5428Joined Dec 2017


Former AI safety research engineer, now AI governance researcher at OpenAI. Blog:


EA Archives Reading List


I've also downvoted other things that promote one-dimensional tribalist conflict on this issue. In general I don't support escalation to even-more-clearly-defined battle lines; this is precisely how conflicts become worse.

Downvoted because I'd prefer not to turn the situation into a one-dimensional tribal conflict of "support or oppose Bostrom", which this post feeds into. I support you expressing your opinions, though, so would encourage a more descriptive title.

I also think people likely significantly underestimate the costs of running conferences. From what I understand, $2000 per person is a pretty standard number even for a 3-day conference, if hosted in a standard hotel convention center. I'm not claiming that EA should host many conferences which are that expensive - just that buying a property to host them could relatively quickly become cost-effective.

Others have commented that you could get a much cheaper venue. I'm skeptical, though. For external-facing events you want a hotel-quality venue, and I don't expect that getting one with 30 rooms anywhere near a major city is going to be much cheaper.

While wider western societies are often a good baseline to use, I think it's important for EA to try to be more moral than our surroundings. Based on my moral standards, if existing norms punish consenting adults for making certain personal decisions about their private lives, we should try to strongly avoid adopting similar hostility (I think this is a pretty widely-endorsed principle in general, and it's just that others aren't as consistent about it).

That's separate from noticing ways in which higher prevalence of polyamory has flow-through effects on other dynamics (like the ones I identified in my previous comment) and trying to ensure those go well; I think that's a more productive discussion than trying to debate about polyamory as a whole.

Thanks for commenting :) I think the dynamics around polyamory are important to think about in these types of discussions.

My own take: I agree that lots of people being poly makes it harder to identify norm violations, compared with traditional environments, and that this is a significant cost. So when thinking about how to set norms about professional boundaries, we should be aware that the "standard" norms are calibrated for primarily-monogamous environments, and therefore err on the side of being more careful than we otherwise would.

De-normalizing is pretty broad, though, so I'm keen to think more about what this might involve. Things like not assuming people are poly by default definitely seem valuable. On the other hand, I wouldn't endorse "opposing" poly more generally - I think we should be very cautious about passing judgement on people's sexual identities (especially when poly people often face hostility from the rest of society).

Hi Isabel, thanks for commenting. I think your frustration is understandable, but I'm more optimistic than you about the usefulness of collecting experiences and examples in a place where people can easily refer back to them, and giving people a space to contribute their perspectives. It's hard for any one person to have a full picture of the situation across many local communities in many countries; even the couple of examples you gave are useful for me in thinking about which interventions to prioritize. (Of course, to the extent that this has already been done, I'm happy to just fall back on that work rather than reinventing the wheel - please let me know if there are specific resources you think more people should see, and I'll link them in the original post).

I also wanted to note that people who read the EA forum are a relatively small proportion of the people who go to EA meetups, attend EA events, and so on. So one of the intentions of my post was to better understand the best ways for the people reading this to spread the message from your comment more broadly.

Thanks for your work KMF! I'm curious if there are any specific patterns or common stories you've noticed from talking to people from underrepresented groups who are involved in EA?

(Also, FYI, your Magnify Mentoring link is broken.)

I'm not sure if I agree, but if we suppose you're right, then I'm interested in thinking through what the bottlenecks are for that person or team existing (if they don't already), and who might be a good fit for that work.

Thanks Ula! Curious what your ideas about improving EAGs are?

Load More