The DAO model is either in crisis or in chrysalis
Aave, Curve, Uniswap and — kind of — Bitcoin are all reconsidering how to proceed
Almost everyone agrees that decentralized autonomous organizations (DAOs) are a dream.
But does that make them an ideal, a place of freedom, alignment and growth? Or are they more a dream like: “Dream on, loser.”
That’s the question.
On Sunday night I published my second-ever podcast, an hour-long deep dive into crypto’s dominant blockchain money market, Aave. I made it to get back into decentralized finance, because Aave is the Godzilla of DeFi. I also did it because Aave was in the news thanks to a controversy.
The point of my podcast: This thing has become big enough that it’s worth fighting over. Between the lines, I did it assuming the controversy would blow over. The fight didn’t seem like that big of a deal.
Incorrect.
Little did I know that, as I was finishing up my edits, Avara was getting set to put in a proposal to the Aave DAO that would kick off a firestorm, a full rift over the future of the project. Avara is the company that set up Aave, did the key early development work, ran its initial coin offering and controls its best brand assets, IP and its website.
Avara is run by Stani Kulechov, the founder of Aave, the one who tends to do most of the speaking on Aave’s behalf in the Ethereum community and in the crypto press.
Avara also pivoted to reinventing social media a few years ago. It did not.
But I’m not writing this post to rectify the fact that I under-indexed the fight inside the Aave family (though I did). I’m writing it because it’s a come-to-Jesus moment in blockchain governance.
Divers alarums
It’s not just Aave.
A sea change is afoot in on-chain governance and the model of collective decision making that ascended amidst the DeFi boom that kicked off in 2020.
The headline grabber has been Aave. The latest wrinkle in that debacle stems from the fact that Avara started keeping fees it earned from giving visitors to aave.com an easy way to interact with the Aave protocol (a.k.a. a frontend), money that previously went to the DAO treasury.
This catalyzed what looks like long-simmering frustrations into a conflagration between the Aave community and Avara, its creator.
This weekend, a proposal went up to move all rights to the Aave name and control of the website to the DAO.
The vote is set to run a few days, ending tomorrow, on Christmas.
The point of this seems to have been to show that the Avara side has enough AAVE, the governance token, at its disposal to vote it down.
To that end, advocates for the DAO side are pushing other holders to vote “abstain,” as a signal of the illegitimacy of the proposal.
As I write this, NAYs are at 58%, YEAs at 4% and ABSTAINs at 38%.1
There’s a lot to unpack in this story, but I’ll return to Aave in detail another day.
The point today is that Q4 has seen some big moves in governance:
Uniswap
Uniswap has completed its “unification” vote, where most of what was once done by the Uniswap Foundation has been moved back into the original company, Uniswap Labs.
Uniswap is the most influential decentralized exchange in crypto, though it isn’t always the largest (but it is always very large).
There’s been this structure where companies would set up a non-profit to promote a new protocol while a company would remain to monetize. This always looked like smoke and mirrors to me and with “Unification’ finally everyone has, by my read, implicitly admitted that.
This doesn’t win me back life I lost arguing with flaks over the obvious fiction of “totally separate organizations.”
But this vote went through easily, passing with 98% support among voting token holders (turnout was low: ~12% — but that’s par for the course – no one got into crypto to vote).
Curve
Curve is another important decentralized exchange, one that specializes in assets that should have nearly the same price (such as stables).
This month, its DAO narrowly voted down a funding proposal from the company that launched it. Which is both like “damnnnn” but also like “okayyyyyy…”
Yearn Finance
The DeFi project that set the standard for set-it-and-forget-it yield back in 2020 has shrunk quite a bit, but it still has a team hard at work pushing returns for the users that remain.
In September, it kicked off a reorg. That recently executed an overhaul to drive more revenue back to token holders and enhance the visibility and transparency of the work of the teams that drive those earnings.
Ethereum Name Service (ENS)
ENS floated a rethink in how discussion flows around the DAO in November. Without any review on this other than reading some of the forum posts (and no significant prior reporting on ENS), it sounds like lots of committees and some forces within the organization had come to believe this was slowing it down.
It’s an example of a very important organization that’s in conversation about its setup here at the end of 2025.
A proposal went to a vote and the vote failed basically 2:1. (roughly 2% of tokens voting)
Bitcoin
The present moment of soul searching over the threat of quantum computing to Bitcoin’s core cryptography is, ultimately, a governance question.
Quantum computing entangles the innovator’s dilemma with Bitcoin
The innovator’s dilemma has come for Bitcoin and you shall know it by the name of Q-day.
There are other examples but that’s enough to make the point.
You are here: Blockchain organizations went from dreamy ideas to actual operations from 2019 to 2025. Endless debates about valuations aside, there’s unquestionably billions of dollars of actual transactions running through these things. They are real.
Now that they are real, they are getting real.
Which way, robots on the internet? Are we living the dream or are we waking up from it?
If you quit reading here, I won’t blame you. I’m just going to sketch out some broad ideas before I go. You’re caught up.
I want to believe
Decentralization is one of those things that takes a long time to understand. As I began work as a crypto reporter, it was one of those concepts that delighted me in the ways it forced my dumb little human brain to enjoy new perspectives on the world.
The crux of it is this: Can you put software out on the internet that runs itself thanks to the substrate of a shared global computer (like Ethereum and Solana) and give humans a little control over it so that it can evolve as needed?
Here’s how I have always understood the basic idea:2
Step one: A company comes along that wants to start some kind of decentralized protocol, to build a robot. They seek funding from investors to pay to develop it and to do the legal and promotional work around it. They sell those investors equity in the company.
Step two: They build the robot. They release the robot. They issue a token that is important to the robot and reflects its value. Not necessarily in that order.
The company gets a lot of that token and so do its investors. Now the team and the investors have both equity and tokens.
But the token is about just the thing. It’s not about the company.
For the company to grow in value from there it needs to either drive value to the thing in order to grow the value of their token stash or create products built on the thing that people will pay to use.
Step three: Holders of the token create a DAO to steward revenue earned by the robot. That revenue is often spent on stewarding or growing the robot, promoting the robot or insuring the robot.
Other companies come along that provide services to the robot, which is all part of decentralization.
But the DAO is not a company, which sits poorly with legal structures all around the world. They are a new secret third thing.
Some people think this uneasy arrangement is a problem and others view it is as a modest rebellion/reformation. Take your pick!
Step four: Theoretically the original company either winds down or becomes just another contributor to robot’s ecosystem, one privileged only by the quality of its ideas.
Or it sells, such as in the case of Interop Labs. Which inevitably leaves the token holders asking why they didn’t get a cut of the sale, rightly or wrongly.3
And I confess, having covered startup founders on the normal web for several years before switching to crypto, it was hard to believe that this, shall we say, highly agentic sub-species of human would willingly redistribute control over their creations.
One problem with step four is all the control the company has in step one. Usually the company and its investors have the lion’s share of the tokens that govern the robot. It gets murky and even if there’s nothing inherently wrong with that, it’s going to inevitably cause tension with the people who join the DAO directly.
Because websites need to be owned by normal companies, the original companies tend to own the website that always ends up being the primary way of using the robot.
Sometimes those companies monetize that access. This has always seemed theoretically fine to me. A gentle reward for kicking the whole thing off, nothing to get upset about.
After all, with robots on the internet, anyone can build a frontend. That’s the beauty of the things. The DAO can make their own! Or other companies can include access in larger apps that reach many protocols!
However, the Aave dispute has illuminated the fact that controlling the canonical “whatever.com” address is a vulnerability for DAOs.
What does it do if animosity forms between the DAO and the original company? What if the state wants to pressure the company to get to the DAO?
Much to consider there.
The main takeaway from the Aave story is that, if DAOs persist, they need to have a serious conversation that pretty much comes down to: Okay but what if this robot actually works? What, at a minimum, does the DAO need to own if the worst case scenario happens and the founding company turns against the DAO?
That wasn’t a scenario that had to be taken too seriously in 2019, but times change. These things worked.
Real talk: A recurring critique of DAOs has always been that they were never real, it was always a fiction. It was a hustle the crypto world came up with to route around hostile regulators in the U.S. But, then again, some DAOs really persist in a DAOish way and, despite headaches, seem to be functioning as intended.
Look, don’t listen to me. Listen to a guy who DAO’ed and lived to tell about it
Joseph DeLong was in the heart of one of the most hardcore DAOs, SushiSwap. He’s a major critic of the model now. I get it. The thread linked to the embedded tweet is great.
You should stop listening to me, go click on it, then come back here.
Believe me, I’m going to sound like an idiot to a lot of people! The folks who have actually built things in the world are very bearish on DAOs.
The further irony of the fact that I am clinging to the DAO model is that I used to work for grassroots non-profits and I know what a nightmare grassroots governance can be.
And yet, yet.
Final point.
The challenge to me is that I don’t see how the blockchain world moves forward productively without DAOs. More specifically, I don’t see how crucial protocols can persist and evolve in a decentralized way, a censorship-resistant way, without organizations that function outside state sanction.
That’s not to say there isn’t a way! I just don’t know what it is, right now, today, as I write this. I’ll talk to some folks about this and report back. Maybe it’s obvious?
End note: Eugene Leventhal has launched a project to discuss evolving governance further. This could be a fruitful time for brainstorming along those lines.
Much has been made of the price action on the AAVE token around this dispute. Yes, it’s down 16% on the month, but tokens are volatile. Its price was also that low in April. Whatever.
As an editorial note, I’m not really writing about the law here. I’m writing about the world and working in the world and people working together. The law is a layer of abstraction that we have added to the world in order to deal with conflicts without violence or coercion. Can’t do without it. But the world is not the world’s laws. If you want a legal treatment of these topics, follow some lawyers. It’s just not how I think or what I write about.
To me, the equity is the company and the team, and they are all owned by the company. Just because something associated with a DAO gets a payday, that doesn’t mean the DAO earns a cut. This seems easy to me, but I’m not in the trenches.