I’m glad you posted this! It’s an interesting idea, and the kind of thing that deserves to be discussed, because there’s a lot to be learned from different proposals for “EA education”.
Where I agree with you: I think it’s true that many people are out there somewhere in the world who agree with the general principles of EA (even if they don’t know what it is yet) and could be shifted onto a high-impact career path with the right nudge.
The question is: How can we find these people, and what’s the right nudge?
--
I don’t think something like an academy would be impossible to run, but it would be very difficult.
Before I lay out more detail, I’ll try to sum up my main concern in one sentence: “It’s really hard to teach people to do the kinds of EA work that are most in-demand.”
If you look at the 80,000 Hours job board, you’ll find that most of the highest-impact jobs they know about require a pretty specific background and/or set of skills—more than can be obtained in a few months. If we have money and time to spend on the education of people who want to make an impact, my best guess is that trying to help them obtain the requisite background/skills directly will be better than teaching EA principles and concepts more generally.
--
Some other points related to risk and difficulty:
Even at schools focused on a simple and easily marketable skillset (programming), many students drop out of the program or graduate and still can’t find a job.
Lambda may be an exception, but for every exception, there is a rule, and the rule of “programming bootcamps” seems to be that many things can go wrong on the road from novice to professional.
Some of the most valuable “skills” for EA organizations are either not teachable (policy experience) or are very difficult to teach (research, general execution skills).
My impression is that the most successful research organizations try to avoid teaching research skills by filtering heavily on that skill ahead of time, even though this greatly increases their hiring costs.
If Open Phil had a way to generate five new Research Analysts to work for two years each using, say, $500,000 and six total months of staff time (“two instructors”), I strongly suspect that they would do so. The fact that they don’t do this makes me think that teaching EA research is probably hard (though I don’t work at Open Phil, and this is speculative).
The two instructors will be hard to find.
They’ll need to have a very strong grasp on EA concepts, be reasonably good at teaching, be reasonably good at curriculum-writing (unless that’s a job for a third talented person with spare time), and not be working on something with higher EV.
The students will be hard to find.
If people apply from within the EA community, they probably know quite a bit of the material already and don’t need the “nudge” (instead, they might be better served joining a school like Lambda, doing a research internship, or something else more skill-building for their specific goals).
If people apply from elsewhere, and have valuable skills, is it better for them to enter an “academy” than to work with 80,000 Hours?
This seems like a bad deal for anyone who plans to enter the private sector, since it charges a lot of money and doesn’t teach any “marketable” skills.
As something like a college course, it might hold appeal for people who like learning for its own sake. But I think most people who would be good candidates for the class are also likely to be capable of learning the material through some combination of reading, participating in discussion online, traveling to an event or two, and maybe Skyping with more experienced people once in a while. (We don’t have an easy way to set up Skype calls like that right now, but building one would be a lot easier than building an academy.)
--
This is me trying to play Devil’s advocate, so I apologize if I sound harsh! There are certainly positive aspects to the plan: It’s possible that finding good students would be easier than I envision, and if the program only succeeded half the time, that would still be a good outcome.
Questions I think you’ll want to answer going forward: How did current employees of EA organizations (or people running independent projects) get to their positions? If they could go back to the end of college (for example) and try to teach themselves to do their own jobs in a few months, how far would they get? What would they try to teach themselves? How do those answers map onto what you think an “EA Academy” should teach?
I was going to raise some similar objections, but you did it better. :) The only extra point I would raise is that my worst fear with project isn’t it being a waste of resources—my worst fear would be that it leads to significantly worse group-think as the “best and brightest” absorb the “right answers” from one or two teachers.
Thanks Aaron, I appreciate your feedback! You point out a lot of potential issues I did not consider. Given those issues and the existing opportunities like the EA Hotel and the collegiate EA Fellowship programs, I think an EA Academy may not be the best option to try right now.
Interesting! I’ve been working on a much lighter thing that’s like the college EA Fellowship program but for working adults, and have been debating whether it should be an ongoing once per week kind of thing or an intense weekend retreat (i.e. the CFAR model). EA Academy (or REACH Academy since we’d be running it out of the Berkeley REACH) came up as a potential name during brainstorming (haven’t landed on a definite name yet, will likely depend on what the program ends up looking like).
I’ve been an instructor at a coding bootcamp in the past and am generally interested in non-traditional education. I like the thought of your proposal, but see some potential flaws, many of which have already been laid out by Aaron. I’d be interested in chatting, though!
It seems to me the plan is based on several assumptions which actually do not hold
Effective altruism wants to grow the number of its members faster. It seems there was something like an attempt to deliberately slow down, limit outreach, international expansion, etc. based on problems with too fast expansion like coordination or diluted knowledge. The academy would likely help with such problems, but generally there is likely not such a strong urge to create new EAs now as you feel.
If EA wants to grow faster, there are cheaper ways.
There seems to be a persistent misconception in the community about how likely is it to get “initial grants” from OPP.
That said, something like EA Academy is a format which may be wort to explore some time in the future. (Other people thought about the idea before)
I think this is an interesting idea, but currently the EA hotel is struggling to obtain sufficient funding, so I wouldn’t so it as beneficial to spring up a second long-term residential program while there’s already one very similar such experiment in progress. If anyone would want to be a student in such a school they can already apply to the EA hotel and learn the skills that they need. It’s more self-directed than your idea, but I suspect that this is a vital skill for most EA jobs.
I wasn’t aware the EA hotel was struggling; that should take priority. We probably want to make sure self-directed people have the resources they need to study and work on EA before funding people who may be less self-directed.
The school could then be sustained by a Lambda school-style deal (Income Share Agreement): If you get employed outside EA, you owe 10% of your salary for 2 years once you make over $50,000. If you work for a think tank or inside EA, the cost is waived.
I’m glad you posted this! It’s an interesting idea, and the kind of thing that deserves to be discussed, because there’s a lot to be learned from different proposals for “EA education”.
Projects you might find valuable to look up, because they had/have certain features in common with your idea: Students for High-Impact Charity, collegiate EA Fellowship programs, and the EA Hotel.
--
Where I agree with you: I think it’s true that many people are out there somewhere in the world who agree with the general principles of EA (even if they don’t know what it is yet) and could be shifted onto a high-impact career path with the right nudge.
The question is: How can we find these people, and what’s the right nudge?
--
I don’t think something like an academy would be impossible to run, but it would be very difficult.
Before I lay out more detail, I’ll try to sum up my main concern in one sentence: “It’s really hard to teach people to do the kinds of EA work that are most in-demand.”
If you look at the 80,000 Hours job board, you’ll find that most of the highest-impact jobs they know about require a pretty specific background and/or set of skills—more than can be obtained in a few months. If we have money and time to spend on the education of people who want to make an impact, my best guess is that trying to help them obtain the requisite background/skills directly will be better than teaching EA principles and concepts more generally.
--
Some other points related to risk and difficulty:
Even at schools focused on a simple and easily marketable skillset (programming), many students drop out of the program or graduate and still can’t find a job.
Lambda may be an exception, but for every exception, there is a rule, and the rule of “programming bootcamps” seems to be that many things can go wrong on the road from novice to professional.
Some of the most valuable “skills” for EA organizations are either not teachable (policy experience) or are very difficult to teach (research, general execution skills).
My impression is that the most successful research organizations try to avoid teaching research skills by filtering heavily on that skill ahead of time, even though this greatly increases their hiring costs.
If Open Phil had a way to generate five new Research Analysts to work for two years each using, say, $500,000 and six total months of staff time (“two instructors”), I strongly suspect that they would do so. The fact that they don’t do this makes me think that teaching EA research is probably hard (though I don’t work at Open Phil, and this is speculative).
The two instructors will be hard to find.
They’ll need to have a very strong grasp on EA concepts, be reasonably good at teaching, be reasonably good at curriculum-writing (unless that’s a job for a third talented person with spare time), and not be working on something with higher EV.
The students will be hard to find.
If people apply from within the EA community, they probably know quite a bit of the material already and don’t need the “nudge” (instead, they might be better served joining a school like Lambda, doing a research internship, or something else more skill-building for their specific goals).
If people apply from elsewhere, and have valuable skills, is it better for them to enter an “academy” than to work with 80,000 Hours?
This seems like a bad deal for anyone who plans to enter the private sector, since it charges a lot of money and doesn’t teach any “marketable” skills.
As something like a college course, it might hold appeal for people who like learning for its own sake. But I think most people who would be good candidates for the class are also likely to be capable of learning the material through some combination of reading, participating in discussion online, traveling to an event or two, and maybe Skyping with more experienced people once in a while. (We don’t have an easy way to set up Skype calls like that right now, but building one would be a lot easier than building an academy.)
--
This is me trying to play Devil’s advocate, so I apologize if I sound harsh! There are certainly positive aspects to the plan: It’s possible that finding good students would be easier than I envision, and if the program only succeeded half the time, that would still be a good outcome.
Questions I think you’ll want to answer going forward: How did current employees of EA organizations (or people running independent projects) get to their positions? If they could go back to the end of college (for example) and try to teach themselves to do their own jobs in a few months, how far would they get? What would they try to teach themselves? How do those answers map onto what you think an “EA Academy” should teach?
I was going to raise some similar objections, but you did it better. :) The only extra point I would raise is that my worst fear with project isn’t it being a waste of resources—my worst fear would be that it leads to significantly worse group-think as the “best and brightest” absorb the “right answers” from one or two teachers.
Or, frankly, that it would make EA seem even more like a cult than it already does.
The EA Hotel hosted an EA Retreat which sounds a bit similar. Here’s a report from a Czech EA retreat.
Thanks Aaron, I appreciate your feedback! You point out a lot of potential issues I did not consider. Given those issues and the existing opportunities like the EA Hotel and the collegiate EA Fellowship programs, I think an EA Academy may not be the best option to try right now.
Interesting! I’ve been working on a much lighter thing that’s like the college EA Fellowship program but for working adults, and have been debating whether it should be an ongoing once per week kind of thing or an intense weekend retreat (i.e. the CFAR model). EA Academy (or REACH Academy since we’d be running it out of the Berkeley REACH) came up as a potential name during brainstorming (haven’t landed on a definite name yet, will likely depend on what the program ends up looking like).
I’ve been an instructor at a coding bootcamp in the past and am generally interested in non-traditional education. I like the thought of your proposal, but see some potential flaws, many of which have already been laid out by Aaron. I’d be interested in chatting, though!
That sounds like a much better approach as it requires much less resources to be committed upfront.
It seems to me the plan is based on several assumptions which actually do not hold
Effective altruism wants to grow the number of its members faster. It seems there was something like an attempt to deliberately slow down, limit outreach, international expansion, etc. based on problems with too fast expansion like coordination or diluted knowledge. The academy would likely help with such problems, but generally there is likely not such a strong urge to create new EAs now as you feel.
If EA wants to grow faster, there are cheaper ways.
There seems to be a persistent misconception in the community about how likely is it to get “initial grants” from OPP.
That said, something like EA Academy is a format which may be wort to explore some time in the future. (Other people thought about the idea before)
I think this is an interesting idea, but currently the EA hotel is struggling to obtain sufficient funding, so I wouldn’t so it as beneficial to spring up a second long-term residential program while there’s already one very similar such experiment in progress. If anyone would want to be a student in such a school they can already apply to the EA hotel and learn the skills that they need. It’s more self-directed than your idea, but I suspect that this is a vital skill for most EA jobs.
I wasn’t aware the EA hotel was struggling; that should take priority. We probably want to make sure self-directed people have the resources they need to study and work on EA before funding people who may be less self-directed.
I can’t imagine anyone signing this contract.
Paradigm Academy comes to mind. Curious about how you see your proposal as being different from that.
The Pareto Fellowship even moreseo for me. Here CEA explains why they discontinued it.
The whole thread around the comment you linked to seems relevant to this.
Oh yeah, good call. Forgot about the Pareto Fellowship.
There is also the EA MOOC. There does not appear to be a counter—does anyone know how many completions of this course there has been?