A response to "Empowering the ENS Foundation" — proposing amendments, not a rejection. Where I'm coming from I'm a developer, not a governance person by trade. ENS DAO's ecosystem is how I got into this space — I build identity infrastructure on ENS and maintain open-source Ethereum tooling. I'd rather ship than argue about org charts. But this decides what building on ENS looks like for years, so it's worth getting right — quickly, and without a fight either side "wins." What everyone actually agrees on The diagnosis in the temp check is correct, and it's not just an ENS problem. Vitalik put it bluntly in January: most DAOs today are "a treasury controlled by token holder voting" — a model he called inefficient, vulnerable to capture, and a failure at mitigating the weaknesses of human politics. Token voting is slow, binary, and low-context; participation in major DAOs frequently sits below 10%; and decision fatigue is real — early excitement decays into people no longer willing to participate or even stay informed. ENS makes too many small operational decisions and too few big strategic ones. All true. But "empower the Foundation" and "keep the sprawling DAO" are both the wrong response I don't want to defend the DAO as it exists — hundreds of delegates, endless proposals, reputation-driven votes. That machine is the problem, not the thing to protect. But the temp check's fix goes too far the other way. Handing the treasury to a Foundation board doesn't fix bad voting so much as move the whole problem out of the community's reach — and it's the kind of change that's easy to make now and very hard to walk back if it doesn't work out. The right answer is a leaner, purpose-built DAO: fewer votes, but each one consequential; operational work delegated to accountable committees; and the treasury staying under community control. Vitalik's framing again — DAOs should be built to solve specific problems like long-term stewardship and dispute resolution, treating governance design as 50% of the job, not 10% — not run as general-purpose voting treasuries. The model: split by strength, thin the DAO, keep the purse 1. The DAO stops voting on everything and oversee the structure instead. Its job shrinks to a few high-leverage decisions: approving budget, appointing and rotating committees, and reviewing their performance — with power to replace members or dissolve a committee outright. This directly attacks decision fatigue: fewer votes, each one that matters. Protocol-level control (contract upgrades, fee structures, registry, root key, constitutional changes) stays fully on-chain with tokenholders, exactly as the temp check proposes. 2. Operations → the Foundation / ENS Labs, funded by an envelope, not a vault. Give the Foundation its operational mandate, but fund it the way ENS already funds working groups: from a DAO-approved budget drawn into a multisig, with unspent funds returning to the treasury. The community keeps the structural lever — expand, shrink, or revoke the envelope at the next funding window — instead of a reactive removal petition. If phased custody transfer is still wanted, do it in tranches gated on real checkpoints, never one motion. 3. A leaner DAO can borrow from Optimism's bicameral split. Optimism deliberately separated decisions by constituency: a Token House for economic/protocol matters and a reputation-based Citizens' House for public-goods funding — precisely because tokenholders are one constituency among many and aren't always the right owners for decisions that aren't about system economics, and because splitting the branches helps avoid concentration of power through checks and balances. ENS doesn't need to copy it wholesale, but the principle — match the decision to the body best suited to it, rather than routing everything through one token vote — is the leaner-DAO blueprint. 4. The bottleneck for builders is technical review — fix it with an SLA'd committee. What actually discourages outside contribution isn't custody; it's that protocol-level review is slow and opaque unless you're inside Labs. Stand up a small, DAO-elected, rotating Technical Review Committee, open to non-Labs contributors, that gives ENSIP-level proposals an initial technical read within a fixed window (say 2 weeks), after which cleared proposals get a shortened ratification path. 5. This isn't hypothetical — SPP3 already proves the committee model works in ENS. The latest Service Provider Program cycle is a live, in-production version of exactly this: a small ratified committee, a published weighted rubric, a binding dated timeline (14-day submission, 14-day review, fixed vote), a hard eligibility screen, structured interviews with a no-backchanneling rule, and unspent funds returning to the treasury. It processed 26 applications on schedule. The temp check wants to absorb SPP into the Foundation; I'd replicate it instead — it already cuts through reputation and marketing to judge substance, with public rationale the community can audit. 6. Bake in the accountability SPP contributors already asked for. The SPP3 thread surfaced the exact gap to close DAO-wide: conflict-of-interest rules covering not just direct compensation but material relationships — former colleagues, business partners, close friendships — with proactive written disclosure and recusal. Any empowered committee or Foundation board under this model adopts that from day one, published and queryable, live before authority is granted — not drafted later by the body being overseen. 7. Protect the on-ramp. Ring-fence a minimum share of DAO-directed funding for developer grants, tooling, and docs — the functions that get new builders in — published and trackable so they aren't quietly deprioritized in a reorg. Pros - Actually thins the DAO — attacks decision fatigue and low participation head-on, instead of preserving the delegate sprawl or abolishing community control. - Gives the Foundation real operational speed and hiring authority without surrendering the DAO's one core job: preventing concentration of control over the treasury. - Every commitment is checkable — envelope, rubric, timeline, disclosures — so a new contributor can verify it with no trust and no governance background. - Evolutionary, not a rebuild: reuses working-group allowance, funding windows, and the live SPP committee, plus a proven external model (Optimism) — low implementation risk. - Lower stakes per step means less all-or-nothing tribal fighting; nobody has to win the whole argument today. Cons - Committee selection is the unsolved hard problem. If seats are decided by reputation-and-momentum token votes, the popularity contest just moves up a level. My provisional answer mirrors SPP: initial slates nominated by Labs and existing stewards, ratified (not micromanaged) by the DAO, with strict term limits, mandatory rotation, and published reasoning as the check. This needs its own scrutiny — I won't hand-wave it. - A leaner DAO concentrates more power in fewer committee seats between elections; the rotation, term limits, and transparency have to be real, not formal. - An envelope model keeps the Foundation on a shorter leash than full custody — some will argue that undercuts the independence the temp check wants. That tradeoff is real and worth debating openly. - Fixed review windows can be gamed or ignored under pressure without a clear fallback when a deadline slips. - A bicameral or committee-heavy structure adds coordination overhead and can itself stall if badly composed — the thing it's meant to cure. Concrete proposal — how to move forward 1. Redefine the DAO's scope to governance-of-governance: approve budgets, appoint/rotate committees, review performance. Protocol-layer control stays on-chain, unchanged. 2. Fund the Foundation via a DAO-approved budget envelope (working-group model), not a custody transfer; DAO retains treasury ownership and adjusts the envelope each funding window. Any custody transfer happens only in checkpoint-gated tranches. 3. Stand up a DAO-elected Technical Review Committee, open to non-Labs contributors, modeled on SPP3: published rubric, fixed SLAs, rotation, public rationale. 4. Explore a lightweight bicameral split (Optimism-style) so public-goods and ecosystem decisions aren't gated on the same token vote as protocol economics. 5. Adopt a DAO-wide conflict-of-interest and disclosure standard covering material relationships, with proactive disclosure and recusal, live before authority is granted. 6. Ring-fence a minimum share of DAO-directed funding for developer/ecosystem grants, tooling, and docs — published and trackable. 7. Route proposals that touch both protocol and treasury through both tracks rather than skipping either. Further horizons (worth flagging, not proposing yet). Vitalik's January note also gestures at more experimental machinery: personal AI voting agents trained on a user's values to cut decision fatigue, zero-knowledge private ballots to stop governance from becoming a social game vulnerable to bribery and whale-watching, and prediction markets that let agents bet on whether proposals will pass, filtering spam by rewarding good signal. I'm not proposing any of these for ENS today — they're early and unproven at this scale — but a leaner DAO built on clean committee and envelope structures is exactly the substrate you'd want in place before experimenting with them later. Worth keeping on the radar, not the ballot. This isn't Labs vs. DAO, and it isn't "keep the old DAO." It's a smaller, sharper DAO that ships — each side doing what it's actually good at, using machinery ENS and Ethereum have already shown can work. ----------------------------------------------------------------------------------------------------------------- Edit 2 — responding to the council context and thread feedback Four changes, one correction, and one thing I got wrong. Correction first: not a single multisig. Point 2 above says Foundation funding is drawn "into a multisig." That was careless — avsa's middle-path proposal specifically warns against a single multisig, and it's the same structure James objects to. What I actually mean: a DAO-approved operating budget held under a structure with distributed signing authority and published thresholds, not one wallet with one signer set. The point was always the renewable budget, not the custody mechanism. What I got wrong: committee selection. My Cons section proposed slates nominated by Labs and stewards, ratified by the DAO. That's the same objection Brantly raises about the board itself — selection by an interested party, with ratification as a formality. Fair hit. Replacing it with: Qualify by contribution (merged PRs, ENSIPs, grants delivered), not token weight or name recognition. Reserve seats each cycle for first-timers, filled by lottery from that qualified pool. You can lobby an election. You can't lobby a lottery. Cooling-off, not just term limits — one term, then sit out a cycle. A low-threshold override letting a modest share of delegates pull any committee decision to a full DAO vote. Rare in practice, but it keeps the community live between elections. On reversibility, which I underweighted. Tranches aren't enough on their own. Adding: cap Endowment withdrawals at 5% per year (avsa's number — I'd rather endorse an existing proposal than invent a competing one), a fixed sunset requiring affirmative renewal rather than continuing by default, and a recall path for individual committee seats that doesn't require the full nuclear option. On voting power, structurally. I'm not going to litigate anyone's delegation. But if a change of this magnitude passes, it shouldn't pass on token weight alone — structural changes to custody or scope should require both a supermajority and a floor on distinct participating delegates. That's a rule that binds everyone equally, including anyone who agrees with me. On where I actually disagree. The temp check wants "a DAO that votes rarely and matters every time it does." So do I — that's not my dispute. My dispute is narrower: I don't think transferring custody is required to get there. A renewable budget with real spending autonomy delivers the same operational speed while leaving the community a live lever. If someone can show me why budget authority alone can't work — a legal constraint, a real operational blocker — that would change my mind. I haven't seen that argument made yet.