DENVER – “I think Uniswap Labs is doing a great job,” Joseph Delong, founder of a forthcoming DeFi project called Astaria said during a talk at ETHDenver 2022 on Friday.
That the team who built Uniswap v3 has done strong work wouldn’t really be controversial across much of the crypto space, but Delong is also the recently departed chief technical officer of SushiSwap, the decentralized exchange (DEX) that was founded in 2020 as a direct response to Uniswap, a DEX for the people as opposed to one then heavily critiqued for being too entwined with venture capital.
Delong made the comment at the end of his talk on why DAOs need hierarchy. Coming out of his time running one of DeFi’s most important projects, it had the feel of a sort of “lessons learned” from his time running one of the most grassroots organizations in DeFi: SushiSwap.
Structural Divisions
Right now, SushiSwap is struggling to find its footing again after Delong and its other chief leader, 0xMaki left. For his part, Delong has come to the conclusion that there’s no real getting around hierarchy and clear structural divisions once organizations start to grow.
“Structurelessness isn’t really a viable methodology,” he said. “I am keenly aware that this is deeply unpopular.”
Delong came to talk about why he thinks this ideal of running completely horizontal organizations without leaders isn’t viable and doesn’t actually work. He made it clear that his critique was directed at organizations working to build highly performant products. DAOs that come together to team up on NFT buys or casually invest weren’t his target. It’s groups that are aiming to work together to build products.
Answer to SushiSwap’s Troubles May Be Transforming into On-Chain DAO
He contended that the reason Fortune 500 software is bad so often is because it does its product design without structure. Too many people get involved, pushing for their particular priorities. It makes the software everything and nothing.
In Agile software development, he said, there’s a product lead who decides what goes in and what doesn’t and what the primary point of the software is. That’s structured leadership, he said, and it’s essential for building a good product, in his view.
“I think what you’ll find in the history of human organization is that there’s has always been hierarchy,” Delong said. “The question is how transparent the hierarchy is.”
Too Many People
Delong spent some time on this facet of his argument, namechecking Feminist Jo Freeman and her essay “The Tyranny of Structurelessness” (one that inevtably shows up on any list of must reads about governance). In particular, he emphasized that in structureless organizations, there ends up be an opaque internal structure.
Which is better, he asked, an ineligible hierarchy or one that everyone can at least see and understand.
Many of the reasons he spelled out for why hierarchy and traditional organizations come down to information. Sometimes, among too many people, it becomes too hard to share information efficiently.
In other cases, like in terms of trade secrets and personnel, secrets need to be kept by a responsible party.
He proposed a new way of thinking about structure and DAOs. “Treat DAOs as an organizational structure,” he said. “People participate in DAOs. People aren’t DAOs.”
Echoing the work of OpenLaw, he said, “My thought is we use legal entities as a wrapper,” he said. “I know that token voting is not a great design but it’s all we have so far.”
A common objection to the legal identity approach, he admitted, is that so many people want to participate in these organizations anonymously or pseudonymously. “I actually think anonymity is a knock on effect of not having a legal identity,” he said. It’s about liability, basically, in his view, but it’s also basically impossible to hide your identity forever. “This fails in so many ways,” he said.
Delong said he used to be very sold on doing everything in public, out in the open and without hierarchy. “I’ve done a complete 180 on this,” he said. “Working in the open is really interesting and fun, but fails to scale.”
