I’ve done informal BOTECs and it seems like the current funding amounts are roughly correct, though we need to be careful with deploying this funding due to concerns like optics and epistemics. Regarding the example, spending $5k on EA group dinners is really not that much if it has even a 2% chance to cause one additional career change. This seems like a failure of communication, because funding dinners is either clearly good and students weren’t doing the BOTEC, or it’s bad due to some optics or other concerns that the students didn’t communicate to CEA.
That’s fair—I’m not the earlier commenter but would suggest (as someone who’s heard some of these conversations but isn’t necessarily representative of others’ thinking):
For dinners: Suppose offering to buy a $15 dinner for someone makes it 10% more likely than they’ll go to a group dinner, and suppose that makes it 1% more likely that they’ll have a very impactful career. Suppose that means counterfactually donating 10% of $100k for 40 years. Then on average the dinner costs $15 and yields $400.
For retreats: Suppose offering to subsidize a $4oo flight makes someone 40% more likely to go to a retreat and that this makes them 5% more likely to have a very impactful career. Again suppose that means counterfactually donating 10% of $100k for 40 years. Then on average the flight costs $400 and yields $8,000.
(And expected returns are 100x higher than that under bolder assumptions about how much impact people will have. Although they’re negative if optics costs are high enough.)
Thanks—this is exactly what I think is useful to have out there, and ideally to refine over time.
My immediate reaction is that the % changes you are assigning look very generous. I doubt a $15 dinner makes some 1% more likely to pursue an impactful career; and especially that a subsidised flight produced a 5% swing. I think these are likely orders of magnitude too high, especially when you consider that other places will also offer free dinners/retreats.
If a $400 investment in anything made someone 5% more likely to pursue an impactful career, that would be amazing.
But I guess what I’m really hoping is that CEA and FTX have exactly this sort of reasoning internally, with some moderate research into the assumptions, and could share that externally.
Thanks! Agree it’s good to refine these and that these are very optimistic—I suspect the optimism is justified by the track record of these events. Anecdotally, it seems nontrivially common for early positive interactions to motivate new community members to continue/deepen their (social and/or motivational) engagement, and that seems to often lead to impactful career plan changes.
(I think there’s steeply diminishing returns here—someone’s first exposure to the community seems much more potentially impactful than later exposures. I tried to account for “how many participants will be having their first exposure” in the earlier estimate.)
In other words, we could (say) break down the ~1% estimate (which is already conditioned on counterfactual dinner attendance) into the following (ignoring benefits for people who are early on but not totally new):
30% chance that this is their first exposure
conditional on the above, 10% chance that the experience kickstarts long/deep engagement
conditional on the above, 50% chance of an impactful career switch
(although early exposures that aren’t quite the first one also seem valuable)
If 1% is far too generous, which of the above factors are too high? (Maybe the second one?)
(Edited to add) And yup, I acknowledge this isn’t the source you were looking for—hopefully still adds to the conversation.
Regarding the example, spending $5k on EA group dinners is really not that much if it has even a 2% chance to cause one additional career change.
How much of the impact generated by the career change are you attributing to CEA spending here? I’m just wondering because counterfactuals run into the issue of double-counting (as discussed here).
Unsure but probably more than 20% if the person wouldn’t be found through other means. I think it’s reasonable to say there are 3 parties: CEA, the group organizers, and the person, and none is replaceable so they get 33% Shapley each. At 2% chance to get a career change this would be a cost of 750k per career which is still clearly good at top unis. The bigger issue is whether the career change is actually counterfactual because often it’s just a speedup.
I do think you have to factor in the potential negative risk of spending too much in that estimate as some of the potential members might be turned of by what seems like inefficient use of money. I think this is especially crucial if you are in the process of explaining the EA principles or when relating to members who not yet are committed to the movement.
Is $750k the market price for 1 expected career change from someone at a top school, excluding compensation costs? Alternatively, is there no cheaper way to cause such a career change? IMO, this is the important question here: if there is a cheaper way, then paying $750k has an opportunity cost of >1 career changes.
edit: misinterpreted what comment above meant by “market price”
I think the market price is a bit higher than that. The mean impact from someone at a top school is worth over $750k/year, which means we should fund all interventions that produce a career change for $750k (unless they have large non-financial costs) since those have a <3 year payback period even if the students take a couple years to graduate or skill up.
In practice, dinners typically produce way more than 2% of a career change for $5k of dinners (33 dinners for 10 people at $15/serving). The situation at universities has non-monetary bottlenecks, like information transmission fidelity, qualified organizers, operational capacity, university regulations, etc., and most things that get you better use of those other resources and aren’t totally extravagant are worth funding, unless they have a hidden cost like optics or attracting greedy people.
I think the market price is a bit higher than that.
Someone else in this thread found a report claiming that employers spend an average of ~$6,100 to hire someone at a US university. I also found this report saying that the average cost per hire in the United States is <$5,000, $15k for an executive. At 1 career = 10 jobs that’s $150,000/career for executive-level talent, or $180,000/career adjusting for inflation since the report was released.
I’m not sure how well those numbers reflect reality (the $15k/executive number looks quite low), but it seems at least fairly plausible that the market price is substantially less than $750k/career.
The mean impact from someone at a top school is worth over $750k/year, which means we should fund all interventions that produce a career change for $750k (unless they have large non-financial costs) since those have a <2 year payback period.
This line of reasoning is precisely what I’m claiming to be misguided. Giving you a gallon of water to drink allows you to live at least two additional days (compared to you having no water), which at $750k of impact/year (~$2000/day ) means, by your reasoning, that EA should fund all interventions that ensure you have 1 gallon of water for <=$4000, up to the amount you need to survive.
If water happened to be that expensive, that would be a worthwhile trade. But given the current market price of water (with the time cost of acquiring it included) being willing to pay anywhere near $4000/gallon is absurd.
In general, if you value something at $x, and its market price is $y, x only matters for deciding whether you should pay for the thing or not, not for deciding how much you should pay for it. If x >= y, then you should pay $y, otherwise you should pay $0.
It looks like I misunderstood a comment above. I meant “market price” as the rate at which CEA should currently trade between money and marginal careers, which is >$750k. I think you mean the average price at which other companies “in the market for talent” buy career changes, which is <$750k.
I think there isn’t really a single price at which we can buy infinite talent. We should do activities as cost-effective as other recruiters, but these can only be scaled up to a limited extent before we run into other bottlenecks. The existence of a cheaper intervention doesn’t mean we shouldn’t fund a more expensive intervention once the cheaper one is exhausted. And we basically want an infinite amount of talent, so in theory the activities that buy career changes at prices between $150k and $750k are also worth funding.
I think we can agree that
different activities have different cost-effectiveness, some of them substantially cheaper than $750k/career
we can use a basically infinite amount of talent, and the supply curve for career changes slopes upwards
we shouldn’t pay more than the market price for any intervention e.g. throw $100k at a university group for dinners when it produces the same effect as $5k spent on dinners
we should fund every activity that has a cost-effectiveness of better than $750k per career change (or whatever the true number is), unless we saturate our demand for talent and lower the marginal benefit of talent, or deplete much of our money and increase the marginal benefit of money
we are unlikely to saturate our demand for talent by throwing more money at EA groups because there are other bottlenecks
Because most of the interventions are much cheaper than $750k/career change, our average cost will be much less than $750k/career change
I’ve done informal BOTECs and it seems like the current funding amounts are roughly correct, though we need to be careful with deploying this funding due to concerns like optics and epistemics. Regarding the example, spending $5k on EA group dinners is really not that much if it has even a 2% chance to cause one additional career change. This seems like a failure of communication, because funding dinners is either clearly good and students weren’t doing the BOTEC, or it’s bad due to some optics or other concerns that the students didn’t communicate to CEA.
In the spirit of this post, maybe you could share these informal BOTECs?
‘Here is a BOTEC’ is going to help more than ‘I’ve done a BOTEC and it checks out’.
(I appreciate the post isn’t actually aimed at you)
That’s fair—I’m not the earlier commenter but would suggest (as someone who’s heard some of these conversations but isn’t necessarily representative of others’ thinking):
For dinners: Suppose offering to buy a $15 dinner for someone makes it 10% more likely than they’ll go to a group dinner, and suppose that makes it 1% more likely that they’ll have a very impactful career. Suppose that means counterfactually donating 10% of $100k for 40 years. Then on average the dinner costs $15 and yields $400.
For retreats: Suppose offering to subsidize a $4oo flight makes someone 40% more likely to go to a retreat and that this makes them 5% more likely to have a very impactful career. Again suppose that means counterfactually donating 10% of $100k for 40 years. Then on average the flight costs $400 and yields $8,000.
(And expected returns are 100x higher than that under bolder assumptions about how much impact people will have. Although they’re negative if optics costs are high enough.)
Thanks—this is exactly what I think is useful to have out there, and ideally to refine over time.
My immediate reaction is that the % changes you are assigning look very generous. I doubt a $15 dinner makes some 1% more likely to pursue an impactful career; and especially that a subsidised flight produced a 5% swing. I think these are likely orders of magnitude too high, especially when you consider that other places will also offer free dinners/retreats.
If a $400 investment in anything made someone 5% more likely to pursue an impactful career, that would be amazing.
But I guess what I’m really hoping is that CEA and FTX have exactly this sort of reasoning internally, with some moderate research into the assumptions, and could share that externally.
Thanks! Agree it’s good to refine these and that these are very optimistic—I suspect the optimism is justified by the track record of these events. Anecdotally, it seems nontrivially common for early positive interactions to motivate new community members to continue/deepen their (social and/or motivational) engagement, and that seems to often lead to impactful career plan changes.
(I think there’s steeply diminishing returns here—someone’s first exposure to the community seems much more potentially impactful than later exposures. I tried to account for “how many participants will be having their first exposure” in the earlier estimate.)
In other words, we could (say) break down the ~1% estimate (which is already conditioned on counterfactual dinner attendance) into the following (ignoring benefits for people who are early on but not totally new):
30% chance that this is their first exposure
conditional on the above, 10% chance that the experience kickstarts long/deep engagement
conditional on the above, 50% chance of an impactful career switch (although early exposures that aren’t quite the first one also seem valuable)
If 1% is far too generous, which of the above factors are too high? (Maybe the second one?)
(Edited to add) And yup, I acknowledge this isn’t the source you were looking for—hopefully still adds to the conversation.
How much of the impact generated by the career change are you attributing to CEA spending here? I’m just wondering because counterfactuals run into the issue of double-counting (as discussed here).
Unsure but probably more than 20% if the person wouldn’t be found through other means. I think it’s reasonable to say there are 3 parties: CEA, the group organizers, and the person, and none is replaceable so they get 33% Shapley each. At 2% chance to get a career change this would be a cost of 750k per career which is still clearly good at top unis. The bigger issue is whether the career change is actually counterfactual because often it’s just a speedup.
I do think you have to factor in the potential negative risk of spending too much in that estimate as some of the potential members might be turned of by what seems like inefficient use of money. I think this is especially crucial if you are in the process of explaining the EA principles or when relating to members who not yet are committed to the movement.
Is $750k the market price for 1 expected career change from someone at a top school, excluding compensation costs? Alternatively, is there no cheaper way to cause such a career change? IMO, this is the important question here: if there is a cheaper way, then paying $750k has an opportunity cost of >1 career changes.
edit: misinterpreted what comment above meant by “market price”
I think the market price is a bit higher than that. The mean impact from someone at a top school is worth over $750k/year, which means we should fund all interventions that produce a career change for $750k (unless they have large non-financial costs) since those have a <3 year payback period even if the students take a couple years to graduate or skill up.
In practice, dinners typically produce way more than 2% of a career change for $5k of dinners (33 dinners for 10 people at $15/serving). The situation at universities has non-monetary bottlenecks, like information transmission fidelity, qualified organizers, operational capacity, university regulations, etc., and most things that get you better use of those other resources and aren’t totally extravagant are worth funding, unless they have a hidden cost like optics or attracting greedy people.
Someone else in this thread found a report claiming that employers spend an average of ~$6,100 to hire someone at a US university. I also found this report saying that the average cost per hire in the United States is <$5,000, $15k for an executive. At 1 career = 10 jobs that’s $150,000/career for executive-level talent, or $180,000/career adjusting for inflation since the report was released.
I’m not sure how well those numbers reflect reality (the $15k/executive number looks quite low), but it seems at least fairly plausible that the market price is substantially less than $750k/career.
This line of reasoning is precisely what I’m claiming to be misguided. Giving you a gallon of water to drink allows you to live at least two additional days (compared to you having no water), which at $750k of impact/year (~$2000/day ) means, by your reasoning, that EA should fund all interventions that ensure you have 1 gallon of water for <=$4000, up to the amount you need to survive.
If water happened to be that expensive, that would be a worthwhile trade. But given the current market price of water (with the time cost of acquiring it included) being willing to pay anywhere near $4000/gallon is absurd.
In general, if you value something at $x, and its market price is $y, x only matters for deciding whether you should pay for the thing or not, not for deciding how much you should pay for it. If x >= y, then you should pay $y, otherwise you should pay $0.
It looks like I misunderstood a comment above. I meant “market price” as the rate at which CEA should currently trade between money and marginal careers, which is >$750k. I think you mean the average price at which other companies “in the market for talent” buy career changes, which is <$750k.
I think there isn’t really a single price at which we can buy infinite talent. We should do activities as cost-effective as other recruiters, but these can only be scaled up to a limited extent before we run into other bottlenecks. The existence of a cheaper intervention doesn’t mean we shouldn’t fund a more expensive intervention once the cheaper one is exhausted. And we basically want an infinite amount of talent, so in theory the activities that buy career changes at prices between $150k and $750k are also worth funding.
I think we can agree that
different activities have different cost-effectiveness, some of them substantially cheaper than $750k/career
we can use a basically infinite amount of talent, and the supply curve for career changes slopes upwards
we shouldn’t pay more than the market price for any intervention e.g. throw $100k at a university group for dinners when it produces the same effect as $5k spent on dinners
we should fund every activity that has a cost-effectiveness of better than $750k per career change (or whatever the true number is), unless we saturate our demand for talent and lower the marginal benefit of talent, or deplete much of our money and increase the marginal benefit of money
we are unlikely to saturate our demand for talent by throwing more money at EA groups because there are other bottlenecks
Because most of the interventions are much cheaper than $750k/career change, our average cost will be much less than $750k/career change