Whither the Foundation?
Three Things #181: August 3, 2025
I’ve been thinking a lot lately about the role of foundations: in NEAR, in Ethereum, and in other ecosystems. I recently reread the excellent, classic essay on the trouble with software foundations called “Governance without foundations” by Nadia Asparouhova. I wrote last week about trouble with the Urbit Foundation, and touched upon similar issues I witnessed firsthand at the Ethereum Foundation.
Today, I work for the NEAR Foundation, and in the past I’ve worked or consulted for a handful of other software foundations. I also helped set up a couple. Today, I’m in the process of deconstructing a foundation, in the sense that I’m setting up a decentralized governance mechanism to take some of the responsibility and workload of governance off the foundation.
All of this reading, writing, and working has caused me to reflect deeply on the good and the bad of foundations: both specific foundations, as well as of the foundation model itself. It’s also got me thinking about the question posed by the above article: what path might we choose if we did away with foundations entirely? I think these questions are more relevant now than ever before, and are also directly relevant to my work on House of Stake.
Thing #1: The Good 🥰
Let’s start with the good. I want to initially make the case that software foundations are a good thing. Then we can question some of these ideas, break them apart a bit, poke some holes, and propose alternatives.
The foundation wasn’t the obvious model for crypto. In the early days of the Ethereum project, the project’s cofounders faced a decision point: should they establish a nonprofit foundation to receive the investment coming from the presale, or should they instead establish a for profit company? The opinions of the cofounders were divided, so Vitalik cast the deciding vote. He chose to set up a foundation, and his two primary cofounders, Joe Lubin and Gavin Wood, each then chose to set up for profit companies instead (Consensys and Parity).
In retrospect, this was probably the right decision for Ethereum at the time for a few reasons. First and foremost, it’s arguably good that the cofounders split up, went their separate ways, and set up three very different organizations in different geographies (EF in Switzerland, the Netherlands, and Germany, Consensys in New York, and Parity in the UK and Germany). This is one of the key points that the author makes in the above linked article: we’re building decentralized software, so it’s essential that that software be governed in a decentralized fashion, something that’s quite difficult when there’s a single, central, all-powerful coordinating entity. At least in the case of Ethereum there have always been a multitude of entities participating in R&D and governance. Having a foundation might be fine as long as there are other poles of governance, influence, and funding.
Secondly, in the beginning, any project and community are nascent and fragile. At this point in time it makes sense for coordination and governance to be centralized. There should be a centralized actor who makes decisions, distributes resources, etc.—another point the author makes. In other words, there needs to be a way to coordinate and fund public goods, which aren’t tied to the bottom line or any particular profit motive, and foundations are a reasonable vehicle for doing this, especially early on.
To continue the Ethereum example, the Ethereum Foundation played a pivotal role in getting the protocol off the ground. Most of the core developers worked directly for EF, and the ones that didn’t (i.e., the ones at places like Consensys and Parity) were partially funded by EF. EF conducted the crowd sale. EF staff led research and coordinated all of the core R&D work among all of the core teams at all organizations. EF sponsored meetups around the world, and the annual DevCon event. EF played an important role in PR and media relations, etc. Someone needed to do this work, and another way to look at the role of the foundation in an ecosystem is that it does the things no one else wants to or can do.
The NEAR Foundation plays a similar, pivotal role in the NEAR ecosystem today: it sponsors core R&D, it hosts events, it manages relationships with investors and exchanges, it runs hackathons and works with talented founders to help them build in the ecosystem, it funds educational materials and other public goods projects, etc. These services are all critical for the health of the ecosystem, and in most of these cases, if the foundation weren’t providing them directly or indirectly, it’s unclear who would.
All of this is true when times are good. But the foundation also has a role to play when things aren’t good. In particular, it can act as a credibly neutral arbiter in times of conflict. I saw Vitalik and EF play this role multiple times in Ethereum. Some decisions, whether technical or social, are inevitably contentious and aren’t always easily resolved through ordinary, decentralized governance mechanisms. In these extraordinary circumstances, it can be helpful to have a benevolent dictator make a decision so that the ecosystem doesn’t get stuck, or have too many contentious forks, social or technical. In other words: there are certain coordination problems that only a government-level actor can solve.
Thing #2: The Bad 😩
The main argument made in “Governance without foundations” is that foundations are bad for decentralization. If the purpose of open source software and open source communities is to have an active ecosystem of multiple contributors, and multiple poles of governance, without a single point of failure, then foundations are a problem. The main reason is that, while there may be many companies contributing to a project, there’s almost always only one foundation. In this respect, in a given ecosystem a foundation plays a role akin to a government.
The worst thing about the government, however, is that it’s a singleton. It’s “endowed with the highest form of authority” and has no competition, by definition. As a result, governments and foundations both tend towards bloat, towards laziness, towards what the article calls “a winner-take-all-mentality” that can cause them to “veer off the path into their own game, fundraising for the sake of fundraising, or prioritizing work that doesn’t benefit the project.” I’ve personally seen this happen multiple times, in multiple ecosystems, even to actors who are extremely values aligned and have the best intentions. Show me the incentives and I’ll show you the outcomes; the incentive structure around a foundation, lacking competition and oversight as it does, simply doesn’t incentivize good behavior over the long run. It favors entrenched wealth, power, and influence, which is antithetical to everything that the open source and decentralization movements stand for.
One common response I often hear is that someone who disagrees should go start their own, competing foundation. On the face of it, this isn’t a bad idea: we do, after all, endorse forking as a form of peaceful resistance. Unlike the government, there’s no legal obstacle to launching a second foundation, but the practical obstacles are rife.
There’s a question of branding and legitimacy: there can only ever be one “Linux Foundation,” one “Ethereum Foundation.” Any competing foundation is playing catch up, and doesn’t have the Lindy effect working in its favor. Then there’s the issue of trademarks and other IP ownership: the Ethereum Foundation, for instance, owns the “Ethereum” trademark, and controls important brand surfaces such as ethereum.org and the @ethereum X handle. A competing foundation would have none of these assets. Finally, there’s the elephant in the room: funding. Foundations typically have massive endowments, often in the hundreds of millions to billions of dollars. This by virtue of being the first mover in an ecosystem, by receiving a large premined endowment, etc. No later foundation can ever compete financially.
I’ve seen what this leads to. Even in an ostensibly decentralized ecosystem, everyone turns to the foundation for guidance and leadership. The foundation’s word is as law: “the highest form of authority,” as Asparouhova put it. Projects and people blessed by the foundation get the lion’s share of attention, funding, etc., and everyone else fights over the scraps. I’ve been on both sides of this phenomenon and this behavior, and I can attest to how bad this is for an ecosystem, especially one that purports to be decentralized. The worst case scenario is that the foundation becomes a wolf in sheep’s clothing, advocating for decentralization on the surface while hypocritically commanding the ecosystem from a position of central authority.
Asparouhova has good advice, especially for younger projects: avoid the foundation entirely and instead lean into companies as fit-for-purpose service providers. There can be a multitude of such companies, and it should be the purpose of the first company, the founding company, to decentralize authority and funding as widely as possible. This is good advice, and it’s all well and good for a new project founded by a decentralization maximalist.
Even in the blockchain space, however, most projects aren’t so fortunate, and most people who preach decentralization don’t practice it very well. It’s also especially hard to do this early in the life of a project, when centralization is the key to moving quickly, and when there are few other trusted actors you can rely on. The result in most ecosystems is a massive, wealthy, influential, centralized foundation that makes most of the decisions, employs most of the key leaders, and provides most of the funding.
There’s no easy fix. Getting out of this situation requires doing the things you should’ve done from the beginning: the most powerful actors in an ecosystem need to constantly look for ways to decentralize that power as widely as possible. It takes a certain kind of leader to actually carry this out, because the default path is always towards more centralization, which is easy, efficient, and familiar. But it’s uninteresting, it isn’t novel, and it’s antithetical to everything that we stand for.
If you’re at the center of a successful, growing project or ecosystem, power will tend to accumulate in your hands. It’s just the way of the world, and in spite of your their intentions, even good people let this happen. You have to actively work against it to prevent it.
We’re working on fixing this for NEAR at House of Stake. We’re only getting started and it’s already bumpy and difficult, but we will persevere. The NEAR ecosystem deserves it. And, eventually, the result should be a much more robust ecosystem topology than the one we have today, dominated as it is by one foundation. Here’s some things we’re trying, and some hint of where we’re going.
Thing #3: The Alternative ✨
There’s lots to criticize about foundations. It’s much harder to propose a better viable model. This is where most conversations critical of foundations end: ask the complainer to propose a better model and they usually shut up. I, too, have been guilty of criticizing without offering a better alternative.
As mentioned, the alternative path proposed by the author of the above article is for the initial coordinating entity, i.e., the founding company, to “find major contributors who aren’t funded or supported by them” and effectively decentralize control and funding away from themselves into the hands of other contributors. Their motivation for doing this is the overall health and success of the project: in the long term creating something that “looks more like a standard than an organization.”
That’s one possible path forward, but in my experience it’s easier said than done. The trouble lies in finding other major contributors, especially those that are competent, values aligned, and can be trusted. The project that’s done arguably the best job of this, Ethereum, has indeed managed to foster an ecosystem of over ten independent organizations that regularly contribute to the core protocol, resulting in a number of independent client implementations. Ethereum deeply values client diversity, so this can be seen as a win.
But if we look more closely, it becomes obvious that these supposedly independent teams aren’t actually independent in all the ways that matter. In particular, none of them is profitable or sustainable on its own without the largesse of the Ethereum Foundation or another billionaire. These companies can also seek acquisition, as in the case of Prysmatic Labs, which was acquired in 2022 by Offchain Labs, but this leads to other challenges.
The reason these initiatives aren’t sustainable is simple: core development work, clearly a public good, simply isn’t profitable, and this isn’t likely to change anytime soon. In this respect, it’s alike all of the other backbreaking, thankless, unprofitable work that needs to be done to build an ecosystem from the ground up: it’s ultimately valuable for the ecosystem over the long term, but it requires a huge upfront investment of time and resources. We need better, more decentralized ways to coordinate this work. It’s hard enough to do this at the launch of a project; the problem is compounded in more mature ecosystems, including Ethereum and NEAR, where there are already foundations that are highly influential and control large treasuries.
There’s no panacea and, I think, no single, straightforward playbook for solving this problem. I haven’t seen this done well in any crypto ecosystem to date. But we shouldn’t despair, for two reasons. First of all, crypto is simply immature and ecosystems haven’t had enough time to run a sufficient number of experiments. Second, there are models other than the one proposed in “Governance without foundations.” For this, we can turn to much older models from the wider world: pre-crypto, even pre-software.
Some examples of successful, sustainable, decentralized real world governance structures include retroactive public goods, the multi-host fiscal network, perpetual endowments, multi-stakeholder oversight bodies, and member-owned service cooperatives. Let’s briefly go through these one by one.
Retroactive public goods funding means offering prizes or other funding opportunities for projects that have already been delivered. This can be done in a decentralized, committee-driven fashion out of a joint treasury. NASA COTS and X Prize are two well known, successful examples that are much older than crypto. Next, Open Collective and the Software Freedom Conservancy are two examples of how legal and fiscal oversight can be decoupled from something like a foundation and competitively supplied to an ecosystem (“multi-host fiscal network”). Third, university endowments are able to provide sustainable ecosystem funding indefinitely, and can be governed more broadly than the university itself. The IETF endowment proves that this model works for software governance as well. Fourth, ICANN and the Global Fund Country Coordinating Mechanisms are both successful examples of multi-stakeholder oversight bodies, from inside and outside the software industry. Finally, there are tons of examples of member-owned service cooperatives, including credit unions and rural electric cooperatives; the most famous historical examples were Visa and Mastercard until the mid-2000s.
These are all examples of successful, sustainable governance models that do not feature or require a centralized, singleton foundation or other privileged actor, and I believe many or most of these would work in the crypto ecosystem as well. We just need to take the time to better study the alternatives and run more experiments.
This is precisely our goal at House of Stake. I’ve spent time recently researching and seeking to understand these models, and this research informs the design, and my stewardship, of the House of Stake. HoS is not quite any one of these models in isolation, but it exhibits characteristics of several of them, and of course it’s designed for the specific strengths and constraints of the NEAR ecosystem.
While HoS is relying on the NEAR Foundation for both legal and fiscal support in the beginning, it will decouple from NF in both of these areas before long. HoS has its own endowment, and we’re already discussing staking that endowment and using the yield for sustainable ecosystem funding. HoS is inherently multi-stakeholder, composed as it is of a multitude of stakeholders from across the NEAR ecosystem with competing interests and diverse opinions. HoS will likely experiment with retroactive public goods funding. And you could understand HoS as a member-owned service cooperative, given the requirement to own NEAR to participate in delegation and voting.
The bottom line is that we have our work cut out for us, both at NEAR and more generally, to do the important work of removing the foundation as a single point of failure and as a bottleneck on the creativity, stewardship, and growth of the ecosystem. But we also have more powerful tools at our disposal than ever before, which is why I’m optimistic that we’ll get the job done. AI also has a big role to play here, and I feel strongly that coupling modern AI tools with the sort of tried-and-true economic and governance models outlined above will give us the ability to meaningfully decentralize governance while simultaneously accelerating ecosystem growth and product development, things that have historically been at odds. Watch this space for more on how we intend to do that—and, as always, I’m here for your ideas and suggestions too!