Mjreard
Bio
Thinking, writing, and tweeting from Berkeley California. Previously, I ran programs at the Institute for Law & AI, worked on the one-on-one advising team at 80,000 Hours in London and as a patent litigator at Sidley Austin in Chicago.
Posts 10
Comments122
We might just fully agree. I don't think there were ever career-long professional benefits to EA for people specializing in specific cause areas that outweigh those at cause-specific conferences (but please come teach community builders/members/young people about your work at EAG).
I think EA has always been for:
- figuring out where you want to specialize, and
- building/maintaining your knowledge and motivation around the world's needs generally
The first is professionally relevant early in your career (or for generalists looking to lateral), but not so much later. The latter is personal/social/intellectual and perhaps a broad way that a specialist can give back by helping people working on the first thing.
If it is settled that AI is the thing to do, maybe point one has become irrelevant. I dispute this,[1] but less so than point two, which I think has strong independent value.
It may also be helpful context that I personally am not an expected utiltity maximizer. I'm doing my project because I want people to engage with EA arguments and then do what they want to do with them, as opposed to doing what I want them to do in a more superficial sense.
- ^
For example, it may just be critical to understand what else might be ITN in the world to understand why AI is important, or to think clearly about what its implications for welfare are. If those other problems aren't really in the room or fully explored, it's easy to miss crucial considerations. Similarly, what do we mean by "AI" and "settled?" Lots of EA epistemology can help here. Relatedly, the moral context of everything happening in the world can provide motivation that might otherwise be lacking.
At the core of my project is the idea that people can disagree and still cooperate.
I agree with you that the people who currently control the talent infrastructure that flows from CG, i.e. CEA and 80k, have for the most part become uninterested in views on cause prio that don't buy into the TAI hypothesis. They are not however, completely uninterested. As you say, they invite people working on global health, animals, and other causes to EAG; they support groups which discuss and invite speakers on these topics.
I understand that this is not much support in material terms compared to AI and this does stack the deck against non-AI causes for people making EA career choices. The question is what you want to do with it.
For my part, I am choosing to leverage that small amount of support to strengthen free-ranging discourse about how to do the most good. The bar may be higher for non-AI projects and people to get CG funding as they emerge from this discourse, but I don't think it is insurmountable. Further, I hope people who engage with my project will use it as a launching pad to reach out to non-CG funders and non-"EA" collaborators for their non-AI projects.
Both of these will be more challenging, but I personally resolve to support people doing what they endorse doing based on how thoughtful and ambitious they are, where I measure neither of those things in terms of how much they agree with me or my funders in substance. It'll be tough re bias, but idk, liberalism conquered the world last century, maybe it'll do it again this century.
We're looking at this more differently than I thought. The question "how does EA meet the needs of people with different worldviews" is strange to me. EA should be the place you go to *form* your worldview, by learning about and comparing different perspectives. Whatever has caused this framing to seem tricky/unnatural is the thing I'm pushing back against.
I have a similar take on TAI skepticism, with some added (perhaps excessively charitable) concerns around how economic value gets created in the first place and what hurdles there are between current AI systems and creating that value.
Yes. I am at-all interested in outcomes such that I will regard myself as having failed a sanity check if very few people go on to do ambitious, impactful work after engaging with my events/programs/groups. But I am very bound to being permissive about what counts here. Thoughtfulness about high impact is the bar, not my EV calculation of impact.
To put it in the form of a critique, I think too many community building programs adopt metrics like "number of participants who go into roles at AIM, MATS, GovAI, etc." and that this is too prescriptive and discourages people from really forming their own world models in an EA context.
My metric is whether I'm impressed with the pushback I get on my takes when I go into these spaces or whether I'm learning new and plausibly very important things about big problems.
The framing of your question suggests EA's role is to prescribe actions. I think EA is centrally a question and a set of abstract tools for understanding the world's needs. Using those tools will take different people in different directions. I want to support people using the tools well and I resolve not to judge how well people use the tools based on the specific conclusions they draw.
In particular, I think one of the biggest gaps in AI safety is having a clear understanding of why the majority of educated people reject the TAI hypothesis. Likewise, I think that many people who wish to do good in the world reject the TAI hypothesis for bad reasons that they would regret on reflection. Where do people go to correct these errors and build better models of which actions are best to take all things considered?
I don't know of a better venue than the best pockets of the current EA community. I want to make those pockets bigger!
I think your read is basically right. Thinking explicitly and granularly about the direct chain from your actions to last-mile impact and being sensitive to perturbations of that measurement is one area I think many current orgs over-invest in. I believe it is inconsistent with the processes that created those orgs in the first place (which I'm now trying to replicate without much focus on the direct, measurable outputs).
The biggest issue I see is people spending up to 20% of their best hours getting bogged down in metrics and explicit planning when they could be spending much of that time doing things they're excited about which they've done a quick sense-check on.
I think this is one of the great strengths of liberal, big-tent projects. Support plausibly great people all playing to their strengths. Some of them will disappoint and under-perform your hyper-planned model, sure, but the over-performers will more than make up for it. I want to embody this principle in my org and the groups we support.
I would have predicted the positive press and basically think this would "work" today if these conditions were met:
I agree you on the overall downsides though. This sets a bad precedent that will be misused by many and burn a ton of social trust that is ultimately more important.