Thanks to
- Daniel Yu (Founder of Wasoko, which just raised $125 million - this is his idea!)
- Sjir Hoeijmakers (Senior Researcher at Founders Pledge and Chair at EA Netherlands)
- Ben Clifford (Product Manager of EffectiveAltruism.org)
and many others for their comments and thoughts.
Epistemic status: we’ve given this some thought, but could easily be wrong about things and some sections are rough. We would love to hear your thoughts and suggestions.
Opportunities in Effective Altruism
- EA needs entrepreneurs. At the EA meta level, there is currently a general supply of good ideas and funding, but a relative lack of “builders” and “operators” as has been detailed in previous analyses of the ecosystem. Explicit support for EA-aligned entrepreneurs has the benefit of driving both strong financial returns and leadership development for individuals who can contribute their operational learnings to the EA movement in future roles even if their particular ventures are unsuccessful.
- ‘Building-to-give’ seems underindexed within EA. All the top 10 richest people in the world, as well as the two individuals (Dustin Moskovitz & Sam Bankman Fried) who account for the vast majority of EA funding, are entrepreneurs who built large successful businesses. For the right people and ideas, entrepreneurship can have huge expected financial value. However, there currently appear to be far fewer EA members explicitly earning to give through entrepreneurship as compared with other career paths, such as quantitative trading.
- EA-aligned funding is currently heavily concentrated. The vast majority comes from a small number of billionaires, with most funding currently disbursed by one organisation: Open Philanthropy. While there are clearly benefits to scale, this concentration could potentially pose a long-term risk to the movement: we are dependent on a small number of individuals, whose judgements (which will include some bias) have disproportionate weight on the movement’s decisions. If we can increase the size and proportion of broad-based funding for the movement, we can not only increase the EA movement’s resources but also mitigate the risk from such a high concentration of funding sources (though it may be that there are cheaper ways to diversify EA funding than encouraging building-to-give).
A potential solution: The Snowball Fund
The Snowball Fund would be a community-funded endowment whose primary growth strategy would be investments into early-stage startups with EA-aligned founders. Though the startups funded may have some direct positive impact in their value proposition, this will not be essential, so long as the startup’s offering is not clearly welfare-negative (e.g. a tobacco company). The fund would aim to accumulate financial returns through the exponential growth potential of early-stage startups. The fund would then be reinvested in further cohorts of early-stage companies. If things go well, the compounding would generate an increasingly large pool of capital, from which a portion of returns could eventually start being disbursed to support EA organizations directly once a milestone of assets under management and returns has been achieved.
At the same time, we hope the Snowball Fund will deepen community relationships with entrepreneurial EAs while also exposing non-EA entrepreneurs to EA. Anecdotally, many entrepreneurs bounce off the EA community, which has been described as overly risk-averse, too academic and theoretical, and ‘biased against action’. The Snowball Fund could increase EA’s soft power among entrepreneurs. For example, one of three co-founders might be EA-adjacent and suggest the Snowball Fund as an investor. The other co-founders and employees would be exposed to Snowball Fund and effective altruism; the funding announcement might provide further exposure through which other entrepreneurs/tech people read about it (and, by extension, EA). The founder is - and feels - more connected to the community and other founders within it. And maybe their next venture is an EA megaproject. This is speculative and there are potentially other avenues to achieving the same result (e.g. a retreat for exited commercial founders to brainstorm ideas with EA thought leaders), however there is likely an advantage in providing explicit EA community support to entrepreneurs early in their journeys rather than trying to approach successful founders once they’re already billionaires. Fundamentally the Snowball Fund should provide a low-cost experiment, since investments would be made on a standard structure once a qualifying amount of co-investment has been secured (see below) and the financial cost is just any difference in expected return between this set of investments and the next best investment (i.e. the S&P 500).
The Snowball Fund will be open to individuals with any amount of capital to contribute and ultimately accountable to the full base of donors in a structure akin to shareholders in public companies. Individual Snowball Fund donors will retain voting rights to collectively decide key governance decisions of the Fund when necessary, similar to company shareholders. The weight of an individual’s voting rights will be capped (e.g. 2% maximum) to ensure broad-based decision power across the donor base.
To reduce overhead and potential bias in making investment decisions, the Snowball Fund will make minority-stake investments into qualifying startups on a standard structure. The initial proposed structure is to invest $100,000 at prevailing terms for any startup that has at least $300,000 in other funding secured. This amount may be adjusted over time and additional heuristics may be developed to govern follow-on investments into successful ventures. This rules-based minority investor strategy has been highly successful for other organized communities that have established venture funds (e.g. Endeavor Catalyst Fund). This derisks our investment strategy by ensuring we’re backing startups that have commercial validation by investors beyond just the EA community. With this model, the only additional due diligence necessary on our side is to confirm that at least one founder is EA-aligned and that the company idea is not obviously negative for global welfare.
Founders who receive funding from the Snowball Fund will be required to sign the Founders Pledge or equivalent, which further ensures alignment and allocation of the founder’s personal returns towards effective causes in the case of a successful exit.
Initial investments will be overseen by an investment committee composed of donors with venture growth and investment experience. Eventual disbursement to EA projects would be researched by dedicated grantmakers and proposed for approval to management ultimately accountable to the donor base. We expect to be diversified across a variety of worldviews, though the specific details of our approach remain to be decided.
Why should EAs build such a fund?
Some people we have spoken to seem to think that almost all of any impact we may have will come from increasing the engagement of entrepreneurial EAs and recruiting new entrepreneurs to EA. We are not sure whether the effect will be large enough to outweigh the benefits of the potential financial gains of the Fund and/or of helping these EA entrepreneurs succeed in their current venture. Maybe having the Fund as an investor and the communication + community feeling that comes with that will be enough to significantly increase the probability that someone launches directly impactful projects later, but maybe not: it might not change things much compared to people e.g. just taking the Founders Pledge and already being a part of the EA community.
The way we now conceptualize the project is that it's a low-cost, low-risk experiment that might yield a range of benefits, none of which clearly stands out above the others / none of which is super high value by itself in expectation, but some of them may turn out to be great (e.g. supporting the next Sam Bankman-Fried, or recruiting an entrepreneur into EA because they hear about the Fund, or the Fund becoming a complementary funder/coordination mechanism in the EA funding landscape), and the median case may still be cost-effective enough given the presence of investment returns, so it's probably worth trying.
The financial cost of this experiment is any difference between our (risk-adjusted) returns and those of whatever else the money would have been invested in, which makes it very cheap per entrepreneur, provided that the opportunity cost of time involved is sufficiently low. The opportunity cost of time may be higher, but are limited to a 30-minute call per founder, plus any time spent fundraising and later disbursing and governing the disbursement of funds as grants. Daniel Yu has committed to initially running the fund.
—
VC provides the opportunity to deliver outsized returns relative to other asset classes of investment. Entrepreneurship can create huge amounts of wealth extraordinarily quickly, so we think that risk-neutral EAs can benefit from the fat-tailed distribution of wealth generation through VC investments.
There are several potential drawbacks to becoming an LP in existing funds, which would ultimately limit our upside potential relative to our direct investment opportunities. First, becoming an LP in proven high-performing VC funds is extraordinarily competitive, so we're unlikely to get allocation in established funds. Second, being an LP in a fund also gives us no control over the investment decision process, meaning a fund we directed money into could theoretically make an investment against EA interests; we intend to be systematic, but still have veto discretion if something looks terrible.
We believe the Snowball Fund has a preferential dealflow advantage with EA founders based on the two following factors:
- Value alignment: If the venture is successful, Snowball Fund proceeds will be reinvested to multiply resources further and provide a guaranteed pool of funds for high-impact initiatives in the future.
- Access to other EA founders/community: It seems reasonable that EAs would be particularly keen to assist startups funded by this fund. EAs have broad expertise and networks that can likely assist ventures across a variety of fields and stages.
Why not instead focus on directly impactful ventures? Firstly, if we assume profit and impact each follow sufficiently uncorrelated heavy-tailed distributions, the optimal strategy to maximise social and financial returns seems to be to allocate some proportion of your portfolio to maximising impact and the rest to maximising financial return, rather than trying to use all the portfolio to maximise jointly for both. Secondly, in practice it’s difficult to know what for-profits should be founded from an impact perspective (especially for longtermists). For-profit startups have a significant social impact, but perhaps not in comparison to the top interventions we consider as EAs. In that case, for-profit entrepreneurship may be best understood as a tool for creating huge amounts of wealth that can then be used to fund high-impact work.
Summary of potential benefits & conditions for them to hold
- Capturing above-market returns through (a) fund and (b) founders’ Founders Pledge (or similar) donations
- (a) requires VC premium or EA entrepreneur premium
- (b) requires significant additional pledges as a result of the fund’s existence
- Helping EA entrepreneurs succeed at their ventures
- Requires additionality of investment (i.e. EA entrepreneurs counterfactually benefit from our investment)
- Increasing engagement of entrepreneurial EAs (e.g. preventing value drift)
- Requires fund facilitating connections and mentorships that add value
- Must provide more than just money
- Recruit new entrepreneurs to EA
- Requires Fund to be a reason for someone to enter EA
- Get information on early-stage EA entrepreneurs as a class
- If it works, this would prove naysayers wrong and allow for scaling up funding (possibly with outside, non-EA capital)
- Requires a large pool of investments (the larger, the stronger the inference)
- Diversify in the EA grantmaking landscape to reduce financial risk
- Requires the Fund to have a large-ish pool of funders and grow large
- Add value to the EA grantmaking landscape
- Requires the Fund to be a competent grantmaker and add something new, e.g. a new decision mechanism
Key questions on theory of change & red-teaming
- What are the most promising ways to have additionality?
- What is the opportunity cost of time spent on the Fund?
- One 30-minute call to verify that the person is qualified (EA-adjacent, etc.) and get to know them.
- Little diligence is required because of systematic approach. Diligence tends to take most of an investor’s time.
- More time might be required for fundraising - and later, disbursing funds through grants and setting up good governance structures for that. But we already have some initial capital committed, so this is not an immediate issue.
- The aim of reducing value drift among EA entrepreneurs overlaps with Founders Pledge.
- There could be room for some healthy competition in this space.
- Snowball is looking for earlier stage entrepreneurs than FP currently targets.
- But, if using the FP pledge infrastructure, why reinvent the wheel? What can Snowball do for early stage entrepreneurs that FP is choosing not to?
- Concern that this could become a low-tier VC, with best founders going elsewhere - especially if we can’t deliver much extra value to entrepreneurs.
- Will this fund save EA-aligned entrepreneurs a significant amount of time finding other investors, or are we mostly just replacing a value-neutral investor that would have stepped in otherwise (given other funding has been secured already), and is the direct investment support of the Fund hence negligible?
- Or is the value of the "support" mentioned here more about the "explicitness" of it, increasing the standing of EA-aligned entrepreneurs in the EA community and/or connecting them more to it, by which they may be more likely to engage in EA-type work and donations later?
- Endeavor Catalyst Fund is a good example up to a point. Certainly it’s an impressive community and the fund seems to be doing well. But their definition of 'impact' seems to be transforming the world (in any way) using entrepreneurship. That's fine, but note that it aligns with the funds mission: the fund does well, it funds more community building, bigger/better community helps the endeavor entrepreneurs do well. Whereas the cycle with an EA fund could be: fund does well, funds normal EA community building, EA community becomes better at grants, ability of EA entrepreneurs to succeed stays flat.
- Possibly negative diversification premium.
- The total EA portfolio is overweight tech & startups, both financially and socially, compared to normal population. How much of a negative that is is arguable.
- So there is probably some (relative) negative return to potentially causing more EA money to be allocated to startups/tech, if it wouldn’t be allocated to these sectors otherwise.
- Possibly low/negative VC return premium.
- Some see current tech/VC valuations indicating lower returns going forward.
- Mechanism by which systematic fund leads to new EA megaprojects is unclear/weak.
- Relies on an idea of soft power and reputation, which seems plausible but might be hard to model.
- Worth betting on still?
- It’s possible we move some potential nonprofit founders to for-profits - the for-profit startup ecosystem is already higher-status outside EA, and we might contribute to this.
- My intuition is that this is unlikely, but I’m unsure how to think about quantifying this risk. (First pass: doing a quick mental scan, most of the nonprofit founders I know seem unlikely to have thought of starting for-profits instead.)
Next steps
Given the limited downside, we believe the most appropriate next steps are to establish a small fund on a trial basis to make initial investments aligned to the thesis above. To this end, a total of $1M has already been committed by Daniel Yu and other EA members to support initial investments. As with most entrepreneurial endeavors, we expect the operating model to adjust over time as we receive feedback from the market.
There are several key practical questions that we aim to explore further and report back on:
- What is the right legal structure for the fund?
- Tax-deductible charitable organizations can in fact make VC investments to drive returns for future activities (VC investments are common practice for university endowments).
- Initial discussions with lawyers suggest we set up charitable entities in the UK, US, and EU that can then facilitate investments
- When can the fund get started?
- Lawyers suggest we can start making investments through a UK entity immediately, though tax-deductible donor status will take up to six months to confirm. Initial $1M is ready to be deployed prior to full tax deductible status.
- What is the initial operating structure of the fund?
- Propose an initial structure of three volunteer investment committee members to review opportunities and unanimously agree on investments. We expect this to evolve over time
- When is the right time to start using fund proceeds to support EA programming work?
- Current belief is that returns should be reinvested until a decision is made by the donor ‘shareholders’ once meaningful returns into the fund are seen, which would probably not be for 5+ years.
Thanks for writing this up. I might suggest targeting much more precisely than "startups"/"entrepreneurs".
You say:
My personal experience with this complaint is that it's usually a feature of reality, not the EA community. Even the most applied AI safety research agendas involve a lot more thinking about linear algebra than they do building a sales team.
Taking Future Fund's list of projects as an example: most of them seem to require very substantial technical domain expertise. E.g. creating better PPE is probably the sort of project which would benefit from an entrepreneur who has started a medical device manufacturing business, but might be even more bound on scientists who can research different materials etc. (and is very much not bound on e.g. entrepreneurs who know a bunch about B2B marketing).
Some of this comment comes from my personal experience: it's pretty rare that I feel like my impact could be increased through being better at the sorts of things which made me better at running a tech startup, and much more common that there is a specific technical skill set which I lack.
tl;dr: most "entrepreneurs" aren't a good fit for starting the highest impact projects, and a lot of what EAs describe as "startups" are wildly different from sort of thing VCs fund.
I think this is an important point and there's some truth to it: EA 'startups' are different from venture-backed startups. When you say 'targeting', what do you mean here exactly, and do you have any ideas on how to do this without increasing time costs a lot? An advantage of our proposed approach is that it's time-cheap because it's systematic (with some caveats, we offer investment if and only if someone else has). But if there were a cheap way to make this better at finding the people most helpful for EA projects, that would be great.
---
If the claim is that, as a result of the differences between EA and venture-backed 'startups', it's not worth doing something like this to attract entrepreneurs to EA, I'm not sure I agree:
1. There are some common skills to building both types of startup - you're creating a new org to do something untested with a small number of people and often not much money, so there are at least some similarities. Maybe the overlapping skillset is quite small. Do you think you had already built that skillset before leaving your startup, such that any extra skill-building in that overlapping core wouldn't be helpful for your work now? I'm unsure about this because I haven't really built a startup of either type.
2. It's not just reality but also the EA community.
a. Not everything can or should be tested by building a minimum viable product and seeing what reality throws at it, but for some types of problems this is useful. And, even for these types of problems, I think the default approach most EAs have is usually that problems are best solved by doing research and analytical thought instead. I think this probably leads to a bias against trying things. (We think hard when it's appropriate, but also often when it would be better to just try it and see.)
b. I know of at least one example of bias against action in the community - someone applied for funding to distribute EA books at UPenn and said it was initially declined because CEA was concerned about possible downside risks. Both they and I thought this seemed crazy.
c. Technical skills are required for many of the types of startups that VCs fund, but there are few technically-skilled entrepreneurs in the EA community. This suggests EA is pushing away even entrepreneurs who could contribute to technically demanding projects.
3. Earning to give can still be high impact, especially if you're right and a traditional entrepreneur wouldn't be well-placed to start an EA (mega)project. If it works, Snowball Fund could either beat market returns (increasing the money in EA) or pay a small price (difference between market returns and its returns) per entrepreneur. If this means a higher number of EA-aligned entrepreneurs earning-to-give, or more giving from each, it may still be very cost-effective.
Hmmm, reading your question, I think our disagreement might not actually be about entrepreneurs' fit in EA, but rather about how much labor your plan requires. I think (correct me if I'm wrong) you and I both believe something like:
Let me briefly explain why I think the antecedent holds:
To be clear: it's good to be better at more things, and of course I would be more impactful if I was better at start up stuff. But it just doesn't seem like the biggest thing holding me, or the people I know, back.
Daniel's comments:
On point #1, there's a critical distinction in the type of "follow-on" strategy we're employing, which is a standard template $100K into a large number of companies (generally known as the "spray and pray" model). This is characterized by low diligence per deal as opposed to most VC's who do still put in a decent amount of effort. Spray and pray of course has it's drawbacks in terms of validating the quality of the deal flow, but that's the crux we're exploring here regarding EA-aligned founders being potentially higher quality than average.
On point #2 not to be autobiographical but I personally am an example of an individual who would not have started an EA project without the onramp from Founders Pledge to actually introduce me to the community networks to get Snowball Fund up and running.
Thanks for the data point about you coming from FP! That's helpful.