Giving 10% without money still loses on some of the signal. Not sure how big of a deal this is, but there is less of an incentive to choose those carefully if you’re not giving any money.
One option is to make them an actual offer of money. Even if they turn it down, it's acredible signal to outsiders that you really did believe in the project. And it also solves the hard cutoff problem, since you can offer a mediocre project $5k and a better one $100k, and everyone else can assign them prestige accordingly.
One option is to make them an actual offer of money. Even if they turn it down, it's acredible signal to outsiders that you really did believe in the project. And it also solves the hard cutoff problem, since you can offer a mediocre project $5k and a better one $100k, and everyone else can assign them prestige accordingly.
Yeah, the prestige-only awards are an interesting question!
Wild guess/knee-jerk response as an outsider to all this:
Is there any way to make a guess as to how much each incremental prestige-only award dilutes the signal? My wild guess/knee-jerk thought is to treat the prestige as if it is a fixed pool, effectively getting awarded to both the cash grant recipients and the prestige-only awards, and getting diluted by the total number of both awards.
So the criterion for the prestige-only award would be something like the financial criterion for the reduction in the potential remaining value of this year's prestige-only awards due to dilution by the incremental recipient. One might need to estimate the effective financial value of the prestige-only award to the recipient group.
Not sure how hard it is to not be affected by bias, but for the seal of approval/signal boost problem, I'd say that the categories would stay the same. $5000 would correspond to "Seal of confirmation that we would give them $5000 if they asked, but they didn't ask". The same for $10k and any other sum. You would judge the application as if they are asking for money, but then just don't give them money.
You can do this blinded. Ask applicants just wanting a seal of approval to say how much money they would want. Then rank all proposals, while not knowing which ones actually want the money, then go through them until you run out of money.
This seems like a sensible approach, though part of evaluating the grant is probably a question of how much money they are asking for, which may skew rankings.
With science funding evaporating in the US I’m left wondering if there’s some private ways to seek it.
ACX grants are really great to see and it’s been interesting to have followed these for the past few years.
I’m a grad student. I think my research is beneficial to humanity but perhaps not so obviously so that I felt it would qualify for this sort of thing. (As these grants seem particularly geared toward the most high impact projects).
Does anyone perhaps have advice of anywhere someone might look for this sort of thing? I’d be very happy to hear it :)
There's plenty of Foundations that give out money for scientific research. The secondary "how to get grants" is actually grubbing for them, like the universities and hospitals do. There are entire teams of people who exist to get donations. Some of the donations come with strings, like "please fund this specific project."
Having a very large building to name is one classic way of getting funding. People pay a lot for their name on a building.
Do you give grants to people outside the US? Because in many (most) countries, there is very little venture capital. Even in my own, Denmark, who ranks just below the US on the Freedom of Doing Business Index, it's just not very common. I'm sure many startups could appreciate your help
I'll write a more detailed comment on this but yes, Tornyol (anti-mosquito drones) is based in France and we couldn't have got the support from ACX Grants from a local VC.
Yes, there is, and it would have the added bonus of beeing more of a signal. A small SAFE would be the ideal, so as not to choke the cap table when the valuation is so small. (We have that in France, as a BSA-AIR).
My main point is that VCs, like potential employees or clients, have reading grids to evaluate pitches. The great thing I found about ACX Grants is that the grid is different, and encourages different kinds of projects. Founder obsession for a problem seems to be encouraged in ACX Grants, although it's clearly discouraged for very early VC funding. VCs like very well made slides, communication abilities, and beautiful people in general, while I've found no such bias for ACX Grants. Being based outside the US is a big minus for American VCs, but ACX Grants almost seems to be favoring it. VCs tend to think a lot by analogy (the Uber for X, the Cursor for Y ...) while I found ACX Grants to be much more thinking from first principles than the median VC I met.
I'm not criticizing the VC reading grid. It obviously comes from experience and it tends to work financially for them. But you have to remember that a large part of the decision comes down to the potential for a quite early (3-4 years) and billion-dollar exit option. Not all projects fit that and it's a good thing to support the other. The other advantage of it is that it selects founders that can go through the hoops of making their project fit the grid. That proves VCs the founders are capable of adapting their message to their interlocutors, which is highly necessary when raising further money, recruiting or discussing with any partner. That's something ACX Grants does not seem to value much.
All in all, ACX Grants is great in that it provides funding with a very unique reading grid, so it helps projects that could get no help anywhere else. That was our case a year ago, and is something I'm very grateful for now. Thank you again, ACX Grants, for your help.
> My main point is that VCs, like potential employees or clients, have reading grids to evaluate pitches. The great thing I found about ACX Grants is that the grid is different
"Reading grid" isn't a turn of phrase that I'm familiar with. Based on the way you're using it, I would guess it means something like "rubric" (per Merriam-Webster: "a guide listing specific criteria for grading or scoring academic papers, projects, or tests").
Is it a direct translation of a French expression, or a reference to something in French culture?
Yes, it comes from « une grille de lecture ». I thought the expression also existed in English. Thank you for the correction. The definition of « Rubric » seems to correspond to the French usage.
French people here, yeah "grille de lecture" (reading grid) is used to mean something like the paradigm, the set of criteria, the assessment process through which you'll view something.
If an entrepreneur is interested in a forprofit company that has a charitable side-effect, that entrepreneur isn't going to be at home neither with VC Capital (full for-profit) nor with traditional NGO funding (full charity).
The ideal should be an agreement like OpenAI, where there's a charity that "owns" shares in a for-profit company.
This points to a very interesting niche: ACX readers could provide VC-style funding, but more on the scale of the "microloan" ecosystem. I would be comfortable putting $5k into a pot that would be distibuted to ACX grantees, but with some possible diversified upside
These days I am not sure it's the availability of venture capital (that used to be a big issue 10, maybe even 5 years ago, there's a lot more of it now, at least in Switzerland) as much the much more difficult scaling compared to the US language, cultural, regulatory differences can be significant hurdles - single market or not.
Possibly related: I have done corporate VC investments (early stage) on rare occasion and would love to do more - the funding has never been the issue in our case (financial services), the credible projects are.
I think the question of nonfinancial grants should be a grant minimum of, say, $500 for individuals and $2,500 for organizations. (Or maybe higher?) Organizations and individuals can use the money, it preserves the grant structure, and in my experience it's very rare that when pressed people don't think they would have more time or be more effective with a bit more money. At the very least, that structure would cover the expected opportunity costs of applying for groups that could use the signal boost, but are effective and fully financed otherwise.
But even if you don't put in place a minimum, and allow signal-only grants the cost-benefit calculation should assume a cost of evaluators time and overhead - if the benefit if $100, the cost isn't zero, for an infinite return on investment, it's at least $100 for the time to consider and write up the grant, making it at best neutral. (This doesn't need to be accounted for explicitly for most grants, but the edge case changes things.)
I recommend against mixing investments with donations, since the legal structure get messy, and misaligned incentives are really tricky. (This would likely be different if you weren't doing well financially anyways, or planned to use the profits to start your own VC business.)
So I think the for-profit invest/donate issue is a chance for VCs to tell you if they think the opportunity is marginal for profitability but likely to succeed, or potentially profitable but higher risk then they'd invest in. At that point, they should either tell you they thin you should donate the money, or you can try a co-investment structure where you donate $1 for every dollar they invest. (If you want, depending on the distribution of outcomes they expect, this could be tied to a gentleman's agreement that the VC support future grant rounds with some proportion of the profits if the business makes them a large profit. I wouldn't try to formalize it, though, because that gets really messy.)
As a consultant at a for profit B Corp (meaning we are responsible to our values), I would be interested in pitching a specific project, rather than requesting investment. Would have the same structure supporting a project at a nonprofit.
I was thinking of investments to start a for profit. This sounds like pitching a contract for service, which seems fine to evaluate traditionally, i.e. similar to a nonprofit project.
I agree with this, the way I would phrase it is I think granting to for profit companies is fine (and you should just do granting) if you support their work and you want to lower their cost of capital to make it more likely to succeed.
There's an existing procedure for returning the money if things don't work out; clout-only applicants could become an alternate form of that, though it might be necessary to make a clear formal distinction between the two. Perhaps returning it same-day, plus an extra 10%, so whoever was next down the list can benefit from the actual cash.
It sounds quite essential to make a clear distinction because "I got grant money but had to return it because it didn't work out" sounds like the polar opposite of a positive endorsement ;)
I've got no knowledge on this, but hell if that'll stop me.
1. I don't know what the definition of "charitable aims" is here, but if it meets them, sure, do it.
2. I'd say invest, less so to make money and more so to keep more control over the use of the grant. Easy for a for-profit to decide something else would be more profitable. Sounds like it's a legal nightmare for some reason, but hey, I'm not the one doing it.
3. No further actions. You're not trying to make money, you're trying to get ideas running, and this is basically asking what happens if an idea gets running too successfully. The answer is, you give a polite golf clap and move on.
4. My take is, make them clear the grant, and then give the money to the next-highest contender. If ten people would get grants but No. 11 just wanted prestige, well, they didn't make the 10 grants, better luck next year. But if No. 5 just wanted prestige, give them the prestige, and give 5's money to No. 11.
5. Complicated web of lies, because I don't understand what you're asking and my policy is "when confused, confuse others."
For the seal of approval question, to solve the "getting 5k becomes easier than getting 0" problem: if you have determined some level of money equivalent to giving them the seal of approval, why not add that to the amount sought by every proposal? Getting 0 is evaluated as asking for 10k, getting 5k is evaluated as asking for 15k, etc., and your evaluation process is monotonous in the amount of money requested again.
A nonprofit company doesn't exactly have equity to give, as least as a 501(c) under the laws of every state i am aware of.
You could get a board seat, or shares in a nonprofit that has shares, but you could not get dividends on those shares.
Also, a nonprofit subject to 501(c) can't simply decide later to become a for-profit.
Finally, if and to the extent ACX is structured as a 501(c)(3), dividends received from an investment would be Unrelated Business Income and subject to tax.
Reg. 2: If you give forprofit companies money, I really think you should donate. Conflicts of interest could become really paralyzing and lead to misaligned incentives, I fear.
I think a VC charitable organization sounds promising, but I don’t think it’s your best personal fit. The legal hurdles will be immense and even if Manifund does the paperwork, you will be responsible to make decisions like whether to sue or whether to settle. You may have to testify on stand. My worry is it will take much more of you time than you expect and generate too much stress that gives you less time to use your other considerable talents. Your writing has made all of this possible and you could easily find this becoming a full time endeavor and take away from your writing and time with your family.
Direct cost of signal-boosting isn't zero, you just haven't got a metric for measuring costs denominated in reputation and signal-to-noise ratio as precisely as you can measure money. May have, in fact, actively worked to avoid developing sensitivity to conventional reputational costs, out of distaste for the deceptive behaviors associated with excessive concern for such. Perhaps the solution would be to formally consult a respected specialist in public relations, or a related field; they might very well have standard tools by which such costs could be assessed, and other off-the-shelf solutions, or at least domain expertise to warn you off from some common pitfalls.
"If we give forprofit companies money, should we donate or invest?"
Is lending a possible middle ground? Of course, it doesn't have the potential massive upside of equity, but you don't have to worry as much about (e.g.) "someone tr[ying] to backstab us to get more shares at our expense." The financial conflict of interest may be less, as you'd predominantly care only about whether the company did well enough to repay the loan.
"What happens if a nonprofit research organization that we donate to later decides to become a forprofit company?"
If it's that brazen, I'd consider reporting them to their state regulators and IRS (in the US). The prohibition against more-than-incidental private benefit (inurement) is one of the most fundamental obligations of a non-profit. I share your concern about "unenforced gentlemen’s agreement[s]" with organizations and organizational leaders who won't comply with their actual legal obligations!
On the organizational side, if I *hadn't* committed inappropriate private benefit, I'd be worried about giving equity in my sufficiently-unrelated new business to donors of the charity I used to work for. It sure looks like conceding that I had violated my duties to the non-profit, or had run the non-profit in a way to benefit my for-profit, or so on.
I want to start a cheaper but worse (not intentionally worse, but because their website seems very classy, I'm less informed, and I just found a cheap source of olivine near the Mediterranean coast and believe I can find a way to dump it to the sea somehow, and also find a way to document this so I can sell carbon credits to people/companies who want to buy, and preferably get subsidies from the Dutch state for each ton of CO2 I sequester) version of Vesta thus my answers are heavily biased:
On the signal, every marginal seal of approval you give out reduces the value of the deal to the people you already grant it to. That makes the cost-benefit analysis not one-sided and you could implement it by (say) fixing the number of approvals.
That being said, probably more trouble than it's worth?
A. If you don't, people who have big visions are unlikely to submit a proposal to you. Even apart from the financial motivation, it's quite difficult to grow a non-profit, since it's harder to get investments or loans for them.
B. I have too much experience with government agencies and contractors going off the rails because nobody cares if they lose money. For just one example of many, a boss of mine refused my plan to replace our supercomputing grid with a few nVidia machines that could do the same job, because he said it was better to spend millions of government dollars every year, than to spend $20,000 of our money. He wouldn't do it even if it were free, because spending millions of government dollars bought people and resources that could be diverted to other projects, and increased the status of the managers involved.
C. I was a member of MIT's entrepreneurship club in the 1990s, and met lots of VCs who told us how they invest, and I don't think they would invest in anything that you would want to fund.
First, the purpose of the VC ecosystem is to make as much money as possible as quickly as possible. Public benefit is lower-priority. Tech VCs do prioritize by coolness and audaciousness; I think they're VCs as much for status and excitement as for money.
Second, most VCs aren't interested in investing less than a million dollars at a time in anything. They would rather invest more than less. It doesn't take take 10 times as long to approve a $10-million investment as a $1-million investment, and they're generally more pressed for time than they are for money.
Third, the most-important factor in investment decisions, to every VC I've talked to, is how many of the founders attended either an Ivy League college, or else MIT or Stanford. That's literally the first or second thing they mention when talking about any company they invested in.
I donated money to a nonprofit that later turned into a for-profit company. At first I felt like a chump, but after that feeling wore off, I reflected, and it’s honestly one of the best outcomes of my donations. I think this is likely to be a good thing rather than a bad thing.
This is cope. You WERE a chump, or at least got successfully scammed. I can see how that could happen to anyone (or at least anyone who doesn't abide by deontological prohibitions against donating significant amount of money), but it is in your power to choose whether or not to learn from it.
Outcome 1. I donate money to a nonprofit, and they subsequently turn into a for-profit company, whose mission I wholeheartedly support, and it seems like they're on a promising path. But I don't make any profit myself.
Outcome 2. I donate money to a nonprofit, which tries earnestly to succeed at its mission, but does not.
I'd feel my generosity was taken advantage of in Outcome 1 but not in 2, so I'd prefer the latter.
But then again, at my level of wealth, I'm only ever purchasing warm fuzzies, not status or actual utility. I can see how other people might be in a different position.
Do you think you’d be a happier person if you got more fuzzies for helping others, and less from not being a chump? You can purchase utility pretty cheaply these days.
There's an old joke about how the best thing to do is change everyone's utility function to value hydrogen instead art, beauty, love, justice, truth, or whatever other complicated things humans actually do. This comment reminds me a bit of that. Perhaps the hypothetical person you describe would be happier, but I think he would no longer be meaningfully ME.
This assumes good-faith on the part of the business. Not "hire a lawyer so I don't do something illegal and go to jail"... I'm going to say that it's possible to make either nonprofits or for profit companies whose sole goal is scamming people, and making it sound REALLY good.
"A fool and his money" was actually published (ten years later or so), so it is possible even the most hopeless of projects can actually succeed.
I feel like related- I see that OpenAI is now a defense contractor ($200 million contract with the department of defense). Not quite the charitable aims of the original funders I think.
On whether to funding for profits would funge against VCs - VC has a specific hits-based business model that looks for companies with fast growth potential and a huge potential market so that one success can cover the loss of many failures. Not all impactful investments fit this mold, even if they'll likely be profitable. I currently helping start a new company with huge impact potential but a small market size (in financial terms), and it likely won't be venture backable. If you support companies like this, you might funge against mission aligned angel investors, but this is a generally a looser and less connected community, and the money is less countefactually valuable. You could also ask a similar question in the application around "do you think you could raise this money from someone else."
I think you ultimately just have to give money to confer a seal of approval, even if the money isn't needed. There's no understandable way for the recipient of a seal of approval to tout it without saying they received money in any normal venue. "I got approved to get $15,000 from ACX, but we didn't need the money so they didn't give it to us" might be fine for signaling to some in-group but will be functionally useless in the wider world, where you'll need to just say "grant recipient of ACX." If they really don't need the money, they can donate it, though I know the bookkeeping is what they'd really prefer to avoid.
I am the cofounder of Ovelle, an IVG company, which I am mostly doing because I believe broad access to IVG is a worthwhile goal. From what I’ve seen of VCs and fundraising, I would advocate for investing in private companies (even though that somewhat conflicts with my own interest in donating to them).
VCs really want to make money right now. This means they tend to avoid slightly longer-term investments that have large positive social impact. However, not all of these types of projects can be carried out in a charitable or academic setting. Take IVG, for example. Part of the reason I founded Ovelle is that no academic labs in the U.S. are currently making a serious effort to solve the problem. This is likely due to poor incentives, both in academia and in U.S. funding structures. So, if this technology is important, which I believe it is, I either need to change the incentive structure of academia and the NIH (which seems very hard) or start a company.
This seems like a significant gap in many areas, and the fact that such efforts are overlooked by VCs often means the returns are much stronger than expected (see, for example, OpenAI, SpaceX, Commonwealth Fusion, etc.). These returns can then be used, as you said, to fund future projects.
I’m not a legal expert, but I know this is possible. For example, the Aphorism Foundation (a private 501(c)(3) grant-making foundation that Reid Hoffman and his wife, Michelle Yee, established) invested about $50 million into the new for-profit OpenAI LP in 2018. U.S. private foundations are allowed to make program-related investments (PRIs) in for-profit entities when the activity directly advances a charitable purpose.
Surprised all the commentators are saying Yes to for-profit companies. Venture fundraising is almost always regarded as significantly easier than non-profit fundraising.
Non-profits have a handicap on them because they can't give equity in exchange for funds like for-profits can. They knowingly accept this poison pill, among others. They will often work the same as for-profits but with pennies to the dollar in $, and no prospect of striking it big with their non-existent equity. Grants help bridge this gap.
Non-profits generally speaking don't have as big an incentive to solve problems.
Take the Cobb Grill. Reduces fires, cheaper way to cook. Also can sell it in America, for those who don't want a full-size grill. You can even let people donate while buying one.
Compare this to a very efficient charity: "No more amputations" (delivers wool socks to the homeless). It needs to be able to find the homeless (as they're not the most sociable crowd, living out of dumpsters and hiding in daylight), and distribute the socks. It lacks the price signals telling it when more socks are needed, and must rely on begging for donations when they are (this is a lucky charity in that folks are willing to donate more around Christmas).
Your example is not a very efficient charity. There's plenty of examples of charities that are efficient. Nonprofits can also sell grills through earned income revenue models.
A nonprofit can run example 1, and a for-profit can run example two. You just laid out different business models. The nonprofit designation just is tax-designation and equity designation.
The nonprofit designation encourages a lot of "charity" upcharging, in order to keep the tax-designation. My employer charges ten times the rate for "charity" than they do for "ordinary work" (yes, that's because "ordinary work" is done under insurance bargaining, but still...).
Making investments requires formalization on your end, but also on theirs. Incorporating, defining ownership all the stakeholders, and so on, shows some additional level of commitment and interest. You had a lot of projects that died out because the recipient seemed to lose interest. I think this would help. Also, you will get good deal flow and becoming a self sustaining endowment might be really powerful long term. Im a VC, so I’m biased. But I think it’d be cool!
For 4, the intuitive answer to my mind is: presumably, for your grantees, there's a step where you ask them for payment information so you can actually send them the money; but, also presumably, you don't need to check that information prior to doing your grants. (Even if you validity-check the information pre-grantmaking or something to that effect, you can presumably assign the validity-check task to a third party who isn't directly responsible for the grantmaking part.) So, at the "specify payment information" step, include an option of "we don't actually need the money, keep it and use it for later grants instead". So you're granting blindly-to-whether-that's-their-position, and reasonably non-inflatedly—if you grant them $5000, that's a clear signal that it's a project you think is worth $5000-ungrantable-to-other-people-this-round-due-to-this-project-taking-up-that-slice-of-your-budget—but *also* you get to non-tax-complication-ladenly keep the money if they don't in fact need the money and just need the project-worth-$5000-as-gauged-by-ACX-grants seal of approval.
It seems like you like all of the options and generally like keeping flexibility.
What I think you should do is think hard about the mission (for now) of ACX grants, and if it involves more than one of; non-profit grants, for-profit investments, or no-money stamps of approval, then bisect (trisect) the ACX grants program into multiple arms.
Of course over time as you discover where you have the most success (however you measure that) you might shift focus between the three. But, if you have clean delineation (even if it shifts) it will get rid of a lot of problems of one arm eliminating the usefulness of the other arms or otherwise cannibalizing them in a way that is negative.
The US govt and universities already face this problem of what to do when they support noncommercial research that may turn into a startup.
US govt tends to be reasonable, but usually that's because they are effectively funding research which they hope will turn into a product that they themselves want to buy. Different incentives.
Universities tend to turn themselves into obstacles, so focused on making sure they get their "fair share" in terms of royalties or equity that it's self-sabotaging. Weird licensing restrictions or a screwed up cap table make life difficult when the company needs to raise again. The incentives seem to flow in this direction.
Maybe one could find the most reasonable university, or the most reasonable private research funder, and see what they do as a legal model. But overall, probably best to just be a chump. The point is to give away money to make good things happen, right?
So I run a for-profit company that's now investigating the possibility of applying for institutional grant funding. I don't know how typical our situation is, but maybe it'll be instructive.
We work with cities, counties, and special districts to overcome bureaucratic obstacles to delivering public services. With all this momentum building around abundance, we're hoping to run parallel improvement projects in a half dozen cities that could make dramatic improvements on, say, permits for multi-family housing units, create a playbook that other cities could use, and also do some joint storytelling or advocacy for state-level policy change. Happy to say more about what we're up to. I don't think we're the best candidate for ACX grants, because we'd be looking for larger dollar amounts, though I suppose ACX could be part of a mix of several funders.
We've been doing this work on a for-profit basis for one-off governments. So why look for funding?
- Cities really struggle to find the budget to pay for strategic work, when they're always in a budget crunch. I regularly scope work for cities who are excited to change but can't figure out how to fund it.
- The cities that DO figure out how to fund it aren't always best set up to change. We would get much better results if, instead of us doing the work for whomever is willing to pay, cities had to apply to US, and we could select those where we see the best readiness for change.
- The size and predictability that would come from institutional grants would let us put our part-time, hourly staff on salary, which would bring their per-hour cost down significantly. That's too risky to do when contract income is unpredictable.
- There's cool strategic work we can do with outside funding that no single client is willing to pay for. For example, creating a convening for multiple cities would be nightmarish in an environment where they're all individually contracting with us, but simple when it's outside funding.
We considered whether we should become a non-profit (or set up some kind of non-profit sister organization), but ultimately decided against it because:
- It's a tremendous pain.
- We have a for-profit model that works and we don't want to stop doing, it's just that it would have much more reach and strategic impact with outside funding.
- Having a for-profit and a non-profit running next to each other creates about a million conflicts of interest.
- We learned that big institutional funders find ways to give to for-profits.
Note that it's not like we have some new thing to invest in. We think that outside funding would create compounding social good, but not really compounding profits. VC funding wouldn't make any sense for us.
One thing we DID do is to register as a public benefit corporation. It's almost entirely symbolic, but it does require the owners of the company to accept a fiduciary responsibility to ensure that the company seeks the public good. This was a part of our mission anyway, so it was a no-brainer for us. It's a signal, but signaling is important.
So YMMV, but take that as a vote for, yes, give to for-profits.
You may have explored this already, but Bloomberg Philanthropies does a lot of work helping improve local governments (making them more efficient, more innovative, helping them help more people, etc.). They also do things like, for example, running conferences that bring multiple cities together. I don't know if they'd fund a for profit, but they might—and there are a *lot* of functionally independent sub-philanthropies, so I think that it's relatively likely that at least one of them would be open to the idea. https://www.bloomberg.org/government-innovation/
This is not ideal for a granting mechanism, mostly for the issues mentioned. I recommend having a second process called "ACX Awards" or what have you, that highlights excellent/innovative/whatever projects. You can either cap the number of awards, or have a quality threshold. Give the awardees a nominal award (e.g. like $100) and/or free travel to present at a special conference session (which would probably cost ~2k per awardee depending on the specifics, so a very small grant). The advantages:
-you can have a separate application process more suited to a ~non-monetary award, set the parameters that make the most sense, have specific annual targets, etc, while not disrupting the ACX Grants
-you would encourage deserving applications from people that see "grant" and say "oh I don't need the money" and move on
-with conference session or something similar, you provide increased prominence/networking opportunities for the awardees, increasing the value for them
Re 1-3: I think you should decide how much annoyance you are willing to bear to make for-profit investments partially as a function of how much money you have and how much you'd otherwise want to donate to startups.
Re 4: I think you should establish some minimum funding amount (there are evaluator time and other logistical difficulties with administering grants) and judge them based on the minimum amount. Then you can offer the grantees that much money with an easy option of refusing the money. Ideally you can even do this in a way that the grant evaluators by default do not even know if this is a grantee that originally asked for minimum vs 0 money.
Re 5: I believe I did warn you about the coherency difficulties! I will now award myself some Bayes points. :P
taking equity makes startups eligible for programs like SSBCI, which provides matches for funds in many US states. also, investors are more likely to invest with “early money.”
A few reflections from being on the board of a large foundation:
1. Some of our most impactful work has been via for profit companies. It's a small minority of what we do but it's been very powerful. On the margin we should probably do more of this.
We structure money to for profits in various ways depending on the circumstances - could be a grant, an equity investment, a loan, a contact, or some combination. Loans, equity investments, and loan guarantees are referred to as program related investments (PRIs).
But of course we are a large organization with staff so can afford to do complicated things. And there are potentially headaches associated with investments, taxes, etc. You probably only want to do equity investments if it's a fairly large check size, enough that it's worth some administrative work that will persist over time. If it's just a small amount of money, I would use one of the other options, which are simpler - loan, grant, or a contract. Also note that if you are directing private foundation money, there are some extra rules around monitoring grants to for profit enterprises to make sure the money is used for a charitable purpose.
2. We also sometimes give grants where the main point is the stamp of approval, not so much the money. I think you still want to give money for the reasons outlined in the post - basically it keeps you honest. Give whatever your minimum grant size is and have them compete with the other ideas. I think that's the only way to maintain high standards.
3. I wouldn't worry about the case where a nonprofit converts to a for profit. Open AI aside, this is incredibly rare and what you will want to do will depend a lot on the details.
> 1: Should we give money to forprofit companies with charitable aims?
I felt a strong visceral negative reaction to this. I don't think these grants should be for forprofit companies. If they later become forprofit, fine, but if the intent is forprofit or if it is already forprofit, I think no.
You already mention the VC industry as a source of funding. There are others, of course -- debt, other kinds of grants.
But IMO the bigger issue in funding forprofit is that you create perverse incentives for the people applying for these grants and for yourself. Moloch is extremely powerful.
On the former, you already mention that people want the signal boost more than other things, it is not hard to imagine people putting money into gaming your application (or spamming it) to try and get access to your network. AND those people will likely do so on the back of existing funding that they may already have from other places! Inb4: but why do nonprofits not have this problem? I think when you select for a class of people who have to spend some time not directly benefiting from their work, you also select out of people who will be susceptible to moloch in this way.
On the latter, if you start seeing returns -- getting shares, whatever -- you personally will have a hard time divorcing that consideration from your grant giving. There will always be some part of your consideration process that will think "well, maybe this is a good investment opportunity independent of the charity value" and over time I think that becomes dominating. And even if you personally are able to avoid this problem, I think the optics are still bad here
imo, go the forprofit route if you want to *make money* off it. Otherwise, you're going to corrupt any secondary or tertiary goals of the grant
On 4: IMO the majority of the value provided to the most successful grantees most likely already is non-monetary. Partially from raising their aspirations, partially from signaling.
When I try to put a price on the market value of the implicit endorsements Valuebase & Manifold got from ACX grants, it's a lot (>$100k, perhaps even >$500k if including other blog posts), def dwarfing the $55k and $20k grants.
There likely already is incentive to bid as needing funding, even if you could entirely make do without it, since the bump to user acquisition, hiring, partnerships or further fundraising is worth it.
My top two proposals would be:
a) Some sort of separate non-grant but endorsed category.
Pretty similar to the non-financial grants, but I think calling them that is a bit odd and might dilute signal. A link-post mention, or an honorable mention associated with ACX grants seems like it might capture much of the value.
b) Give them the money anyways.
Presumably this is going to be amongst your less useful dollars, but it's not necessarily amongst your less useful grants if in fact value from signaling already dominates, and unbundling might weaken this costly signal.
I’ll make the argument against nonfinancial grants:
1) The prestige comes almost entirely from the costly signal, which is that you’ve really given them money! There is little-to-no dilution effect here—even if you only give one nonfinancial grant, that grant will be less prestigious than the smallest “real” grant in that batch.
2) Money is useful to everybody! Filtering on those requesting a $0 grant is adverse selection for applicants who have not really thought about their project in enough detail yet to find something relevant to spend it on.
I think it makes sense to highlight interesting applicants separately, whether or not they need the money, but you shouldn’t expect it to be any more impactful than a links post shoutout, no matter how strongly you try to tie it to the real ACX grants.
I'm not sure what process you actually used for deciding how much money to give someone, but in principle you could evaluate cost-effectiveness (good per dollar) separately from the size of the grant, and then give a seal of approval based on cost-effectiveness.
Of course, if you measure cost-effectiveness by dividing by the size of the grant, and they're asking for $0, you get infinite cost-effectiveness. You'd need to evaluate based on the total resources the project is using, regardless of whether those resources are coming from you or from elsewhere. But I think you want to be doing that anyway, in order to avoid double-counting (e.g. if a project goes through 2 rounds of fundraising, you don't want the donors in each round to act as if their donation is responsible for 100% of whatever the project accomplishes, even if the project genuinely has a chance of dying at each round).
> How to handle applicants who want prestige / our “seal of approval”, but not money?
I hate to say it, but money IS prestige. Or at least the most potent form of prestige. There’s nothing better to signal value than giving something, well, of value for it. No shortcut is ever going to be as effective.
Seal of approval is a gradient just like money is.
There's a fifty dollar level (present in a massive post listing a short summary of everyone that qualifies) and a ten thousand dollar level (get a dedicated post featuring them where they explain what they're doing, what they need, have a Q&A in comments)
Pro for investing in for-profits: VC typically invest in companies with potential for 10x plus returns whose goal is to get to large scale quickly. There may be a gap in funding for socially good companies who require a slower growth model or who have smaller ambitions in terms of scale.
Also, given the influence of Scott and proven ability to judge talent, this may be more successful than other similar ventures.
I strongly disagree with this sentence: "Either those VCs will fund these companies (in which case our help isn’t required), or they will turn them down (and since they’re the experts, we should be skeptical that their rejects really deserve funding)". The median VC is not very smart, has a poor understanding of new technology, and has no idea how to value companies that don't neatly pattern-match an existing successful company. There's a reason Paul Graham refers to investors as "fickle idiots".
It seems to be that 4 could be solved by publishing a ranking (with added noise/fuzziness to make it easier for the ranker and more palatable for the rankee) of grant proposals. In this way everyone can receive the rank (no cutoffs/choices made by ACX), and the signaling is transparent, maximizing the value for the grantee
It seems to be that 4 could be solved by publishing a ranking (with added noise/fuzziness to make it easier for the ranker and more palatable for the rankee) of grant proposals. In this way everyone can receive the rank (no cutoffs/choices made by ACX), and the signaling is transparent, maximizing the value for the grantee
Here some thoughts on what I think would be easiest to you, least likely to create headaches, and give you a sense of fair-minded moral superiority.
1. For profit vs nonprofit doesn't matter to you, since that is a legal designation chosen upon the goals of the project. If you like the project, just don't count it for or against it being a nonprofit or profit. Make sure their choice fits the situation though.
2. Just donate. Keep life simple and then you can think of everything as a loss even easier.
3. Just leave it as a public monument of your civic mindedness.
4. You usually link post the stuff you find interesting anyway. If something is worthy of a link post, and yet is still so worthy that they deserve more than that, you can offer them a nominal grant for being awesome. Maybe think of it as an application to be in a linx post.
Prestige grants: there's an established principle in the rationalist community that when YOU put your own money down, YOU feel differently about it. When you give some amount of money to a cause, you're making a bet, and the money leaving your hand changes the way you think about it the interaction by giving you skin in the game.
If you choose to give a prestige grant with no money, it will change the way you think about the grant, sacrificing more than just the external signaling effect of the grant. The clarifying emails your send privately (which are helpful, I think) won't be of the same caliber, and the whole project will be not as good. Think of a prestige grant as a 'bet', and consider how much you're willing to bet on a grantee. What would be meaningful to you?
Giving 10% without money still loses on some of the signal. Not sure how big of a deal this is, but there is less of an incentive to choose those carefully if you’re not giving any money.
One option is to make them an actual offer of money. Even if they turn it down, it's acredible signal to outsiders that you really did believe in the project. And it also solves the hard cutoff problem, since you can offer a mediocre project $5k and a better one $100k, and everyone else can assign them prestige accordingly.
One option is to make them an actual offer of money. Even if they turn it down, it's acredible signal to outsiders that you really did believe in the project. And it also solves the hard cutoff problem, since you can offer a mediocre project $5k and a better one $100k, and everyone else can assign them prestige accordingly.
Yeah, the prestige-only awards are an interesting question!
Wild guess/knee-jerk response as an outsider to all this:
Is there any way to make a guess as to how much each incremental prestige-only award dilutes the signal? My wild guess/knee-jerk thought is to treat the prestige as if it is a fixed pool, effectively getting awarded to both the cash grant recipients and the prestige-only awards, and getting diluted by the total number of both awards.
So the criterion for the prestige-only award would be something like the financial criterion for the reduction in the potential remaining value of this year's prestige-only awards due to dilution by the incremental recipient. One might need to estimate the effective financial value of the prestige-only award to the recipient group.
'How to handle applicants who want prestige / our “seal of approval”, but not money?'
You should send them a seal plushie along with the endorsement.
A $10,000 plushie, and use the usual methods to evaluate whether it's worth it.
$10K would be chump change these days. Think of something like a $10M NFT depicting something exceptionally stupid.
Not sure how hard it is to not be affected by bias, but for the seal of approval/signal boost problem, I'd say that the categories would stay the same. $5000 would correspond to "Seal of confirmation that we would give them $5000 if they asked, but they didn't ask". The same for $10k and any other sum. You would judge the application as if they are asking for money, but then just don't give them money.
Just burn the money as a sign of good faith.
You can do this blinded. Ask applicants just wanting a seal of approval to say how much money they would want. Then rank all proposals, while not knowing which ones actually want the money, then go through them until you run out of money.
This seems like a sensible approach, though part of evaluating the grant is probably a question of how much money they are asking for, which may skew rankings.
With science funding evaporating in the US I’m left wondering if there’s some private ways to seek it.
ACX grants are really great to see and it’s been interesting to have followed these for the past few years.
I’m a grad student. I think my research is beneficial to humanity but perhaps not so obviously so that I felt it would qualify for this sort of thing. (As these grants seem particularly geared toward the most high impact projects).
Does anyone perhaps have advice of anywhere someone might look for this sort of thing? I’d be very happy to hear it :)
There's plenty of Foundations that give out money for scientific research. The secondary "how to get grants" is actually grubbing for them, like the universities and hospitals do. There are entire teams of people who exist to get donations. Some of the donations come with strings, like "please fund this specific project."
Having a very large building to name is one classic way of getting funding. People pay a lot for their name on a building.
Do you give grants to people outside the US? Because in many (most) countries, there is very little venture capital. Even in my own, Denmark, who ranks just below the US on the Freedom of Doing Business Index, it's just not very common. I'm sure many startups could appreciate your help
There's a reason that startups go to places where they can succeed, and a reason that startups in some places do better than others.
But capital flows across borders very easily, and if VCs thought startups in Denmark were worth investing in, they can invest there. And to some extent, they already do - https://www.basetemplates.com/investors/top-vc-investors-in-copenhagen
I'll write a more detailed comment on this but yes, Tornyol (anti-mosquito drones) is based in France and we couldn't have got the support from ACX Grants from a local VC.
Does France have good enough investment law that we could have invested in you if we'd wanted?
Yes, there is, and it would have the added bonus of beeing more of a signal. A small SAFE would be the ideal, so as not to choke the cap table when the valuation is so small. (We have that in France, as a BSA-AIR).
My main point is that VCs, like potential employees or clients, have reading grids to evaluate pitches. The great thing I found about ACX Grants is that the grid is different, and encourages different kinds of projects. Founder obsession for a problem seems to be encouraged in ACX Grants, although it's clearly discouraged for very early VC funding. VCs like very well made slides, communication abilities, and beautiful people in general, while I've found no such bias for ACX Grants. Being based outside the US is a big minus for American VCs, but ACX Grants almost seems to be favoring it. VCs tend to think a lot by analogy (the Uber for X, the Cursor for Y ...) while I found ACX Grants to be much more thinking from first principles than the median VC I met.
I'm not criticizing the VC reading grid. It obviously comes from experience and it tends to work financially for them. But you have to remember that a large part of the decision comes down to the potential for a quite early (3-4 years) and billion-dollar exit option. Not all projects fit that and it's a good thing to support the other. The other advantage of it is that it selects founders that can go through the hoops of making their project fit the grid. That proves VCs the founders are capable of adapting their message to their interlocutors, which is highly necessary when raising further money, recruiting or discussing with any partner. That's something ACX Grants does not seem to value much.
All in all, ACX Grants is great in that it provides funding with a very unique reading grid, so it helps projects that could get no help anywhere else. That was our case a year ago, and is something I'm very grateful for now. Thank you again, ACX Grants, for your help.
> My main point is that VCs, like potential employees or clients, have reading grids to evaluate pitches. The great thing I found about ACX Grants is that the grid is different
"Reading grid" isn't a turn of phrase that I'm familiar with. Based on the way you're using it, I would guess it means something like "rubric" (per Merriam-Webster: "a guide listing specific criteria for grading or scoring academic papers, projects, or tests").
Is it a direct translation of a French expression, or a reference to something in French culture?
Yes, it comes from « une grille de lecture ». I thought the expression also existed in English. Thank you for the correction. The definition of « Rubric » seems to correspond to the French usage.
French people here, yeah "grille de lecture" (reading grid) is used to mean something like the paradigm, the set of criteria, the assessment process through which you'll view something.
Replying to this to give more visibility:
If an entrepreneur is interested in a forprofit company that has a charitable side-effect, that entrepreneur isn't going to be at home neither with VC Capital (full for-profit) nor with traditional NGO funding (full charity).
The ideal should be an agreement like OpenAI, where there's a charity that "owns" shares in a for-profit company.
This points to a very interesting niche: ACX readers could provide VC-style funding, but more on the scale of the "microloan" ecosystem. I would be comfortable putting $5k into a pot that would be distibuted to ACX grantees, but with some possible diversified upside
These days I am not sure it's the availability of venture capital (that used to be a big issue 10, maybe even 5 years ago, there's a lot more of it now, at least in Switzerland) as much the much more difficult scaling compared to the US language, cultural, regulatory differences can be significant hurdles - single market or not.
Possibly related: I have done corporate VC investments (early stage) on rare occasion and would love to do more - the funding has never been the issue in our case (financial services), the credible projects are.
I think the question of nonfinancial grants should be a grant minimum of, say, $500 for individuals and $2,500 for organizations. (Or maybe higher?) Organizations and individuals can use the money, it preserves the grant structure, and in my experience it's very rare that when pressed people don't think they would have more time or be more effective with a bit more money. At the very least, that structure would cover the expected opportunity costs of applying for groups that could use the signal boost, but are effective and fully financed otherwise.
But even if you don't put in place a minimum, and allow signal-only grants the cost-benefit calculation should assume a cost of evaluators time and overhead - if the benefit if $100, the cost isn't zero, for an infinite return on investment, it's at least $100 for the time to consider and write up the grant, making it at best neutral. (This doesn't need to be accounted for explicitly for most grants, but the edge case changes things.)
I recommend against mixing investments with donations, since the legal structure get messy, and misaligned incentives are really tricky. (This would likely be different if you weren't doing well financially anyways, or planned to use the profits to start your own VC business.)
So I think the for-profit invest/donate issue is a chance for VCs to tell you if they think the opportunity is marginal for profitability but likely to succeed, or potentially profitable but higher risk then they'd invest in. At that point, they should either tell you they thin you should donate the money, or you can try a co-investment structure where you donate $1 for every dollar they invest. (If you want, depending on the distribution of outcomes they expect, this could be tied to a gentleman's agreement that the VC support future grant rounds with some proportion of the profits if the business makes them a large profit. I wouldn't try to formalize it, though, because that gets really messy.)
As a consultant at a for profit B Corp (meaning we are responsible to our values), I would be interested in pitching a specific project, rather than requesting investment. Would have the same structure supporting a project at a nonprofit.
I was thinking of investments to start a for profit. This sounds like pitching a contract for service, which seems fine to evaluate traditionally, i.e. similar to a nonprofit project.
I agree with this, the way I would phrase it is I think granting to for profit companies is fine (and you should just do granting) if you support their work and you want to lower their cost of capital to make it more likely to succeed.
Maybe applications could be blinded w.r.t. whether they are financial or clout-only.
There's an existing procedure for returning the money if things don't work out; clout-only applicants could become an alternate form of that, though it might be necessary to make a clear formal distinction between the two. Perhaps returning it same-day, plus an extra 10%, so whoever was next down the list can benefit from the actual cash.
It sounds quite essential to make a clear distinction because "I got grant money but had to return it because it didn't work out" sounds like the polar opposite of a positive endorsement ;)
The profit motive ruins everything. Stay away, far away.
I've got no knowledge on this, but hell if that'll stop me.
1. I don't know what the definition of "charitable aims" is here, but if it meets them, sure, do it.
2. I'd say invest, less so to make money and more so to keep more control over the use of the grant. Easy for a for-profit to decide something else would be more profitable. Sounds like it's a legal nightmare for some reason, but hey, I'm not the one doing it.
3. No further actions. You're not trying to make money, you're trying to get ideas running, and this is basically asking what happens if an idea gets running too successfully. The answer is, you give a polite golf clap and move on.
4. My take is, make them clear the grant, and then give the money to the next-highest contender. If ten people would get grants but No. 11 just wanted prestige, well, they didn't make the 10 grants, better luck next year. But if No. 5 just wanted prestige, give them the prestige, and give 5's money to No. 11.
5. Complicated web of lies, because I don't understand what you're asking and my policy is "when confused, confuse others."
> 2. I'd say invest, less so to make money and more so to keep more control over the use of the grant.
That would involve a large additional commitment of Scott's time and effort. It might be a good argument for just making grants.
For the seal of approval question, to solve the "getting 5k becomes easier than getting 0" problem: if you have determined some level of money equivalent to giving them the seal of approval, why not add that to the amount sought by every proposal? Getting 0 is evaluated as asking for 10k, getting 5k is evaluated as asking for 15k, etc., and your evaluation process is monotonous in the amount of money requested again.
A nonprofit company doesn't exactly have equity to give, as least as a 501(c) under the laws of every state i am aware of.
You could get a board seat, or shares in a nonprofit that has shares, but you could not get dividends on those shares.
Also, a nonprofit subject to 501(c) can't simply decide later to become a for-profit.
Finally, if and to the extent ACX is structured as a 501(c)(3), dividends received from an investment would be Unrelated Business Income and subject to tax.
Reg. 2: If you give forprofit companies money, I really think you should donate. Conflicts of interest could become really paralyzing and lead to misaligned incentives, I fear.
I think a VC charitable organization sounds promising, but I don’t think it’s your best personal fit. The legal hurdles will be immense and even if Manifund does the paperwork, you will be responsible to make decisions like whether to sue or whether to settle. You may have to testify on stand. My worry is it will take much more of you time than you expect and generate too much stress that gives you less time to use your other considerable talents. Your writing has made all of this possible and you could easily find this becoming a full time endeavor and take away from your writing and time with your family.
Direct cost of signal-boosting isn't zero, you just haven't got a metric for measuring costs denominated in reputation and signal-to-noise ratio as precisely as you can measure money. May have, in fact, actively worked to avoid developing sensitivity to conventional reputational costs, out of distaste for the deceptive behaviors associated with excessive concern for such. Perhaps the solution would be to formally consult a respected specialist in public relations, or a related field; they might very well have standard tools by which such costs could be assessed, and other off-the-shelf solutions, or at least domain expertise to warn you off from some common pitfalls.
"If we give forprofit companies money, should we donate or invest?"
Is lending a possible middle ground? Of course, it doesn't have the potential massive upside of equity, but you don't have to worry as much about (e.g.) "someone tr[ying] to backstab us to get more shares at our expense." The financial conflict of interest may be less, as you'd predominantly care only about whether the company did well enough to repay the loan.
"What happens if a nonprofit research organization that we donate to later decides to become a forprofit company?"
If it's that brazen, I'd consider reporting them to their state regulators and IRS (in the US). The prohibition against more-than-incidental private benefit (inurement) is one of the most fundamental obligations of a non-profit. I share your concern about "unenforced gentlemen’s agreement[s]" with organizations and organizational leaders who won't comply with their actual legal obligations!
On the organizational side, if I *hadn't* committed inappropriate private benefit, I'd be worried about giving equity in my sufficiently-unrelated new business to donors of the charity I used to work for. It sure looks like conceding that I had violated my duties to the non-profit, or had run the non-profit in a way to benefit my for-profit, or so on.
I want to start a cheaper but worse (not intentionally worse, but because their website seems very classy, I'm less informed, and I just found a cheap source of olivine near the Mediterranean coast and believe I can find a way to dump it to the sea somehow, and also find a way to document this so I can sell carbon credits to people/companies who want to buy, and preferably get subsidies from the Dutch state for each ton of CO2 I sequester) version of Vesta thus my answers are heavily biased:
1- Yes
2- Donation
On the signal, every marginal seal of approval you give out reduces the value of the deal to the people you already grant it to. That makes the cost-benefit analysis not one-sided and you could implement it by (say) fixing the number of approvals.
That being said, probably more trouble than it's worth?
"Should we give money to for-profits?"
I think so.
A. If you don't, people who have big visions are unlikely to submit a proposal to you. Even apart from the financial motivation, it's quite difficult to grow a non-profit, since it's harder to get investments or loans for them.
B. I have too much experience with government agencies and contractors going off the rails because nobody cares if they lose money. For just one example of many, a boss of mine refused my plan to replace our supercomputing grid with a few nVidia machines that could do the same job, because he said it was better to spend millions of government dollars every year, than to spend $20,000 of our money. He wouldn't do it even if it were free, because spending millions of government dollars bought people and resources that could be diverted to other projects, and increased the status of the managers involved.
C. I was a member of MIT's entrepreneurship club in the 1990s, and met lots of VCs who told us how they invest, and I don't think they would invest in anything that you would want to fund.
First, the purpose of the VC ecosystem is to make as much money as possible as quickly as possible. Public benefit is lower-priority. Tech VCs do prioritize by coolness and audaciousness; I think they're VCs as much for status and excitement as for money.
Second, most VCs aren't interested in investing less than a million dollars at a time in anything. They would rather invest more than less. It doesn't take take 10 times as long to approve a $10-million investment as a $1-million investment, and they're generally more pressed for time than they are for money.
Third, the most-important factor in investment decisions, to every VC I've talked to, is how many of the founders attended either an Ivy League college, or else MIT or Stanford. That's literally the first or second thing they mention when talking about any company they invested in.
I donated money to a nonprofit that later turned into a for-profit company. At first I felt like a chump, but after that feeling wore off, I reflected, and it’s honestly one of the best outcomes of my donations. I think this is likely to be a good thing rather than a bad thing.
This is cope. You WERE a chump, or at least got successfully scammed. I can see how that could happen to anyone (or at least anyone who doesn't abide by deontological prohibitions against donating significant amount of money), but it is in your power to choose whether or not to learn from it.
Which is better.
Outcome 1. I donate money to a nonprofit, and they subsequently turn into a for-profit company, whose mission I wholeheartedly support, and it seems like they're on a promising path. But I don't make any profit myself.
Outcome 2. I donate money to a nonprofit, which tries earnestly to succeed at its mission, but does not.
I'd feel my generosity was taken advantage of in Outcome 1 but not in 2, so I'd prefer the latter.
But then again, at my level of wealth, I'm only ever purchasing warm fuzzies, not status or actual utility. I can see how other people might be in a different position.
(I agree broadly with the message of https://www.lesswrong.com/posts/3p3CYauiX8oLjmwRF/purchase-fuzzies-and-utilons-separately)
Do you think you’d be a happier person if you got more fuzzies for helping others, and less from not being a chump? You can purchase utility pretty cheaply these days.
There's an old joke about how the best thing to do is change everyone's utility function to value hydrogen instead art, beauty, love, justice, truth, or whatever other complicated things humans actually do. This comment reminds me a bit of that. Perhaps the hypothetical person you describe would be happier, but I think he would no longer be meaningfully ME.
This assumes good-faith on the part of the business. Not "hire a lawyer so I don't do something illegal and go to jail"... I'm going to say that it's possible to make either nonprofits or for profit companies whose sole goal is scamming people, and making it sound REALLY good.
"A fool and his money" was actually published (ten years later or so), so it is possible even the most hopeless of projects can actually succeed.
I feel like related- I see that OpenAI is now a defense contractor ($200 million contract with the department of defense). Not quite the charitable aims of the original funders I think.
On whether to funding for profits would funge against VCs - VC has a specific hits-based business model that looks for companies with fast growth potential and a huge potential market so that one success can cover the loss of many failures. Not all impactful investments fit this mold, even if they'll likely be profitable. I currently helping start a new company with huge impact potential but a small market size (in financial terms), and it likely won't be venture backable. If you support companies like this, you might funge against mission aligned angel investors, but this is a generally a looser and less connected community, and the money is less countefactually valuable. You could also ask a similar question in the application around "do you think you could raise this money from someone else."
I think you ultimately just have to give money to confer a seal of approval, even if the money isn't needed. There's no understandable way for the recipient of a seal of approval to tout it without saying they received money in any normal venue. "I got approved to get $15,000 from ACX, but we didn't need the money so they didn't give it to us" might be fine for signaling to some in-group but will be functionally useless in the wider world, where you'll need to just say "grant recipient of ACX." If they really don't need the money, they can donate it, though I know the bookkeeping is what they'd really prefer to avoid.
I am the cofounder of Ovelle, an IVG company, which I am mostly doing because I believe broad access to IVG is a worthwhile goal. From what I’ve seen of VCs and fundraising, I would advocate for investing in private companies (even though that somewhat conflicts with my own interest in donating to them).
VCs really want to make money right now. This means they tend to avoid slightly longer-term investments that have large positive social impact. However, not all of these types of projects can be carried out in a charitable or academic setting. Take IVG, for example. Part of the reason I founded Ovelle is that no academic labs in the U.S. are currently making a serious effort to solve the problem. This is likely due to poor incentives, both in academia and in U.S. funding structures. So, if this technology is important, which I believe it is, I either need to change the incentive structure of academia and the NIH (which seems very hard) or start a company.
This seems like a significant gap in many areas, and the fact that such efforts are overlooked by VCs often means the returns are much stronger than expected (see, for example, OpenAI, SpaceX, Commonwealth Fusion, etc.). These returns can then be used, as you said, to fund future projects.
I’m not a legal expert, but I know this is possible. For example, the Aphorism Foundation (a private 501(c)(3) grant-making foundation that Reid Hoffman and his wife, Michelle Yee, established) invested about $50 million into the new for-profit OpenAI LP in 2018. U.S. private foundations are allowed to make program-related investments (PRIs) in for-profit entities when the activity directly advances a charitable purpose.
Surprised all the commentators are saying Yes to for-profit companies. Venture fundraising is almost always regarded as significantly easier than non-profit fundraising.
Non-profits have a handicap on them because they can't give equity in exchange for funds like for-profits can. They knowingly accept this poison pill, among others. They will often work the same as for-profits but with pennies to the dollar in $, and no prospect of striking it big with their non-existent equity. Grants help bridge this gap.
Keep grants to non-profits.
Non-profits generally speaking don't have as big an incentive to solve problems.
Take the Cobb Grill. Reduces fires, cheaper way to cook. Also can sell it in America, for those who don't want a full-size grill. You can even let people donate while buying one.
Compare this to a very efficient charity: "No more amputations" (delivers wool socks to the homeless). It needs to be able to find the homeless (as they're not the most sociable crowd, living out of dumpsters and hiding in daylight), and distribute the socks. It lacks the price signals telling it when more socks are needed, and must rely on begging for donations when they are (this is a lucky charity in that folks are willing to donate more around Christmas).
Your example is not a very efficient charity. There's plenty of examples of charities that are efficient. Nonprofits can also sell grills through earned income revenue models.
A nonprofit can run example 1, and a for-profit can run example two. You just laid out different business models. The nonprofit designation just is tax-designation and equity designation.
The nonprofit designation encourages a lot of "charity" upcharging, in order to keep the tax-designation. My employer charges ten times the rate for "charity" than they do for "ordinary work" (yes, that's because "ordinary work" is done under insurance bargaining, but still...).
Making investments requires formalization on your end, but also on theirs. Incorporating, defining ownership all the stakeholders, and so on, shows some additional level of commitment and interest. You had a lot of projects that died out because the recipient seemed to lose interest. I think this would help. Also, you will get good deal flow and becoming a self sustaining endowment might be really powerful long term. Im a VC, so I’m biased. But I think it’d be cool!
For 4, the intuitive answer to my mind is: presumably, for your grantees, there's a step where you ask them for payment information so you can actually send them the money; but, also presumably, you don't need to check that information prior to doing your grants. (Even if you validity-check the information pre-grantmaking or something to that effect, you can presumably assign the validity-check task to a third party who isn't directly responsible for the grantmaking part.) So, at the "specify payment information" step, include an option of "we don't actually need the money, keep it and use it for later grants instead". So you're granting blindly-to-whether-that's-their-position, and reasonably non-inflatedly—if you grant them $5000, that's a clear signal that it's a project you think is worth $5000-ungrantable-to-other-people-this-round-due-to-this-project-taking-up-that-slice-of-your-budget—but *also* you get to non-tax-complication-ladenly keep the money if they don't in fact need the money and just need the project-worth-$5000-as-gauged-by-ACX-grants seal of approval.
It seems like you like all of the options and generally like keeping flexibility.
What I think you should do is think hard about the mission (for now) of ACX grants, and if it involves more than one of; non-profit grants, for-profit investments, or no-money stamps of approval, then bisect (trisect) the ACX grants program into multiple arms.
Of course over time as you discover where you have the most success (however you measure that) you might shift focus between the three. But, if you have clean delineation (even if it shifts) it will get rid of a lot of problems of one arm eliminating the usefulness of the other arms or otherwise cannibalizing them in a way that is negative.
I'd be happy to help.
The US govt and universities already face this problem of what to do when they support noncommercial research that may turn into a startup.
US govt tends to be reasonable, but usually that's because they are effectively funding research which they hope will turn into a product that they themselves want to buy. Different incentives.
Universities tend to turn themselves into obstacles, so focused on making sure they get their "fair share" in terms of royalties or equity that it's self-sabotaging. Weird licensing restrictions or a screwed up cap table make life difficult when the company needs to raise again. The incentives seem to flow in this direction.
Maybe one could find the most reasonable university, or the most reasonable private research funder, and see what they do as a legal model. But overall, probably best to just be a chump. The point is to give away money to make good things happen, right?
So I run a for-profit company that's now investigating the possibility of applying for institutional grant funding. I don't know how typical our situation is, but maybe it'll be instructive.
We work with cities, counties, and special districts to overcome bureaucratic obstacles to delivering public services. With all this momentum building around abundance, we're hoping to run parallel improvement projects in a half dozen cities that could make dramatic improvements on, say, permits for multi-family housing units, create a playbook that other cities could use, and also do some joint storytelling or advocacy for state-level policy change. Happy to say more about what we're up to. I don't think we're the best candidate for ACX grants, because we'd be looking for larger dollar amounts, though I suppose ACX could be part of a mix of several funders.
We've been doing this work on a for-profit basis for one-off governments. So why look for funding?
- Cities really struggle to find the budget to pay for strategic work, when they're always in a budget crunch. I regularly scope work for cities who are excited to change but can't figure out how to fund it.
- The cities that DO figure out how to fund it aren't always best set up to change. We would get much better results if, instead of us doing the work for whomever is willing to pay, cities had to apply to US, and we could select those where we see the best readiness for change.
- The size and predictability that would come from institutional grants would let us put our part-time, hourly staff on salary, which would bring their per-hour cost down significantly. That's too risky to do when contract income is unpredictable.
- There's cool strategic work we can do with outside funding that no single client is willing to pay for. For example, creating a convening for multiple cities would be nightmarish in an environment where they're all individually contracting with us, but simple when it's outside funding.
We considered whether we should become a non-profit (or set up some kind of non-profit sister organization), but ultimately decided against it because:
- It's a tremendous pain.
- We have a for-profit model that works and we don't want to stop doing, it's just that it would have much more reach and strategic impact with outside funding.
- Having a for-profit and a non-profit running next to each other creates about a million conflicts of interest.
- We learned that big institutional funders find ways to give to for-profits.
Note that it's not like we have some new thing to invest in. We think that outside funding would create compounding social good, but not really compounding profits. VC funding wouldn't make any sense for us.
One thing we DID do is to register as a public benefit corporation. It's almost entirely symbolic, but it does require the owners of the company to accept a fiduciary responsibility to ensure that the company seeks the public good. This was a part of our mission anyway, so it was a no-brainer for us. It's a signal, but signaling is important.
So YMMV, but take that as a vote for, yes, give to for-profits.
You may have explored this already, but Bloomberg Philanthropies does a lot of work helping improve local governments (making them more efficient, more innovative, helping them help more people, etc.). They also do things like, for example, running conferences that bring multiple cities together. I don't know if they'd fund a for profit, but they might—and there are a *lot* of functionally independent sub-philanthropies, so I think that it's relatively likely that at least one of them would be open to the idea. https://www.bloomberg.org/government-innovation/
Yup, definitely aware of Bloomberg and What Works Cities. They definitely have a Way They Do Things already, but they're certainly on the list!
RE #4 "seal of approval" issue:
This is not ideal for a granting mechanism, mostly for the issues mentioned. I recommend having a second process called "ACX Awards" or what have you, that highlights excellent/innovative/whatever projects. You can either cap the number of awards, or have a quality threshold. Give the awardees a nominal award (e.g. like $100) and/or free travel to present at a special conference session (which would probably cost ~2k per awardee depending on the specifics, so a very small grant). The advantages:
-you can have a separate application process more suited to a ~non-monetary award, set the parameters that make the most sense, have specific annual targets, etc, while not disrupting the ACX Grants
-you would encourage deserving applications from people that see "grant" and say "oh I don't need the money" and move on
-with conference session or something similar, you provide increased prominence/networking opportunities for the awardees, increasing the value for them
Re 1-3: I think you should decide how much annoyance you are willing to bear to make for-profit investments partially as a function of how much money you have and how much you'd otherwise want to donate to startups.
Re 4: I think you should establish some minimum funding amount (there are evaluator time and other logistical difficulties with administering grants) and judge them based on the minimum amount. Then you can offer the grantees that much money with an easy option of refusing the money. Ideally you can even do this in a way that the grant evaluators by default do not even know if this is a grantee that originally asked for minimum vs 0 money.
Re 5: I believe I did warn you about the coherency difficulties! I will now award myself some Bayes points. :P
taking equity makes startups eligible for programs like SSBCI, which provides matches for funds in many US states. also, investors are more likely to invest with “early money.”
A few reflections from being on the board of a large foundation:
1. Some of our most impactful work has been via for profit companies. It's a small minority of what we do but it's been very powerful. On the margin we should probably do more of this.
We structure money to for profits in various ways depending on the circumstances - could be a grant, an equity investment, a loan, a contact, or some combination. Loans, equity investments, and loan guarantees are referred to as program related investments (PRIs).
But of course we are a large organization with staff so can afford to do complicated things. And there are potentially headaches associated with investments, taxes, etc. You probably only want to do equity investments if it's a fairly large check size, enough that it's worth some administrative work that will persist over time. If it's just a small amount of money, I would use one of the other options, which are simpler - loan, grant, or a contract. Also note that if you are directing private foundation money, there are some extra rules around monitoring grants to for profit enterprises to make sure the money is used for a charitable purpose.
2. We also sometimes give grants where the main point is the stamp of approval, not so much the money. I think you still want to give money for the reasons outlined in the post - basically it keeps you honest. Give whatever your minimum grant size is and have them compete with the other ideas. I think that's the only way to maintain high standards.
3. I wouldn't worry about the case where a nonprofit converts to a for profit. Open AI aside, this is incredibly rare and what you will want to do will depend a lot on the details.
> 1: Should we give money to forprofit companies with charitable aims?
I felt a strong visceral negative reaction to this. I don't think these grants should be for forprofit companies. If they later become forprofit, fine, but if the intent is forprofit or if it is already forprofit, I think no.
You already mention the VC industry as a source of funding. There are others, of course -- debt, other kinds of grants.
But IMO the bigger issue in funding forprofit is that you create perverse incentives for the people applying for these grants and for yourself. Moloch is extremely powerful.
On the former, you already mention that people want the signal boost more than other things, it is not hard to imagine people putting money into gaming your application (or spamming it) to try and get access to your network. AND those people will likely do so on the back of existing funding that they may already have from other places! Inb4: but why do nonprofits not have this problem? I think when you select for a class of people who have to spend some time not directly benefiting from their work, you also select out of people who will be susceptible to moloch in this way.
On the latter, if you start seeing returns -- getting shares, whatever -- you personally will have a hard time divorcing that consideration from your grant giving. There will always be some part of your consideration process that will think "well, maybe this is a good investment opportunity independent of the charity value" and over time I think that becomes dominating. And even if you personally are able to avoid this problem, I think the optics are still bad here
imo, go the forprofit route if you want to *make money* off it. Otherwise, you're going to corrupt any secondary or tertiary goals of the grant
On 4: IMO the majority of the value provided to the most successful grantees most likely already is non-monetary. Partially from raising their aspirations, partially from signaling.
When I try to put a price on the market value of the implicit endorsements Valuebase & Manifold got from ACX grants, it's a lot (>$100k, perhaps even >$500k if including other blog posts), def dwarfing the $55k and $20k grants.
There likely already is incentive to bid as needing funding, even if you could entirely make do without it, since the bump to user acquisition, hiring, partnerships or further fundraising is worth it.
My top two proposals would be:
a) Some sort of separate non-grant but endorsed category.
Pretty similar to the non-financial grants, but I think calling them that is a bit odd and might dilute signal. A link-post mention, or an honorable mention associated with ACX grants seems like it might capture much of the value.
b) Give them the money anyways.
Presumably this is going to be amongst your less useful dollars, but it's not necessarily amongst your less useful grants if in fact value from signaling already dominates, and unbundling might weaken this costly signal.
I’ll make the argument against nonfinancial grants:
1) The prestige comes almost entirely from the costly signal, which is that you’ve really given them money! There is little-to-no dilution effect here—even if you only give one nonfinancial grant, that grant will be less prestigious than the smallest “real” grant in that batch.
2) Money is useful to everybody! Filtering on those requesting a $0 grant is adverse selection for applicants who have not really thought about their project in enough detail yet to find something relevant to spend it on.
I think it makes sense to highlight interesting applicants separately, whether or not they need the money, but you shouldn’t expect it to be any more impactful than a links post shoutout, no matter how strongly you try to tie it to the real ACX grants.
I'm not sure what process you actually used for deciding how much money to give someone, but in principle you could evaluate cost-effectiveness (good per dollar) separately from the size of the grant, and then give a seal of approval based on cost-effectiveness.
Of course, if you measure cost-effectiveness by dividing by the size of the grant, and they're asking for $0, you get infinite cost-effectiveness. You'd need to evaluate based on the total resources the project is using, regardless of whether those resources are coming from you or from elsewhere. But I think you want to be doing that anyway, in order to avoid double-counting (e.g. if a project goes through 2 rounds of fundraising, you don't want the donors in each round to act as if their donation is responsible for 100% of whatever the project accomplishes, even if the project genuinely has a chance of dying at each round).
> How to handle applicants who want prestige / our “seal of approval”, but not money?
I hate to say it, but money IS prestige. Or at least the most potent form of prestige. There’s nothing better to signal value than giving something, well, of value for it. No shortcut is ever going to be as effective.
Seal of approval is a gradient just like money is.
There's a fifty dollar level (present in a massive post listing a short summary of everyone that qualifies) and a ten thousand dollar level (get a dedicated post featuring them where they explain what they're doing, what they need, have a Q&A in comments)
Pro for investing in for-profits: VC typically invest in companies with potential for 10x plus returns whose goal is to get to large scale quickly. There may be a gap in funding for socially good companies who require a slower growth model or who have smaller ambitions in terms of scale.
Also, given the influence of Scott and proven ability to judge talent, this may be more successful than other similar ventures.
I strongly disagree with this sentence: "Either those VCs will fund these companies (in which case our help isn’t required), or they will turn them down (and since they’re the experts, we should be skeptical that their rejects really deserve funding)". The median VC is not very smart, has a poor understanding of new technology, and has no idea how to value companies that don't neatly pattern-match an existing successful company. There's a reason Paul Graham refers to investors as "fickle idiots".
It seems to be that 4 could be solved by publishing a ranking (with added noise/fuzziness to make it easier for the ranker and more palatable for the rankee) of grant proposals. In this way everyone can receive the rank (no cutoffs/choices made by ACX), and the signaling is transparent, maximizing the value for the grantee
It seems to be that 4 could be solved by publishing a ranking (with added noise/fuzziness to make it easier for the ranker and more palatable for the rankee) of grant proposals. In this way everyone can receive the rank (no cutoffs/choices made by ACX), and the signaling is transparent, maximizing the value for the grantee
Here some thoughts on what I think would be easiest to you, least likely to create headaches, and give you a sense of fair-minded moral superiority.
1. For profit vs nonprofit doesn't matter to you, since that is a legal designation chosen upon the goals of the project. If you like the project, just don't count it for or against it being a nonprofit or profit. Make sure their choice fits the situation though.
2. Just donate. Keep life simple and then you can think of everything as a loss even easier.
3. Just leave it as a public monument of your civic mindedness.
4. You usually link post the stuff you find interesting anyway. If something is worthy of a link post, and yet is still so worthy that they deserve more than that, you can offer them a nominal grant for being awesome. Maybe think of it as an application to be in a linx post.
there's way too much profiteering in the world as is, I strongly believe that ACX grants should be nonprofit
Prestige grants: there's an established principle in the rationalist community that when YOU put your own money down, YOU feel differently about it. When you give some amount of money to a cause, you're making a bet, and the money leaving your hand changes the way you think about it the interaction by giving you skin in the game.
If you choose to give a prestige grant with no money, it will change the way you think about the grant, sacrificing more than just the external signaling effect of the grant. The clarifying emails your send privately (which are helpful, I think) won't be of the same caliber, and the whole project will be not as good. Think of a prestige grant as a 'bet', and consider how much you're willing to bet on a grantee. What would be meaningful to you?