Founder of Ultra Philanthropy, an independent advisory that helps major donors—increasingly from tech and AI—give for maximum impact; fund manager of its mid-stage global health fund; Chair of Trustees at High Impact Athletes.
Jack Lewars
Karma: 2,702
Two Coefficient Givings beat one twice as big
I agree that there is certainly gatekeeping happening around AI donors and that this relates to issues around power concentration (related post coming later today on my Substack).
This post wasn’t particularly intended to address either problem.
One related issue is that most people on this forum probably do want some gatekeeping—that is, they wouldn’t want Bridgespan or Blue Meridian to have unfettered access to AI donors, because they don’t focus on cost-effectiveness enough.
If that’s true, it becomes an argument about where to draw the line, and most parties want the line to extend to exactly where they are but no further 😂
Hey Siobhan—thanks for this. These are definitely concerns I take seriously. You are right that my approach doesn’t solve the ‘access’ problem for lots of fundraisers, by implicitly gatekeeping access—that is a real issue, just not one I was trying to address here. I also recognise that “intermediary says using intermediaries is best strategy” is at best convenient.
However, I think you understate some of the things that undermine your position. Things like:
Donors are annoyed by this. You can see a public expression of this here (https://fundinganthropalypse.com/p/how-not-to-fundraise-from-anthropic/comment/296180159?utm_source=share&utm_medium=android&r=4b7xoz), but also talk to anyone who interacts with these donors regularly and you will hear the same. It’s also intuitive that this sort of outreach would be annoying.
You said donors are receptive to “thoughtful, targeted” outreach. This is a very generous description of cold emailing, or indeed any sort of unsolicited pitching anyone you have never met or talked to.
I think you understate the wider harms of this sort of approach.
I do take these concerns seriously and I am glad you raised them, but I don’t agree with your conclusions. This type of outreach very likely won’t work and is therefore net negative, in my view.
How (not) to fundraise from Anthropic staff
Congratulations on your many amazing achievements at GWWC Luke. You’ll be sorely missed but I’m also really excited for your next steps. I also know the team at GWWC will handle the transition with their usual professionalism.
This is an absolutely staggering challenge—anyone who has ever run even a half marathon will know how long that feels, and this is tens of them back to back!
I also know as a former colleague how much effort Emma puts into training, getting up at 5am every day so she can be at her desk ready to go for 9am.
This may not be the type of fundraising EA usually does but a pledge of even $1 per mile could help to crowd in new donors who wouldn’t give on a recurring basis to New Incentives. I really encourage promoting this campaign as HIA trials campaign fundraising through sport as a way to widen the base of effective donors.
Thank you!
Thanks mate, this is really lovely to read. Will you be at EAG London?
One for the World has had the Double The Donation widget for a couple of years. Unfortunately, it is about to become considerably more expensive, as they are upgrading everyone to a service called Match365. The plus side of this is that it will search your database for people who could be getting matching and email them proactively, and try to smooth the process of them getting matching. The downside is that it’s fairly expensive (~$4k/year), but I think it’d still be positive ROI, at least in the first year (probably with diminishing returns after that).
One thing to point out is that DTD massively overclaims its success rate. For example, it emails me every month saying “52 people claimed matching at their companies via the widget on your website”, but it turns out this means 52 people searched for matching, and I would estimate 1-2 per month actually end up accessing it (based on our overall volume of incoming corporate matching).
One final note - @Neil Warren as you are at Google, is there a reason you aren’t using Benevity to do your donations, which automatically adds matching? If this is new to you, see if you can log in at google.benevity.org
I believe GWWC do include ‘income you have sacrificed to do direct work’ in their pledge. (Correct me if I’m wrong, GWWC folks.)
Although, of course, your main argument (which I endorse) would still be true—you’d almost certainly still be earning substantially above the minimum wage in your region, and could most probably still give up 10% of your income, without materially affecting your wellbeing. So it’s a bit debatable if you should then feel like you’ve “fulfilled a pledge” by doing direct work at a lower salary.
Thanks for clarifying, Lizka. Just for clarity on what follows—I absolutely don’t think you’re thinking this through in bad faith, so I don’t want to come across as suggesting that. I do wonder if there might be some blind spots in your reasoning though, so I’m testing out the following to shine a light into those grey areas.
-
On your point 1 - what if was just “before my expected natural death/in my will”? I guess my point is: would you accept that it’s reasonable to pledge to give away financial resources that you ultimately didn’t need, after they had provided you with a safety net throughout your life?
More generally: I see your reasoning here but I think my reaction is still as follows. You want financial security to protect you in various scenarios, one of which is to protect against biases. There might be other ways to protect against these biases (your point 3.c is basically my point of view, for example—you definitely could blow the whistle on your employer anyway, even if you’ve donated some portion of your income over time, although I agree that this is harder on the margin if you’re more financially dependent on your employer).
Ultimately, in each of the scenarios you outline, you’re deciding whether to donate based on how you weigh one set of benefits (e.g. avoiding bias if your employer goes rogue) over another one (donating the money effectively).Some of your scenarios seem completely reasonable to me (e.g. “I could donate but I want to make sure I can provide financial security to my family first.”). Candidly, though, I think there is a strong whiff of motivated reasoning when the argument is something like “I could pledge to donate my money, but what if life expectancy has gone up to 150 in the future and I might need it?”, or “I could pledge to donate my money, but what if my employer turns out to be net negative and I don’t feel able to do anything about it because I didn’t keep my resources to get financial security?”.
It seems to me that if we allow this level of speculation in the justifications for not donating, we have to conclude that no one should ever be expected to donate anything, because clearly we could always construct some plausible scenario where they later regret doing that. (To put it a different way, this line of reasoning seems very vulnerable to reductio ad absurdum.)
In reality, lots of people donate 1-10% of their income, and I have never actually heard of someone who suffered significant hardship and said “if only I had not donated that money, I’d have been fine” (although NB: these people clearly could exist—I’m just sceptical that it’s true in the vast majority of cases).
For one thing, most affluent people in high-income countries spend at least 1-10% of their income on stuff they clearly don’t need. So for these arguments to be convincing, I’d also need to be convinced that you are spending 100% of your income either on essentials or on savings or investments. I have no idea about your financial circumstances, so that could be true. In my own life, it definitely isn’t true. So if I said “well, I could donate 10% of my income, but that would undermine my financial security”, this would be a false choice. I’d be better off not getting Ubers or takeaway food so often, joining a cheaper gym, drinking less wine and putting that money into savings; and donating the final 10% anyway. (Another way of putting this—people tend to theorise about this as if they are making decisions on the margin, but we are almost never, ever, ever actually acting on the margin.)
In summary, I guess every time we decide to keep a unit of resources to ourselves rather than donate it, we tell ourselves some form of justification for that. It’s just that I tend to view justifications that are second- or third-order insurances with suspicion, because they look and sound a lot like “I’m going to keep even the final 1% of my income to myself but it’s not for my own benefit, it’s for impact”.
Thanks for writing this—it definitely makes sense to me and resonates with another discussion we had in the Berlin EA office recently on what counts as “disposable income”.
I would just note three things:
If you are not pledging in order to build up financial security, would you consider pledging to donate these funds later in life if it turns out you don’t need them? A lot of your reasoning seems to be about self-insurance, which makes sense, but in the event that you don’t need the insurance (e.g. because you stay impactfully employed largely without a break for your whole career), it’d be great if you then donated those reserves steadily in your 70s and older. If you don’t want to do this, e.g. because you’d prefer to pass these reserves onto your kids, I think that does seem like it’s close to motivated reasoning (i.e. “it turns out the most impactful thing I can do is build up personal and intergenerational wealth” seems pretty convenient).
And then relatedly:
Depending on your personal circumstances, you might be surprised at how feasible it is to do the GWWC pledge and build up financial security at the same time. I definitely have less money set aside than if I hadn’t pledged, but I equally haven’t had to choose directly between pledging and being financially secure. Something to think about / factor in to your spreadsheet. The excellent Fi-lanthropy Calculator from Yield and Spread is great for showing how donating certain amounts extends your timelines for complete financial security, and it usually makes less of a difference than you think—https://www.yieldandspread.org/free-resources
And then finally, and a minor point:
FWIW, I think the only unconvincing scenario you mention is “if I was worried about EV/CEA/the usefulness of my work, I can imagine leaving without another opportunity lined up.”
In each scenario you mention, I think the correct trade off is “is the security for me not to suffer X more important than the benefit of donating this money?”. So when it comes to caring for your family, I think it’s fair enough to prioritise that pretty highly. But in this scenario, you could just carry on working at CEA while you do your job search, and I think it’s pretty indulgent to say “the ability to resign immediately rather than do my job and look for other jobs on the side is worth more than the value of donating money to help people in extreme poverty”. People become disillusioned in their jobs and start applying for other ones all the time, and I think you could do this as well, without undue hardship :-)
David Milliband, CEO of IRC, for an EA-adjacent view on how to be most effective in global health.
David can speak to why he doesn’t just follow EA orthodoxy in running a very large development org with a massive budget. These reasons might prove to be good or bad or just thought-provoking
Thanks, I agree with your clarification on the point I was trying to make
Thanks, this also made me pause. I can imagine some occasions where you might encourage employees to break the law (although this still seems super ethically fraught) - for example, some direct action in e.g. animal welfare. However, the examples here are ‘to gain recreational and productivity drugs’ and to drive around doing menial tasks’.
So if you’re saying “it isn’t always unambiguously ethically wrong to encourage employees to commit crimes” then I guess yes, in some very limited cases I can see that.
But if you’re saying “in these instances it was honest, honourable and conscientious to encourage employees to break the law” then I very strongly disagree.
I’m particularly interested in whether or not they were encouraged to break the law for people who had financial and professional power over them, which seems less nuanced than ‘how threatening is or isn’t this WhatsApp exchange’.
This gave me pause for thought, so thank you for writing it. I also respect that you likely won’t engage with this response to protect your own wellbeing.
I just want to raise, however, that I think you have almost completely failed to address a) the power dynamics involved; and b) the apparently uncontroversial claim that people were asked to break laws by people who had professional and financial power over them.
It seems impossible to square the latter with being “honest, honourable and conscientious”.
I understand that you are using this as an example of something you think is untrue and to demonstrate the asymmetrical burden of refuting a lot of claims.
However, if you’re prioritising, I would be most interested in whether it is true that you a) encouraged someone who you had financial and professional power over to drive without a driving licence; and b) encouraged someone in the same situation to smuggle drugs across international borders for you.
Whether or not they are formally an employee, encouraging people you have financial and professional power over to commit crimes unconnected to your mission is deeply unethical (and encouraging them to do this for crimes connected to your mission is also, at best, extremely ethically fraught).
Beautifully written, thank you for writing it
Thanks for writing this—extremely interesting.
Disclosure: I recently advised a client to fund a small grant to Dimagi, to do emergency cholera response in Nigeria.
Overall, this is a plausible way to activate a lot of small actors, while overcoming some of the difficulties of vetting very large numbers of small grants, or requiring overly burdensome due diligence from small actors.
I would also note that many of these interventions are externally validated, e.g. vitamin A, treating childhood malnutrition, kangaroo mother care. They are funded by organisations EAs have learnt to trust, like GiveWell (which specifically funds all three of these).
It is also often true that verification is a critical factor in delivering services or commodities, and that digital platforms like Connect can ensure more cost-effective delivery.
I would be interested to see actual cost-effectiveness figures for your table, rather than just rate per unit, if that is available? E.g. a cost per DALY averted estimate?