This is a (late) Draft Amnesty Week draft |
I claim that there’s a special thing about some EAs. It’s really hard to describe this using a low number of bits, which should make me skeptical that it exists. I’ve here settled on describing the collection of skills that people with that thing have. I claim that this set of highly correlated traits predicts success in EA about as well as raw intelligence (or more), and is much more trainable. I won’t justify this, and instead, please accept this draft amnesty post gesturing at a thing.
Epistemic status: I find this model useful, and make this claim with the part of me that buys into it, but I’m aware that it doesn’t land for many people I describe it to, including those who I would describe as having this skillset. So take it as a model but not one with my full endorsement.
The “skills”:
- Highly altruistic
- Impact comes first
- Alliance mentality is easy with other EAs
- Scout mindset
- This and the above factors kinda work well together. When you’re really dedicated to the mission, it’s easier to put down your pride, realize you were wrong and start focusing on the right thing.
- This person probably speaks in a lot of probabilities
- Has learned to differentiate between their inside view and when they’re deferring
- Related: good at reasoning transparency
- Generalists
- You can count on them having a breadth of knowledge about whatever area they’re in. When you believe something is the most important thing in the world, it’s easier to motivate yourself to learn about adjacent things. And they have done enough heroically diving into to help their allies that they’ve learned areas outside their main specialty.
- Knows EA ideas very well
- Need someone to explain EA 101? Easy. Biosecurity 101? They gotchu. The cause area they know least about? They probably can do a half-decent job.
- If a new consideration is coming down the pipe that hasn’t made its way into their team’s official strategy yet, they’re probably tracking it, and may already be adjusting their behavior
- Has engaged with the EA community
- They have a network and a reputation
- If their work comes under fire on the EA Forum, they can write a reasonable response, because they know how to communicate with EAs
- Cluster of EA-ish traits that are less central to the definition, but are correlated:
- Growth mindset
- Probably does better than you’d expect at talking about feelings, controlling for whatever level of analytical they are
- Probably skeptical of the sorts of established wisdom I think are dumb :p
- Agentic?
- Knows there are no adults in the room
The EA Spark
Nicole Ross called the potential for the EA Core Skillset “the EA Spark”, which I also love, and I expect may evoke the right idea if my previous writings have failed.
Disclaimer
I think there are people who are committed EAs, who really don’t buy into the whole enchilada. I’m sympathetic to that being an option. So: “What the hell, JP, you’re telling me that I can only ‘have this skillset’ if I accept all these independent claims that EAs make?” Am I? I’m not sure. I do think that many people can have the EA Core Skillset without buying into everything, or engaging much with EA. But many do, and without a better definition I just want to gesture at this cluster.
I really like the thought of the EA spark :) 🔥 I’m on week 6 of my EA intro course which is really interesting. I’m feeling stuck in terms of finding so many issues worth pursuing, knowing my skill set and also how much more I want to and plan to learn whilst balancing with just getting out and adding to the action to help make change. I can;t seem to find how to help as traditional job applications have not been successful which is surprising since I resigned from a senior position in the tech space (“soft AI” ) to pursue opportunities in climate friendly areas plus AI - since I was in the field. Any advice as to how I can unstick and get out to help make change??
Strong upvoted, I'm glad I read this and really wish I'd read this or something like this back in 2017. Finding out about some of these skills and their importance, at all, was a very long slog for me and many others; the process of trial and error often yields valuable insight, but for most of these skills, it didn't have to happen and the waste from the delay was not worth it.