DAO do you do?

DAO do you do?

I've spent a fair amount of words being mean about DAOs. The Bored Ape token controlled by a16z and a few jpeg sellers, the golf course DAO, the constitution DAO, the whole Cool DAO bros genre. So this is the constructive companion piece: if you actually want to coordinate a group of people on-chain without it collapsing into either chaos or plutocracy, how do you do it?

What problem are DAOs even solving

Strip away the hype and a DAO is an answer to an old question: how does a group that doesn't fully trust each other make decisions and spend shared money without a boss? Who decides what gets funded, who can change the rules, and how do you keep any one party from capturing the whole thing.

That's a real problem. Open source projects, protocols, and treasuries genuinely need to solve it. The disappointment is that the dominant answer the space landed on, token voting, mostly doesn't.

Why token voting becomes plutocracy

One-token-one-vote sounds democratic and is the opposite. It's one-dollar-one-vote with extra steps.

Whoever holds the most tokens decides, and the people who hold the most tokens are usually the founders and the VCs who got their allocation early and cheap. So "decentralized governance" repeatedly resolves to "the cap table votes." Worse, voter turnout in most DAOs is dismal, which means a small, motivated, well-funded minority can pass essentially whatever it wants. You've rebuilt a corporation with worse accountability and a token ticker stapled on.

This is the part the brochure leaves out: a governance token is a financial instrument first and a voting right a distant second, and financial instruments concentrate.

Top row: 'one token, one vote' leads to 'whoever holds the most decides' leads to 'the cap table (plutocracy).' Bottom row: an open-funding flow, propose then discuss in public then escrow and pay on milestones, labeled Monero CCS and MAGIC Grants, with no token.
Token voting concentrates into the cap table. Open, milestone-based funding (CCS, MAGIC Grants) does the same job without a token.

Things that actually work

The mechanisms that hold up tend to deemphasize "everyone trades a vote token" in favor of legitimacy that money can't trivially buy.

  • Rough consensus over formal voting. Plenty of the most decentralized, longest-lived crypto projects barely "vote" at all. They argue in public, ship code, and let running software and adoption settle disputes. It's slow and messy and it works.
  • Reputation and contribution weighting. Influence that accrues to people who actually do the work, rather than people who bought the most tokens, is much harder to capture and much closer to what you actually want.
  • Funding mechanisms that are transparent and human. Monero's Community Crowdfunding System is a good model: contributors propose work, the community discusses it openly, money is escrowed and released against milestones. There's no governance token at all. It's just legitimacy, transparency, and accountability doing the job a token was supposed to do and couldn't.

The MAGIC Grants model, which I actually work on

I'm on the board at MAGIC Grants, so this is the part where I'm talking from experience rather than from a chart.

MAGIC is a 501(c)(3) nonprofit that funds privacy and cryptocurrency work. It is, in a sense, the boring answer to "how do you do a DAO well": a real legal entity, with a real board, real fiscal responsibility, and grant committees that evaluate proposals on their merits. There's no token to pump, no plutocrat outvoting the community by accident, and no smart contract that drains the treasury because someone fat-fingered an upgrade. The coordination problem gets solved by structure and stewardship, not by a voting token nobody actually votes with.

It is much less exciting than a DAO with a Discord and a moon roadmap. It also still exists and still funds useful work, which is more than most 2021-era DAOs can say.

The legal wrapper question

If you do want on-chain governance with legal standing, the wrapper matters. Wyoming and Colorado both have DAO LLC statutes now, which let a DAO hold the legal protections of an LLC: limited liability for members, the ability to sign contracts, hold a bank account, and not have every member personally on the hook when something goes wrong. If your DAO touches the real world at all, an unwrapped DAO can get treated as a general partnership, which is the worst of all worlds. Pick a wrapper on purpose.

Pragmatic advice

So, DAO do you do?

  1. Be honest about whether you need a DAO or just need a transparent funding process and a legal entity. Usually it's the latter.
  2. If you issue a governance token, assume it will concentrate, and design assuming the cap table will win every vote, because it will.
  3. Prefer rough consensus, public discussion, and contribution-weighted legitimacy over one-token-one-vote.
  4. Use a real legal wrapper if you touch money or contracts.
  5. Look at what's worked: the Monero CCS, MAGIC Grants. They're unglamorous and they last.

The good DAOs mostly look like well-run cooperatives with good plumbing. The bad ones look like a16z's portfolio with a vote button. Build the first kind.

If you want my crankier take on the whole genre, it's over in Monero is a DAO. And if you're curious what I build now, that's Such Software.