I need to think about this issue more, but I think there might be a couple of problems with the estimates.
1) Let’s divide problems into the world into ‘funding constrained’ and ‘talented constrained’. What you’ve done is pick the most funding constrained causes we know (Givedirectly and AMF) and then say “wow these can absorb a lot of funds”, which is not surprising, because they were selected for that property.
But there are other causes where it looks like a talented person could make a big difference but where it’s not easy for money to buy progress. These are causes that are more constrained by innovation, leadership, coordination, and so on. Some areas that might fall in this category include EA movement building, much of research, green energy, much of policy, international relations. We asked Holden to speculate on what they might be here:
https://80000hours.org/2014/10/interview-holden-karnofsky-on-cause-selection/
We asked biomedical researchers to estimate how much money they would trade for a researcher with good personal fit, and they often named figures of around $1m per year, more than most people could donate.
Taking talent gaps into account too, it becomes far less clear where the ideal balance lies.
It seems likely the world is more talent constrained than funding constrained, if that question makes sense.
2) The figures for how much the typical etg person will donate might be a big underestimate. You can’t easily infer the long-term average from the 80k surveys because those are surveys of ppl very early in the career—indeed some of the ppl are still at college. Many EAs have long-term earning potential over $1m, so will be donating $100-$500k per year, so your estimate could be out by a factor of 10.
3) You’re comparing the most talented direct workers (Rob Mather) with the typical etger. It would be more fair to compare equally talented people. The people with best fit for earning to give will be able to donate many millions within a couple of years, which is similar to the amount of room for funding created by a staff member at AMF. So that might suggest a 1:1 ratio of etg to direct work.
And if you think of the typical salaries at an EA org (~$50k per year), one talented etger will be able to cover the salaries of ~20 people.
4) The EA movement is pretty small so it seems very achievable to pull in funds from elsewhere, and there’s been a strong track record of doing this e.g. most of SCI’s funding has come from Gates; Thiel funded a bunch of things; CEA has a bunch of external donors.
5) What about value of information? An EA movement where 95% of people etg as software engineers while 5% do direct work is going to have very stunted learning opportunities. I’d prefer to see EAs working in a wide variety of causes and sectors, then sharing what they learn with each other. A similar consideration applies to the EA movement building a wide portfolio of skills so it can address big problems in the future.
6) I’m unsure about career capital. I’m tempted to agree that for the median person etg might normally offer better career capital, but if you’re especially talented it may be better just to focus on doing something impressive in an important cause.
https://80000hours.org/2015/07/what-people-miss-about-career-capital-exceptional-achievements/
I also think people underestimate the career capital you get from working at EA orgs. e.g. I think I gained far better career capital from working at 80k than I could have done in finance, and I had good options there.
7) I’m unsure etg fits more people. Bear in mind that the common sense position is that earning to give is bizarre and no-one does it. Whereas loads of people want to work in teaching, nonprofits, research and so on.
Also, if you find it hard to stay altruistically motivated, then it’s probably better to be among lots of other altruists rather than being the only person in your company etg.
I want to push back a bit against point #1 (“Let’s divide problems into ‘funding constrained’ and ‘talent constrained’.) In my experience recruiting for MIRI, these constraints are tightly intertwined. To hire talent, you need money (and to get money, you often need results, which requires talent).
I think the “are they funding constrained or talent constrained?” model is incorrect, and potentially harmful. In the case of MIRI, imagine we’re trying to hire a world-class researcher for $50k/year, and can’t find one. Are we talent constrained, or funding constrained? (Our actual researcher salaries are higher than this, but they weren’t last year, and they still aren’t anywhere near competitive with industry rates.)
Furthermore, there are all sorts of things I could be doing to loosen the talent bottleneck, but only if I knew the money was going to be there. I could be setting up a researcher stewardship program, having seminars run at Berkeley and Stanford, and hiring dedicated recruiting-focused researchers who know the technical work very well and spend a lot of time practicing getting people excited—but I can only do this if I know we’re going to have the money to sustain that program alongside our core research team, and if I know we’re going to have the money to make hires. If we reliably bring in only enough funding to sustain modest growth, I’m going to have a very hard time breaking the talent constraint.
And that’s ignoring the opportunity costs of being under-funded, which I think are substantial. For example, at MIRI there are numerous additional programs we could be setting up, such as a visiting professor + postdoc program, or a separate team that is dedicated to working closely with all the major industry leaders, or a dedicated team that’s taking a different research approach, or any number of other projects that I’d be able to start if I knew the funding would appear. All those things would lead to new and different job openings, letting us draw from a wider pool of talented people (rather than the hyper-narrow pool we currently draw from), and so this too would loosen the talent constraint—but again, only if the funding was there.
Right now, we have more trouble finding top-notch math talent excited about our approach to technical AI alignment problems than we have raising money, but don’t let this fool you—the talent constraint would be much, much easier to address with more money, and there are many things we aren’t doing (for lack of funding) that I think would be high impact.
Ben, between your comments these ones I made, and AGB’s comments above, I’m thinking of writing not a direct rebuttal to Peter Hurford’s estimates of ideal proportion of ETG:direct-work, but a post called “When Should You Go Into Direct Work”, which would be a list of heuristic considerations of when someone should consider going into direct work vs. earning to give. I think it’s important to make a visible response to Peter undoing a potential misconception earning to give is a better fit than some kind of direct work than a few disparate comments we’ve made. I especially think your points 3 and 5-7 are important considerations for individuals EAs making career choices, significant plan changes, etc.
Would you like to read or comment on the draft of such a post when it’s available?
Interesting post, thanks Peter!
I need to think about this issue more, but I think there might be a couple of problems with the estimates.
1) Let’s divide problems into the world into ‘funding constrained’ and ‘talented constrained’. What you’ve done is pick the most funding constrained causes we know (Givedirectly and AMF) and then say “wow these can absorb a lot of funds”, which is not surprising, because they were selected for that property.
But there are other causes where it looks like a talented person could make a big difference but where it’s not easy for money to buy progress. These are causes that are more constrained by innovation, leadership, coordination, and so on. Some areas that might fall in this category include EA movement building, much of research, green energy, much of policy, international relations. We asked Holden to speculate on what they might be here: https://80000hours.org/2014/10/interview-holden-karnofsky-on-cause-selection/
We asked biomedical researchers to estimate how much money they would trade for a researcher with good personal fit, and they often named figures of around $1m per year, more than most people could donate.
Taking talent gaps into account too, it becomes far less clear where the ideal balance lies.
It seems likely the world is more talent constrained than funding constrained, if that question makes sense.
2) The figures for how much the typical etg person will donate might be a big underestimate. You can’t easily infer the long-term average from the 80k surveys because those are surveys of ppl very early in the career—indeed some of the ppl are still at college. Many EAs have long-term earning potential over $1m, so will be donating $100-$500k per year, so your estimate could be out by a factor of 10.
3) You’re comparing the most talented direct workers (Rob Mather) with the typical etger. It would be more fair to compare equally talented people. The people with best fit for earning to give will be able to donate many millions within a couple of years, which is similar to the amount of room for funding created by a staff member at AMF. So that might suggest a 1:1 ratio of etg to direct work.
And if you think of the typical salaries at an EA org (~$50k per year), one talented etger will be able to cover the salaries of ~20 people.
4) The EA movement is pretty small so it seems very achievable to pull in funds from elsewhere, and there’s been a strong track record of doing this e.g. most of SCI’s funding has come from Gates; Thiel funded a bunch of things; CEA has a bunch of external donors.
5) What about value of information? An EA movement where 95% of people etg as software engineers while 5% do direct work is going to have very stunted learning opportunities. I’d prefer to see EAs working in a wide variety of causes and sectors, then sharing what they learn with each other. A similar consideration applies to the EA movement building a wide portfolio of skills so it can address big problems in the future.
6) I’m unsure about career capital. I’m tempted to agree that for the median person etg might normally offer better career capital, but if you’re especially talented it may be better just to focus on doing something impressive in an important cause. https://80000hours.org/2015/07/what-people-miss-about-career-capital-exceptional-achievements/ I also think people underestimate the career capital you get from working at EA orgs. e.g. I think I gained far better career capital from working at 80k than I could have done in finance, and I had good options there.
7) I’m unsure etg fits more people. Bear in mind that the common sense position is that earning to give is bizarre and no-one does it. Whereas loads of people want to work in teaching, nonprofits, research and so on.
Also, if you find it hard to stay altruistically motivated, then it’s probably better to be among lots of other altruists rather than being the only person in your company etg.
I want to push back a bit against point #1 (“Let’s divide problems into ‘funding constrained’ and ‘talent constrained’.) In my experience recruiting for MIRI, these constraints are tightly intertwined. To hire talent, you need money (and to get money, you often need results, which requires talent).
I think the “are they funding constrained or talent constrained?” model is incorrect, and potentially harmful. In the case of MIRI, imagine we’re trying to hire a world-class researcher for $50k/year, and can’t find one. Are we talent constrained, or funding constrained? (Our actual researcher salaries are higher than this, but they weren’t last year, and they still aren’t anywhere near competitive with industry rates.)
Furthermore, there are all sorts of things I could be doing to loosen the talent bottleneck, but only if I knew the money was going to be there. I could be setting up a researcher stewardship program, having seminars run at Berkeley and Stanford, and hiring dedicated recruiting-focused researchers who know the technical work very well and spend a lot of time practicing getting people excited—but I can only do this if I know we’re going to have the money to sustain that program alongside our core research team, and if I know we’re going to have the money to make hires. If we reliably bring in only enough funding to sustain modest growth, I’m going to have a very hard time breaking the talent constraint.
And that’s ignoring the opportunity costs of being under-funded, which I think are substantial. For example, at MIRI there are numerous additional programs we could be setting up, such as a visiting professor + postdoc program, or a separate team that is dedicated to working closely with all the major industry leaders, or a dedicated team that’s taking a different research approach, or any number of other projects that I’d be able to start if I knew the funding would appear. All those things would lead to new and different job openings, letting us draw from a wider pool of talented people (rather than the hyper-narrow pool we currently draw from), and so this too would loosen the talent constraint—but again, only if the funding was there.
Right now, we have more trouble finding top-notch math talent excited about our approach to technical AI alignment problems than we have raising money, but don’t let this fool you—the talent constraint would be much, much easier to address with more money, and there are many things we aren’t doing (for lack of funding) that I think would be high impact.
I agree many things are both talent and constrained and funding constrained.
I think you can have the whole spectrum from mainly constrained by a certain type of talent, to constrained by both, to mainly constrained by funding.
Ben, between your comments these ones I made, and AGB’s comments above, I’m thinking of writing not a direct rebuttal to Peter Hurford’s estimates of ideal proportion of ETG:direct-work, but a post called “When Should You Go Into Direct Work”, which would be a list of heuristic considerations of when someone should consider going into direct work vs. earning to give. I think it’s important to make a visible response to Peter undoing a potential misconception earning to give is a better fit than some kind of direct work than a few disparate comments we’ve made. I especially think your points 3 and 5-7 are important considerations for individuals EAs making career choices, significant plan changes, etc.
Would you like to read or comment on the draft of such a post when it’s available?