27. What I learned from the Effective Altruism Global conference
I flew to London to attend Burning Man for nerds
I am briefly exiting my graduate school hibernation to share my thoughts on EA Global. The below is not different from my regular post — it is more informal observations, part journal, part anthropology experiment.
A quick definition: Effective Altruism (EA throughout this blog) is a philosophy and social movement that advocates the use of evidence and reasoning to determine the most effective ways to benefit others.
If you’re new here, subscribe!
Last weekend was one of the best weekends of my life. I flew to London to attend the three day Effective Altruism Global Conference, an annual gathering of people who think about the best ways of doing good.
I have cared a lot about EA for a while. I discovered EA after college and read a lot of the standard books, and the movement helped me answer some deep questions that I was struggling with (how do I do something good with my life? which jobs should I look for that aren’t consulting? etc). And while I’ve considered myself an effective altruist, before this weekend, I had never met actual EAs or attended an EA event.
Imagine a cartoon where the small animal main character gets separated from her small animal family but by the end of the movie discovers that a utopian place where indeed, there are tons of small animals just like her and she — sing it with me — belongs. That’s kinda how I felt last weekend. Anyway, that means these thoughts are likely overly rosy.
THAT BEING SAID. I’ve never felt like I belonged quite as much as I belonged amongst the EAs. It’s not that I’m a loner in general and this was a change of pace. I generally pick places where I fit in. But at University of Chicago, people were nerdy, but they didn’t uniformly care about the world. At Berkeley, people care about the world, but aren’t uniformly nerdy. EAs are both.
The people I met were researchers, philosophers, activists, students, and computer scientists. But they all shared many traits: genuine care for the world, technical aptitude, philosophical thinking, being straightforward. I had some of the best conversations I’ve ever had in my life. The people I met made me feel smart, and kind. And doesn’t everybody just want to feel smart and kind?
I had to submit a short application to get admission. My understanding was that they screened to select more mature attendees — people who were further along in their career and knew more about EA. Either way, with 1,000 attendees, it was the largest conference to date. My flight and Airbnb was incredibly generously paid by EA Berkeley (shoutout Dylan Miars). The reason that undergrads have $$ to throw around (as I understand) is that EA believes that recruiting college students is an effective way to get people into the movement, and so funds EA university chapters well.
If you’re interested in an EA conference but are new to EA, EA Global Prague is happening in December and they’re accepting applications until Nov 18.
Major takeaway: Nobody networks better than EAs
And that’s because they know that the key source of impact from EA events comes from personal connections.
Before the conference, the organizers asked me whether I want to commit to having X number of 1-1 meetings (I think I answered 3-5, ultimately an underestimate), and whether I want to commit to making X number of introductions (I said 1-2, and made 0). I found this an odd and somewhat intimidating question (I don’t know anybody!) but my fears were soon put to rest.
First off, all the sessions were recorded, and it felt like we were encouraged to skip the sessions and instead go meet people. Some people I talked to had as many as ten 30 min meetings per day.
The conference app, Swapcard, also facilitated networking. It made it easy to see profiles of other attendees and message them. You could see the other attendees’ schedule and request to book a meeting with them during a free slot, all within the app. Contrary to my expectation, I felt very little social hesitation in booking meetings. One man requested a meeting with me simply because I had asked him good questions about AI at the career fair. I requested someone because they were an Oxford student in philosophy and physics (according to their Swapcard profile), which just sounded cool (we ended up talking about relativity and time worms, thanks Max!). It was easy to search attendees according to their tags (like “animal welfare”), affiliation, or job. I ended up making incredibly valuable connections with people who I foresee can help me with research for my PhD.
This is in contrast to my usual experience at conferences, where reaching out to people you admire and are more successful than you is scary, and reaching out to peers makes you seem overeager or intruding. What a shame.
It seems EAs are aware of human biases, and that awareness enables them to actively work around those biases to get better results. Meeting people is hard, especially when they’re strangers or more powerful than you. So the conference was organized to make meeting people the default. I appreciate this frame as it helps most people overcome our natural tendency to be shy or aloof.
I think observations from newcomers contain valuable information for an organization/movement/individual. This is because more experienced people tend to get used to the bad stuff, while newcomers have fresh eyes and are unlikely to have already gotten used to oddities, mistakes, and inefficiencies in your organization.
Here are my (perhaps obvious) newcomer observations:
Way more men than women are in EA. The conference seemed around 65-70% men. In the AI focused events, I was one of only a handful of women.
Relatedly: Why? And is anybody doing anything about it?
A lot of people I met talked in probabilities (“I’m 70% sure in this belief”), as well as in terms of updating (“I haven’t updated my beliefs but that doesn’t mean I didn’t get any value”).
A quote from an old guard EA I met: “When EA started, we were talent rich and resource poor. Now, we’re talent poor and resource rich.” This makes me wonder how effective small individual donations, like the ones I give, are.
What the hell is forecasting and why should I care about it? Email me.
Brits care way less about wearing masks than people in the Bay Area. Hardly anyone in this conference of 1000 people wore a mask. More generally in London, only ~30% of the people in public transit are wearing masks.
Relatedly, 15-20% of the attendees seemed Bay Area based or related.
The organizers did a good job of meetings attendees’ non-conference needs. People are not automatons, and everybody at some point was either jet lagged, socially overstimulated, or both. There was a nap room with bean bags, always full of slightly snoring EAs. I slept there for an hour every day and always felt incredibly refreshed afterward. Why don’t economics conferences have nap rooms?
All the food was vegan, and it was delicious. The venue served us everything from apricot tagine to Swedish meatballs to vermicelli noodles with king oyster mushrooms.
Stuff I learned
There is a non-trivial probability of human-like AI by 2050.
Joslyn Barnhart made an interesting argument for why we need more women in AI. Women tend to be more risk averse, and that’s something we want in AI research. In one survey, women were more likely to say that “we should invest more in AI security”. I take this to mean that if we leave AI to men the world will blow up.
A researcher at Oxford shared their idea of having a suffering tax on meat. Like a carbon tax, it would tax the meat proportional to the amount of suffering the animal had endured. Same thing can be done by issuing a limited number of cap and trade certificates for facilities producing caged eggs. I think this won’t work but I like it nonetheless.
Apparently a lot of people are freaked out by GPT3. Why? Email me.
The climate philanthropy space is crowded. Unlike animal welfare, where the next animal welfare charity is probably having additive impact, the average climate organization is probably not high impact.
Claire Walsh (Policy Director @ JPAL) taught me that 90% of climate funding currently goes to mitigation, and only 7% goes to adaptation. Since we’re locked into some amount of warming already, we should increase the total dollars going to adaptation (but not at the expense of mitigation).
Facts from EA movement co-founder Will MacAskill
“If it weren’t for the potato, we’d have a billion fewer people on the planet” (economists back him up)
GiveWell and Open Philanthropy have hit a milestone of moving a total of a billion dollars to effective causes.
He said we don’t know the answer to basic questions like, “Of the humans alive today, how many have lives that are above zero versus below zero?”. i.e., how many peoples’ lives feel worse than not existing (would answer yes to the question, “If you could choose not to be born, would you?”). I had not yet considered this a basic question.
We need more EA historians to tell us what things are persistent (values, information, some institutions) and which things are not persistent (most companies). That way we can figure out which types of things can create lasting change. So if you’re thinking about long term impact, starting a company is probably not as effective as changing values.
One argument for leaving fossil fuels in the ground is that if humanity self destructs and we need to re-industrialize, we need to leave those future humans some coal and oil to burn. So maybe EA should buy the North Antelope Rochelle Mine in Wyoming — which has enough coal to power the first 40 years of industrialization — and just leave it there.
That’s it folks. Thanks again to EA Berkeley and my workplace (CEGA) for sponsoring an amazing weekend. If I met you at EAGL and you’re reading this, well met.