Everybody is (was?) always talking how important distillation is. A field of the size of AI safety needs efficient onboarding, and people writing a textbook with exercises & (afaik?) flashcards is overdue.
Impactful giving,
efficient funding.
Manifund offers charitable funding infrastructure designed to improve incentives, efficiency, and transparency.
Regranting
We delegate grantmaking budgets to regrantors who are experts in their fields, who then recommend grants based on their expertise. We ran one round of regranting in H2 2023, and are renewing the program for 2024!ACX Grants 2024
Scott Alexander's second grants round, this time directed partially through impact certificates. You can top off his grants or invest in the certificates now!Adrian Regenfuss
about 19 hours ago
Everybody is (was?) always talking how important distillation is. A field of the size of AI safety needs efficient onboarding, and people writing a textbook with exercises & (afaik?) flashcards is overdue.
Adam Gleave
2 days ago
Good paper that would be exciting to disseminate. This is a very cheap way of supporting that.
The paper is very theoretical in nature and may not end up improving AI safety in practice.
I asked Joar for an estimate of the travel costs, checked they were reasonable, and then added my own estimate for costs for food/sundries in Vienna.
Please disclose e.g. any romantic, professional, financial, housemate, or familial relationships you have with the grant recipient(s).
I previously supervised Joar Skalse on a related project. Joar is currently a contractor working on a separate project for my organization, FAR AI.
Austin Chen
4 days ago
Okay, I have set up this prediction market; let's move the discussion on operationalization there. I'm open to a cash bet of that size once we've figured out terms.
Vasco Grilo
4 days ago
I'd be open to a bet or prediction market of some kind like "will Alex raise further funding from an EA funder" if that seems cruxy to you.
@Austin , what do you think about the following bet. If Alex is not funded by Charity Entrepeneurship (one can apply to their incubation program with one's own idea) nor GiveWell until 1 year from now (i.e. until the end of 1 May 2025) in the context of his drone project, you send me 100 $. Otherwise, I send you 900 $ (= (1 - 0.1)/0.1*100), as I think that is less than 10 % likely.
Austin Chen
4 days ago
@vascoamaralgrilo shrug, once again, I think you're missing the point of Manifund (early stage interventions look very different than established ones; charity entrepreneurship charities might be a better comparison) and also missing the point of ACX Grants (which is mostly not to compete against GiveWell - see Scott's writeups and look at the other ACX Grants made).
I'm unfortunately not sure how to orient this conversation to be more productive here. I'd be open to a bet or prediction market of some kind like "will Alex raise further funding from an EA funder" if that seems cruxy to you. Otherwise I encourage you to come to Manifest, where both Alex and I will be; perhaps talking it out in person would be helpful!
Alex Toussaint
4 days ago
Bednets serve as a barrier against mosquitoes during sleep. Our goal is their eradication.
Your extensive arguments reduce analysis to a single metric, cost-effectiveness, to the point of being comical to the average reader. Most would happily pay $5/month to rid their city of mosquitoes instead of relying on nets. I'm not sure why you fail to see that.
Vasco Grilo
4 days ago
Thanks, @Austin.
I don't expect people like Alex to look at a complicated calculation that concludes "and therefore gene drives are 10x as effective as anti-mosquito drones" and think "okay, so I should drop what I'm doing and work on gene drives"
My Fermi estimate compared Alex's project with Against Malaria Foundation's (AMF's) bednests in DRC, not with gene drives. I am also not expecting Alex to drop the project, but I would appreciate it if @ScottAlexander / @acx-grants could explain their rationale for funding this project. Is the project being funded because it could realistically be funded by GiveWell in the future, i.e. because it could be at least 10 times as cost-effective as donating to people in extreme poverty via GiveDirectly?
Alex might just not believe the calculations -- for plausible reasons! He has a bunch more insight into what's good about drones than an outside observer.
Note my conclusion of AMF's bednets distribution in DRC being 15.3 times as cost-effective as anti-mosquito drones relied on Alex's own calculations. Are you suggesting Alex does not believe in his own calculations, or that my comparison is flawed in some way? I guess the latter, but then it would be nice if you could be more specific. I basically just relied on GiveWell's numbers, which are usually considered quite trustworthy. It is also the case that the real cost-effectiveness tends to be much lower than what is suggested by preliminary results produced by the people proposing the project. One has to control for a thinker's big idea (https://forum.effectivealtruism.org/posts/GW3cxBurTNKHs352S/controlling-for-a-thinker-s-big-idea). So, even if Alex's own calculations are uncertain, I think they will tend to overestimate the cost-effectiveness of the project.
I do think anti-mosquito drones have a realistic chance (eg >1%) of being sufficiently cost effective to be part of the fight against malaria
What do you think is the probability of GiveWell funding Alex's anti-mosquito drones this year? I guess it is around 1 %, and this seems to low to be worth funding. One could contribute to the seed funding of the charities incubated by Charity Entrepreneurship (https://www.givingwhatwecan.org/charities/charity-entrepreneurship-incubated-charities), and I am much more optimistic about their chances of becoming at least as cost-effective as GiveWell's interventions. Charity Entrepreneurship is quite aligned with maximising impact, and their founders go through a very selective process (e.g. I am pretty confident they would understand the concept of marginal cost-effectiveness).
His current set of skills sets him up really well for doing this particular project; founder-market fit is super important in making projects go well
I agree. However, I am thinking a project going well is a necessary rather than sufficient condition for funding it. If a project has the main goal of saving/improving lives, it (or a future iteration of it) still has to do it more cheaply than the best alternatives (like bednets). To illustrate, I am quite confident GiveDirectly makes unconditional cash transfers go well, but I think people wanting to save/improve lives as much as possible had better donate to GiveWell's funds, which fund projects at least 10 times as cost-effective as GiveDirectly.
I think you're underrating "because it sounds cool", I guess. To shore this up a bit more, sounding cool is an important factor in getting buzz, being able to raise further funding, get stakeholders to agree to participate, etc.
As above, I agree these considerations are relevant, but they are not enough. PlayPump sounded like a cool cost-effective way of supplying water to people in low income countries, and it attracted lots of funding on this basis, but it turned out to be way less cost-effective that the best alternatives (https://forum.effectivealtruism.org/topics/playpump). I believe greater scrutiny would have revealed this earlier.
I think the general approach of analyzing cost effectiveness just doesn't really make sense for projects at this stage, or is at best one of many different lens. Cost effectieness is a reasonable sanity check, maybe; other than that, I'd look to other signals like founder quality and project velocity
I think cost-effectiveness analyses make sense for early stage projects if they are easy to produce, and I suppose it was quite easy for Alex to get the estimate of 50 $/person/year in his 1st comment in this thread. In addition, I would say it would be worth it to spend more time coming up with a better estimate, as this would very much inform the requirements of the drone. It may be that the requirements for the drones to save lives as cost-effectively as GiveWell's interventions are very unrealistic (e.g. the cost per drone would have to be unreasonably low), which may prompt rethinking or dropping the project.
I agree other considerations besides cost-effectiveness are also relevant, but I feel like you are underweighting it because it seems quite quantifiable in the case of Alex's project (as Alex nicely illustrated). Elie Hassenfeld, who is the CEO of GiveWell (whose evaluations are considered the gold-standard in global health and development), said (https://podcast.clearerthinking.org/episode/096/elie-hassenfeld-why-it-s-so-hard-to-have-confidence-that-charities-are-doing-good/):
GiveWell cost- effectiveness estimates are not the only input into our decisions to fund malaria programs and deworming programs, there are some other factors, but they're certainly 80% plus of the case.
GiveWell said (https://forum.effectivealtruism.org/posts/CDt5ShpdABZRn8Tvi/my-quick-thoughts-on-donating-to-ea-funds-global-health-and?commentId=CXspdihsLLSepbyjC):
The numerical cost-effectiveness estimate in the spreadsheet is nearly always the most important factor in our recommendations, but not the only factor. That is, we don’t solely rely on our spreadsheet-based analysis of cost-effectiveness when making grants.
I assume I am also less optimistic than you about Alex. I get the impression Alex may want to go ahead with the project even if he was certain it would be less cost-effective than GiveWell's interventions, and his project absorbed funding which would otherwise go to GiveWell's interventions, which I think would be bad. I am also not sure Alex understands the concept of marginal cost-effectiveness based on his 2nd comment in this thread.
Austin Chen
5 days ago
@vascoamaralgrilo idk, I feel like we're drawing from pretty different worldviews here.
I don't expect people like Alex to look at a complicated calculation that concludes "and therefore gene drives are 10x as effective as anti-mosquito drones" and think "okay, so I should drop what I'm doing and work on gene drives"
Alex might just not believe the calculations -- for plausible reasons! He has a bunch more insight into what's good about drones than an outside observer. I do think anti-mosquito drones have a realistic chance (eg >1%) of being sufficiently cost effective to be part of the fight against malaria
His current set of skills sets him up really well for doing this particular project; founder-market fit is super important in making projects go well
I think you're underrating "because it sounds cool", I guess. To shore this up a bit more, sounding cool is an important factor in getting buzz, being able to raise further funding, get stakeholders to agree to participate, etc.
I think the general approach of analyzing cost effectiveness just doesn't really make sense for projects at this stage, or is at best one of many different lens. Cost effectieness is a reasonable sanity check, maybe; other than that, I'd look to other signals like founder quality and project velocity
Vasco Grilo
5 days ago
@alextouss Sorry for being too negative. I was assuming the goal of this project was preventing malaria deaths more cost-effectively than the most competitive alternatives (like bednets), but I appreciate others may want to fund the project for other reasons (e.g. because it sounds cool).
Thanks for the the feedback, @Austin . If I was proposing a project, I would find it super helpful if someone pointed out to me there are way more cost-effective interventions. I want to contribute to a better world as much as possible, so I would want to know if the money which would support my project could go towards other interventions that would e.g. save more lives. If I was persuaded by the arguments, I would consider halting the project, returning the money to the donors, and then encourage them to donate to a more promising project.
I agree one should account for the value of information of projects. However, do you think anti-mosquito drones have a realistic chance of being more cost-effective than GiveWell's top charities? If not, there is not much relevant information to be gained? Or are you referring to the value of information which is not related to the cost-effectiveness of the intervention, like Alex getting a better picture of what he wants to do in the future? Are there more cost-effective ways of gaining such other information (for example, for e.g. career plans, checking e.g. 80,000 Hours' career guide (https://80000hours.org/career-guide/))?
I understand Manifund has the goal of supporting early stage projects, I agree it makes sense to support novel interventions, and I personally do not think the platform should be restricted to cost-effective projects. However, I think users should also be welcome to point out whether they think a given project is cost-effective or not.
As a side note, I also think GiveWell's top charities are a pretty low bar. I believe the best animal welfare interventions are 1 k times as cost-effective (https://forum.effectivealtruism.org/posts/vBcT7i7AkNJ6u9BcQ/prioritising-animal-welfare-over-global-health-and#Corporate_campaigns_for_chicken_welfare_increase_nearterm_wellbeing_way_more_cost_effectively_than_GiveWell_s_top_charities).
Austin Chen
5 days ago
I agree with @alextouss; @vascoamaralgrilo, while I appreciate critiques that are intended to help out projects, I think your stance is unnecessarily combative here, or at least shortsighted/missing the point of Manifund. Manifund is primarily a platform for early stage projects, where there is a lot of information value to be gained from people like Alex trying new and novel things and then telling the world about their results. We don't generally expect interventions at these stages to compare favorably to established interventions like bednets on a spreadsheet (though it's a nice benefit if they do!)
I do also think that mosquito gene drives are extremely promising and I would be excited to see project proposals of that sort on Manifund -- if you are yourself interested in working on this or know of people are, encourage them to apply!
Tom O’Haire
5 days ago
@aaron Never too late to add technical skills. If you have an idea the tools are often available to make it happen.
Jason
6 days ago
Making a small offer for social-proof purposes. [I am generally hesitant to fund AI safety stuff, mainly because (a) there seems to be a lot of money in the space, (b) there is more risk of projects ending up net-negative than in other cause areas, (c) I feel less qualified to evaluate, and and (d) in light of these factors I am more inclined to defer to a project not having other funding on board. But I'm convinced that the social-proof element here outweighs my hesitancy at the small offer level.]
Emily Kerr-Finell
6 days ago
@Jason I see! Thank you for explaining more fully what you were asking about.
Even at this small scale, I'm committed to using this money to deliver the program to 1 person for each $500 invested, even down to the minimum funding level of $1K (which would cover two people). So $7.5K will let 15 people start microbusinesses (with all the accompanying positive impacts for each of those people/families, as I describe above.) It's possible that things will become even more cost effective over time -- but even at this scale, I can and will deliver it for $500/person (including covering $200 of their startup costs, effectively a cash transfer.) So I don't think the cost effectiveness calculation needs to include a promise of future efficiency.
Another important piece of the way I look at cost effectiveness is that once these new business owners are profitable (within 6 months), they'll pay for others to go through the program. So the $7.5K could, if I'm successful, fund 15 people immediately and then (hopefully) fund many more people than that with that repayment model.
The way I explain that above is -- each dollar isn’t “burned” but rather funds an ongoing (and hopefully even >1x) multiplier of the funds. And it’s reasonable to expect a 50-100x revenue yield for the migrant in just 12 months (for example, if the program costs $500 and he has revenues of $25K-$50K in his first year. )
It almost sounds too good to be true -- a $500 investment can 2x-5x a person's income within 30 days and give them a business they can grow over time -- and then that person will turn around and fund that same income lift for the next person in line. But these are the impacts I'm seeing (and why I'm devoting these years of my work life to the project, despite the opportunity costs.)
Thanks again for talking it over. I'm still working on the clearest ways to explain/conceptualize all of this and your questions help me refine that!
Alex Toussaint
6 days ago
@vascoamaralgrilo I'm not sure I get the reason for your hostility. Your donation criteria is yours, and other people have different ones. If you can't see the value of entirely getting rid of mosquitoes in cities for $50/year/person, than I can't do anything for you.
Jason
6 days ago
@briico Thanks, Emily! That makes sense. I think what I'm suggesting here is that evaluating the cost-effectiveness for the project as-stated ($25K for 50) is easier because we can factor in the direct benefits to the 50 participants, which are less speculative than the potential for future expansion. If it scales down to -- say, $7.5K for five participants due to fixed elements of staffing costs -- then the benefit comes predominantly from information value / benefit to potentially scaling the intervention down the road. By the classic EA metric, $7,500 is 1.5 lives worth of bednets ~ I take that mode of analysis seriously but not literally by the way. It's tragic that cost-effective global health is so underfunded that the bar is that high . . .
If the argument for funding is heavily linked to increasing the probability of future, more cost-effective rounds, then it's a somewhat more complex analysis for the donor. Most notably, we have to think about the probability that you'll be able to get funding for expansion down the road. So any evidence you could share about ability to get non-EA funding would be helpful, as that would allow an update on the "likelihood that expansion will happen" factor in the back-of-envelope cost-effectiveness calculation.