having polyamorous relationships now, but if we decided one day that this had net negative consequences, we'd shun you' - then I'd feel way less good in the social community now, because my acceptance wouldnt feel secure. I think people need to feel safe and like their acceptance is "unconditional", rather than feeling like if their presence is no longer deemed to be net positive for the world, they'll be rejected from their social network [cf "I didn't get into EAG and am sad" discourse] . 

I put "unconditional" in inverted commas because obviously some conditions are always present and appropriate - if I went on a murdering spree (or committed a billion-dollar fraud :p) it would be reasonable for the community to shun me. But I think this bar should be pretty high, because the costs are bigger than the obvious costs to the people being shunned.

I would say the goal of EA is twofold:

  • Improve the lot of humanity as much as we can given the resources we can gather and are willing to put in,
  • Gather and put in a lot of resources.

In particular, pretty much no-one puts in all their resources, or as much as they can possibly afford. Most EAs are not willing to entirely forego being really happy themselves in the pursuit of a better world.

(There are instrumental reasons why being infinitely self-sacrificing is a bad idea anyway, but even if there weren't, most people just aren't that hardcore about their utilitarianism.)

I think I've recently got a reputation as one of the people really against personal and professional mixing, so I wanted to clarify more about how I think about this.

I definitely think it's ok (beautiful even) for EAs to date other EAs. There's nothing wrong with this. I think where it goes wrong is how it is handled. It needs to be handled with a lot of thought and care.

Essentially, a few things:

  • I think we as a community need to make clearer differentiations between personal and professional spaces and then avoid asking people out in professional spaces. EA Global is a professional space and I'm glad we now have a "Swapcard is not for dating" rule. Other events, like EA Global afterparties, that feel more ambiguous should probably do more to clearly set expectations upfront so attendees know what they're getting into.

  • We shouldn't be able to make decisions of power (e.g., grants, hiring decisions, management, performance reviews) to people we are dating. That is a clear conflict of interest. Any grantmaking body or organization has to have some sort of system to handle this. For example, if you date someone on the EA Infrastructure Fund, you still should apply, but someone else other than the person you are dating should review the application and your relationship to the grantee should be disclosed confidentially to the Fund when relevant.

  • EAs who are in positions of power (e.g., CEOs of research organizations, grantmakers at OP) need to be extra careful and extra thoughtful about how they go about their relationships.

  • People should be able to engage fully within the professional aspects of EA without having to give up their personal life too. EA needs to be structured to allow this. I think we've made a lot of progress on this.

All of these things apply equally well to any community or professional group, not just EA. There's nothing EA-specific about how to navigate conflicts of interest.

I think "Power dynamics in EA" was a really thoughtful post that captures a lot of the relevant trade-offs here.

This all seems right! 

Re differentiating social from professional spaces, do you think (unofficial, non-CEA-sponsored) dating events around the time of EAGs is a step in the right or the wrong direction? 

Arguments that it's unhelpful: it sets an expectation that you can/should be viewing the conference and adjacent socials as a way to find partners, which isn't true of the conference proper; you may encounter people you've had professional 1-on-1s with at the dating event which may make things blurry/awkward. 

Arguments that it's helpful: it quarantines all the flirty energy in one place so people feel "less of a need" to hit on each other at the conferences and general afterparties; by making that space be explicitly social, it helps the conference itself be more implicitly professional? 


I do personally think it would be helpful to have explicit dating events, so that people who are in to that thing have an explicit place to be and go and people who are not in to that thing can know to steer clear.

I personally think that people are personal and professional people, so it shouldn't be weird or bad to see people in personal and professional contexts and to interact with them very differently depending on the context.

What exactly does "EA Global is a professional space" entail for you? Professional evokes a certain image for me that might not be the same for you, which is why I ask. In my mind, "professional space" implies a place where sidebar or causal conversation is quite limited, where interaction amongst people comes in the form of conferences instead of conversations between friends, the type of space where I might get in trouble for talking to someone for too long about League of Legends or some other random interest of mine rather than my effective way of improving the world. If this is your proposition, I think I stand in stark opposition, but I imagine you might have a different idea in mind, one I'd like to hear more about. 

I kind of skimmed this post, so hopefully I am not making of fool pf myself but I think you didn’t really address a key point which is raised by „critics“ and that are the challenges associated with the tendency for centralization in EA.

There are basically two to three handful of people who control massive amounts of wealth, many of which are interweaved in a web of difficult to untangle relationships ranging from friendly to romantic. The denser this web is, the more difficult it is for people to understand what is going on. Are rejections or grants based on emotion or merit? It’s simply difficult to say the more complex the interactions are.

I think having friendships and relationships is great if the people involved are happy but we have to develop appropriate means for dealing with the complexity of it all. For instance, in cooperatives relationships are often very highly valued and central to the whole experience of being part of the cooperative. There are formalized mechanisms in place to afford systematic discussion of relationships and negotiation of mutually acceptable forms of organization. In EA, we are lacking this kind of structure. While some participatory islands might exist, there are often streamlined but opaque processes in place that allow a few people to make huge decisions affecting countless people with very limited involvement from the community at large (or the people affected for that matter). This becomes pretty tricky to justify as “EA” if you cannot demonstrate that the decisions being made are “above reproach” and not influenced by romantic relationships, in-group favoritism or the like.

In sum, I think I broadly agree that having friendly or even romantic relationships within the EA community can also have a good side but I am very skeptical that our current ways of organizing can handle all the complexity that is entailed by strong versions of this. If we want deeper and more relationships within the community, we should adapt our spaces and institutions to be ready for that. We owe it to ourselves and others to figure out how we can behave responsibly in this context.

Yeah, this seems very reasonable. I'd be in favour of less centralization and more transparency. It does seem like there are issues where grantmakers have to decide about whether to give a grant to present or former partners or metamours, or close friends. Maybe there could be a system where people's grant proposals must always be assessed by someone who doesn't live in the same hub as them (if they live in a hub). 

I'm not really sure this particularly applies here though. I think power being concentrated is not a runoff effect of their being dense relationship clusters within EA, and instead that not having systems of diffuse decision making just is the key problem for the issue of centrality that you mention. Sure, there are probably some cases where friendships have allowed people to bypass more formal and open means of communicating about decisions between orgs, but I still think the effect that has on perpetuating this system is minimal at best. 

But to run with your comment a bit further, what do you think might be the best way to solve the centrality issue of EA?

People choose whom they date and befriend - no-one is forcing EAs to date each other, live together, or be friends. EAs associate socially because they share values and character traits.

To an extent, but this doesn't engage with the second counterpoint you mentioned: 

2. The work/social overlap means that people who are engaged with EA professionally, but not part of the social community, may miss out on opportunities.

I think it would be more accurate to say that, there are subtle pressures that do heavily encourage EAs to date each other, live together, and be friends (I removed the word 'force' because 'force' feels a bit strong here). For example, as you mentioned, people working/wanting to work in AI safety are aware that moving to the Bay Area will open up opportunities. Some of these opportunities are quite likely to come from living in an EA house, socialising with other EAs, and, in some cases, dating other EAs. For many people in the community, this creates 'invisible glass ceilings,' as Sonia Joseph put it. For example, a woman is likely to be more put-off by the prospect of living in an EA house with 9 men than another man would be (and for good reasons, as we saw in the Times article). It is not necessarily the case that everyone's preference is living in an EA house, but that some people feel they will miss opportunities if they don't.  Likewise, this creates barriers for people who, for religious/cultural reasons, can't or don't want to have roommates who aren't the same gender, people who struggle with social anxiety/sensory overload, or people who just don't want to share a big house with people that they also work and socialise with. 

If you're going to talk about the benefits of these practices, you also need to engage with the downfalls that affect people who, for whatever reason, choose not to become a part of the tight-knit community. I think this will disproportionately be people who don't look like the existing community. 

I don't know if the framing of it "creating barriers" completely captures the dynamic. I would suggest that there is already a barrier (opportunities to exchange ideas/network with like-minded people) and the main effect of starting a group house is to lower these barriers for the people who end up joining these and then maybe there is a secondary effect where some of these people might be less accessible than they would be otherwise since they have a lower need for connecting with outside people, however, this seems like a secondary effect. And I guess I see conflating the two as increasing the chance that people talk past each other.

I think this is a reasonable concern (as someone who would avoid moving into a big group house like the plague :p). I'd be in favour of more blinding when people make hiring decisions. Hiring agencies, as well as saving people time here, might also make the process fairer, since they can be more objective and will be less tempted to hire friends.

As an empirical matter, do you think people in EA do disproportionately hire friends, or does the causation go the other way? (e.g., people move into group houses with friendly colleagues). 

I appreciate this post. But one mistake this post makes, which I think is an extremely common mistake, is assuming that there can exist a community without (soft) norms.

Every community has norms. It is impossible to get out of having norms. And so I don't think we should be averse to trying to consciously choose them.

For example, in American society it is a norm to eat meat. Sometimes this is in fact because people actively are trying to get you to eat meat. But mostly, nobody is telling other people what to eat  -- people are allowed to exercise their free choice (though animals aren't). But this norm, while freeing for some, is constricting for others. If I go to a restaurant in many places, there won't be much good vegetarian food. In some places, there is a norm to have vegetarian food. But there is no place where there is no norm: in some places, there is a norm to have it, and in others, there is a norm not to have it. The norms can be stronger or weaker but there is no place without norms.

Currently, there is the non-coercive but "soft" norm in EA that young people interested in AI safety research will go to Berkeley. The post you link is an example of that. People are being actively encouraged to go to Berkeley. They are being paid specifically to go to Berkeley in some cases. For the reasons you give, this could potentially be really good, but the comments on that post also give reasons why it might not be!

You gave the following reason why norms are often not so good:

they prevent people from doing harmless things that they want to do.

This is true. But one could just as easily say of other norms:

they encourage people to do slightly harmful things they wouldn't otherwise want to do.

Or perhaps:

they fail to discourage people from doing harmful things that they want to do.

The "default norm" is what the community happened to settle on. But it is a norm as much as any other. And it isn't necessarily the best one.

I'm not sure Amber is saying that a community should have no norms at all? It sounded more like she was saying that in the domain of personal relationships there shouldn't be norms.

One could argue, of course, that there are implicit norms around this already in the community, so maybe the argument I think is stronger is something like:

"There are implicit norms in the community around personal relationships already - we should evaluate this norms and figure out if we think they are actually good or not"

(I think this might be a fruitful exercise that it think many individual organizations are doing now as a result of recent events, but is harder to do for the more nebulous "community" and network of people)

Yes, I think it's impossible not to have norms about personal relationships (or really, anything socially important). I should perhaps have provided an example of this. Here is one:

If you move to a new place with a lot of EAs, you will likely at some point be asked if you want to live in a large group house with other EAs. These group houses are a norm, and a lot of people live  in them. This is not a norm outside of EA (though it is in maybe some other communities), so it's certainly a positive norm that has been created.

Even if EAs tended to live overwhelmingly in smaller houses, or lived with people who weren't EAs, then that would just be another norm. So I really don't think there is a way to escape norms.

This is a fair point. I think there are maybe two different meanings of norms at play that might be useful to disambiguate:

(1) what's normal in a community, in the sense of 'what most people do'
(2) what's expected, approved of, recommended in the community 

(1) can bleed into (2), because if you are the odd one out, you might feel like an outsider, even if no-one is actively expressing disapproval of what you're doing. Vegans in an majority-omnivore space, or omnivores in a majority-vegan space, might feel kind of awkward, even if no-one criticizes or remarks on their dietary choices. Similarly, I've heard some people say they felt ambient social pressure to be poly in the Bay Area just because loads of other people were, or because people assumed it of them, etc. 

I think what I'm against is not norms existing, but people trying to intervene in the norms 'top down', as it were, by talking about what the norms should be. I think the correct way to contribute to community norms is just by "being the change you want to see". So if any individual EA wants the community norms to be less overlap-y and/or less polyamorous, what they should do is not date multiple people, and not date other EAs. But it's not legitimate for them to tell other people what to do. 


It's worth noting that some people (like me)! don't join the in person EA community at all because the norms are so off-putting.

I expect they are particularly off-putting to monogamous mid career/late career professionals. And it seems to me that it would be good for EA if they had more experienced people around - stuff around risk management would probably improve.

Hard agree with this. It's why I think healthy handling of relationships at work is so important as a stepping stone to a healthy community overall.

Out of interest, which norms are off-putting to you? And which ones might be off-putting to monogamous mid/late career professionals? 

I guess I'm in favour of a meta-norm of openness/permissiveness, meaning that a diverse range of people all feel comfortable in the community. 

What does "more experienced" mean here? Finding it hard to read in a way that is not "EA is comprised too much of younger people who simply can't ascertain the wisdom of the well aged sages". 

I also think that power dynamics are the source of the biggest problems in the work/social overlap, so a flatter power structure might be a good way of avoiding some of the pitfalls and abuses of the work/social overlap.


Do you think that in abstract that professional/social overlap is less of a problem when the power structure is flatter, or that having a flatter power structure is something that EA could actually achieve?

I'm curious because, to deal with potential abuse of power, I would prefer a more explicit power structure (which sounds like an opposite conclusion to your suggestion).

My first assumption is that power structures are an unavoidable fact in any group of people. I assume that trying to enact a flatter power structure might actually cash out as pretending the power structure doesn't exist [this might be where we disagree!].

Pretending that power structures are flat leads to plausibly permissable abuse of the actual underlying power structure. However strictly acknowledging a power structure means one is forced to acknowledge the power dynamic.

So to encourage healthy relationships, I would have called for making power structures explicit, in EA or any group.

Strongly agree with this. 

"the idea becomes a smokescreen for the strong or the lucky to establish unquestioned hegemony over others. This hegemony can be so easily established because the idea of "structurelessness" does not prevent the formation of informal structures, only formal ones."

The Tyranny of Structurelessnes

Additionally what might work well for a group of 10  may quickly become impractical when the group scales.

(replying to both skyblue20 and Jamie)

I do think that professional/social overlap is less of a problem when the power structure is flatter. I agree that informal power structures can arise, but I don't think formal power structures help with this (by making it more explicit), because often the formal structure is a different thing to the informal structure. E.g., you can imagine a person who has various managers and superiors at work,  but also feels less powerful or lower-status relative to colleagues who are nominally on the same level, or friends that they don't work with, because (e.g.) they've been in the community longer, or they're twitter-famous, or they're just more socially dominant, or whatever reason. So I do think getting rid of formal power structures would mitigate the problems, because it would get rid of one avenue for abuse and complication (even though informal power structures would mean that there still was some potential for abuse).

As for whether it could work in EA - I'm not sure, but I think other movements and organizations have experimented with flatter power structures. I think the EA community might be a good place to experiment with this, both because EAs are generally open to experimentation/doing things a bit differently, and because I don't think the average EA has a strong will to power for power's sake (like there's limited macho posturing, for example). 

If you want to get rid of something, you should first try to make sure that it’s not importantly load-bearing first;


I'd be interested in trying to figure this out (maybe through a survey).

What value do different groups in the community get from various kinds of experiences in EA spaces?

For example, I'm curious how most women would weigh being in an EA space (including EAGs, local EA events, EA houses, EA camps and training events etc)

a) that lets them access healthy professional networks free from the tensions of inappropriate* sexual/romantic advances


b) being in an EA space where they are able to date EAs.

(There is a tradeoff here.)

I am also curious if men in the community have an opposing view - if so, it might be important to think about how  the existing state of the community  (that may have been shaped by the views of the majority gender) may make it less attractive to women currently in or considering joining the community.

(This example is obviously hypothetical but I do have a suspicion that a greater percentage of women would weigh a) higher than b) )

Similarly, I am also curious about how different locations/EA organizations/cause area sub groups weigh these choices.

*example of inappropriate - young EA job seekers being propositioned by potential bosses/seniors in their field after making it clear to them that they were looking for job opportunities/contacts/mentoring in that field.


[Edit: This project has been passed on to the CEA Community Heath Team, and folded in to a research project which will gather and analyze data from many other sources in addition to a survey. I've passed on your comment to Catherine Low who is orchestrating the major investigation. You can read her announcement here, posted Feb 14th: https://forum.effectivealtruism.org/posts/mEkRrDweNSdNdrmvx/plans-for-investigating-and-improving-the-experience-of ]

I (and other key players) are working on a gender experiences in EA survey, and I will add this to it, thanks!

Yeah, that would be interesting. I'm not sure there inherently has to be a tradeoff between 'being able to date EAs' and 'having access to work networks free of professional advances.' Granted the situation you mentioned is inappropriate. But there are lots of different fields in EA, and lots of people who are at similar power levels. If a young AI safety researcher and a young animal-welfare person meet at their university group and start dating, that's 'dating within EA' but it isn't, in my opinion, at all inappropriate, and the risks are pretty low. 

I agree that two young (low seniority) EAs from different fields dating is low risk to them. It avoids issues created by power imbalances. I don't see issues in such propositioning, especially in more casual/social settings.

If two senior EAs from different fields date, the risk of harassment stemming from power imbalances is much lower than the senior and junior EA from same field scenario. It could be viewed as problematic from 'a gatekeeping/preferential treatment at the expense of other EAs' perspective. But people could argue this scenario still has more rewards than risks.

If we find that there is a way for people in the low risk scenarios to date each other, while at the same time assuring against the high risk one, I'd be on board with it. My main issue is if our answer to this is to say 'let the current norms/policies be' because we can  defer to the goodness or good judgement of high seniority EAs to act appropriately.

I do believe that the majority of them might even behave ethically, but policies and systems are usually not created with them in mind. They are in place so that the minority of people who engage in bad behavior are not attracted to our spaces knowing that it's easy for them to slip through the cracks here.

I think it's worth keeping in mind that every functioning culture includes a vast list of taboos which people enforce on others. In both regular Western culture, and EA culture, all of the following are considered inappropriate:

- Making sexist jokes
- Not being sympathetic when someone else experiences a personal tragedy
- Nuidity in the workplace
- Answering somebody else's phone
- Smoking weed in the office
- Bragging about money
- Chewing with your mouth open
- Refusing to shake somebody's hand when you're introduced
- Making prolonged, intense eye contact with people
- Not saying "pardon me" after farting or burping loudly
- Over-sharing personal information with someone you don't know very well
- Asking an employee to carry out a duty that is far outside their normal role (e.g. "For the second half of your shift, can you please go to my parent's house and change the bedsheets")
- Giving people unsolicited gifts which are too expensive
- Swearing 
- Having sex with somebody much younger than you even if they're above legal age of consent
- Offering to buy somebody's pet from them

Some of these taboos are hard to justify on the spot, but very few people seriously want to eliminate them all.

Perhaps we want EA culture to lean slightly more towards personal autonomy than normal Western culture does, but there's never going to be a cohesive collection of humans who can productively interact together without  seriously regulating each other's behaviour.

In terms of what's considered appropriate in "regular western culture," a lot of this is not true enough to justify the generalizations you're making:

  • There are variations within cultures in any country, never mind between western, and all other countries, whereby the extent to which crude sexism is considered appropriate. I've met many men from some different walks of life just in Canada whose sense of what's normal is such that they'll look down on other men who don't tow the line with their chauvinistic attitudes and misogynistic comments.

  • While it's far from being all of them, there are a lot of sections of the upper class where bragging about how much money one makes is considered respectable, and this influences other aspects of culture too, especially in North America.

  • Making intense eye contact during a normal interaction is considered inappropriate in most cultures, though there is relative nuance here. Spending a longer amount of time making direct eye contact as part of a back-and-forth in conversation is much more accepted in western cultures, for example, compared to in Russia or China, to the point that to avoid too much making eye contact during conversation in western cultures is often considered rude.

This is consistent with the point I'm trying to make - all human interactions in all contexts are happening within a super complex web of norms and taboos, and any proposal as simple as "just let people do whatever they want" is a non-starter

So first, I do in fact want EA culture to lean substantially more towards personal autonomy than Western culture does - I like autonomy a lot!

I think some regulation is inevitable as people interact. For example, if I go around calling everyone names and never showering, people might be like "Amber is smelly and rude, so I'm not going to invite her to my parties". And then, if I noticed people never invited me to parties, I might be like 'huh! maybe I should shower, and be less rude'. So in a way, people there are 'controlling' and 'regulating' my behaviour, but that doesn't seem overly coercive.

What I object to is if one person thinks I'm smelly and rude and is like "no-one else should invite Amber to parties". This seems to me to be illegitimately hijacking the norm-creation process. Like ideally norms should arise out of the majority preferences of the group; it's bad if a minority decide This Is How It's Going To Be and the rest have to conform. 


I agree that if all the concerns about relationship norms in EA culture were comming from small minority, then this would not justify changing them (but the minority are still entitled to try and advocate/persuade). 

But when it comes to culture/status, I think the dynamics mean majority rule is pretty much baked in by default anyway! So we might not have to worry much about that.

Interestingly, we might have different impressions about what the median attitude is in the community when it comes to questions like 

- "Is it a bad idea to have sex with your manager?"
- "Is it a red flag when a local university group organiser is regularly hooking up with newcommers?"
- "Do long term, stable, committed relationships generally lead to better community health overall?"

Maybe this depends on how big/small you draw the boundary for who counts as part of the EA community, but most people I know who engage with EA would answer "yes" to those questions. (I'm in Australia, perhaps Bay Area is very different)

Thanks for the post - I really appreciated hearing a perspective in favour of this overlap! 

I just wanted to elaborate a bit on how "conflict of interest", "missing opportunities" and "all eggs in one basket" can come together to make the experience especially harmful.

In particular, if you face interpersonal issues (or worse e.g. harassment etc.) from someone who is in the same social graph (professionally and socially) as you, you're put in a very tricky spot. As you acknowledge, it is in fact how it works that if you aren't plugged into the social community, you will miss out on opportunities. It is also the case that many of your friends will be in the same social graph.

Thus you may be put in a position where you have to trade-off having less impact and not attending social events which all of your friends attend, versus your personal happiness and safety. While I think this possibility is very much implicit in your comments, I thought it might be useful to highlight it, because I think in practice this pressure to associate socially is very strong and therefore can be very damaging in the especially egregious cases of interpersonal issues within the community.

Well written, thanks!

If the head of the NIH had had a romantic relationship with the head of the Wuhan Institute of Virology, would that have been reasonable?

People choose whom they date and befriend - no-one is forcing EAs to date each other, live together, or be friends.

I don't think the issue is once of explicit 'force'. It is one of implicit expectation, social pressure from 'soft norms'.


Ordinary workplaces do team-building exercises because they recognise that there are productivity and morale benefits when colleagues trust each other, respect each other, and have fun together. If you work with friends - or befriend your coworkers, e.g. by regularly hanging out with them outside of work, or sharing a group house together - you get these benefits for free.

IMO there is a lot of difference between team building exercises vs encouraging (including 'soft norm' pressure) living with and dating your co-workers.

There is a lot more exposure to risks, especially to vulnerable groups from the latter.

Ordinary workplaces err on the side of workplace romance policies that limit risks to their employees and the reputation of the workplace. I don't think we can reasonable argue that EA organizations/workplaces/professional spaces are immune from these risks. 

'Ordinary workplaces' have learnt how some behaviors increase discrimination and harm towards vulnerable groups the hard way (after decades of harmful behavior towards vulnerable groups). For example, going to a strip club is no longer an acceptable form of team building exercise. I believe there is a lot to be learnt from conventional workplaces that have spent a very long time harming these groups and limiting their access to workplaces. The last thing we would want to do is repeat these mistakes because of biased beliefs that we are too good or too smart to make them.

I guess I'm sceptical that there is really "soft norm" pressure to date one's co-workers (except inasmuch as there is less of a norm not to date one's co-workers, but that's different). Like, do people really think (implicitly or explicitly) "I'll do better at work if I date this person, so I'm going to date them (whether or not I like them)"? This seems very weird to me.  I'd find it really hard to date someone for a long time if I wasn't actually into them. 

Why not document the relationships between people in EA orgs? That might at least clarify things for people who are less central.

You can also document potential conflicts of interest (e.g. where someone hires someone they are in a relationship with), and write down the justifications for doing so. 

Why not document the relationships between people in EA orgs?

It's very common for organizations to have a conflict of interest policy where people within the org are required to disclose conflicts. For example, here's the top search result for "givewell conflict of interest policy" which includes disclosure, and I see it in the one ACE has on their site as well. [1]

Unless you're proposing that these be disclosed publicly, instead of internally? I doubt most prospective employees of these orgs would be comfortable with that level of visibility into their private lives, so you'd lose out on most people who would do a good job at the org.

[1] I don't see public ones for the other EA orgs I tried, which isn't great. I do think they should be making their COI policies public.

[EDIT: wrote something longer about how I'd like to see this.]

I'm trying to decide if I think there is a class of COIs that can be waived only publicly, or if I think that any COI of that magnitude is just non-waivable.

While I wouldn't suggest a general publication of relationships, questions from the public about whether e.g., Employee X recuses from grant decisions about Person Y would sometimes seem appropriate (although they should be generally asked in a private manner). Although the organization doesn't need to explain the basic for a recusal, the answer still provides some information about Employee X.

I like GiveWell's Relationship Disclosures page:

On this page, we disclose the following sorts of relationships:

  1. Social, romantic, family or financial relationships between (a) GiveWell staff and (b) people at our top charities, former top charities, former standout charities, or other grant recipients.

  2. Romantic, family or financial relationships between (a) GiveWell Board members and (b) GiveWell staff.

  3. Social relationships between (a) GiveWell Board members and (b) GiveWell's Chief Executive Officer that pre-dated GiveWell-based relationships between same. (For example, people who were friends with GiveWell's Chief Executive Officer prior to joining GiveWell's Board.)

  4. Other relationships we feel are worth disclosing.

As with most things around transparency, though, we should probably view GiveWell's approach as being towards the upper end of where we could get the broader EA community to.

Curated and popular this week
Relevant opportunities