Part 7 of our 7-part series with the team at BlockScience features a conversation with Jessica Zartler, Communications Advisor and Researcher at BlockScience. Jessica delves into various areas, including:
- What is “decentralization”?
- What is “progressive decentralization”?
- What is the relationship between decentralization and centralization?
- How do decentralization, ethics, and regulation overlap?
Jessica Zartler (JZ) is a Communications Advisor and Researcher at BlockScience. She is also a Founding Steward of the Token Engineering Commons, an Advisor to the Token Engineering Academy, and soon-to-be course creator and former Ecosystem Development Director at CommonStack. Her background is in investigative journalism and political science.
The interview was conducted by Eugene Leventhal, Executive Director of SCRF.
Audio: Spotify, Apple, Spreaker
Video: YouTube
Key takeaways from the conversation:
What does the term “decentralization” mean?
Today many people orient themselves around the term “decentralization,” but in itself, that term isn’t particularly helpful. “Decentralization” is a spectrum. You can decentralize policy-making, community, programming skills, ownership, infrastructure, jurisdiction, consensus mechanisms, and much more. In that light, what does “decentralization” really mean?
BlockScience’s understanding of it is inspired by the field of cybernetics, which applies fruitfully to the world of complex socio-technical systems like those in web3. In particular, the work of Stafford Beer is particularly noteworthy. Beer was a far-sighted management cybernetics philosopher who said as far back as the 1960s that it was a naive dichotomy to call any complex organization of people and technology “centralized” or “decentralized.” Rather, it’s always both things at once, in different dimensions. Today, this realization applies to new structures like Decentralized Autonomous Organizations (DAOs).
“Decentralization” is indeed a frustratingly vague term. And yet it signifies something very real to many people. What are some specific things that have been helpful to you as you’ve tried to explain useful ways to understand it?
Part of it is untangling dimensions or layers, and for this, mapping has proved to be crucial. Researchers are always trying to figure out what systems say they do and then following on with the hard work of ethnography, diagramming computational models, and so on to see those systems in action. What is going on in all of those layers? Is the technical infrastructure all that decentralized? In web3, for example, a great deal of the actual computation is performed via Amazon’s AWS service, which is as centralized as can be.
So we have a social layer, a technological layer, and an economic layer, and what BlockScience does with its new governance research arm is to analyze the projected narrative. What are the crypto news articles saying, what are the press releases saying, and then what is actually happening inside? That kind of work has to be done with a first-principles approach, taking all the different tools, like computation modeling, and trying to get a picture, by looking through a few different lenses, of what’s actually going on.
The more people working on this who understand the complexity and respect the nuance of the various layers of decentralization, the more visible the projected narrative becomes. Stafford Beer had a concept called “POSIWID”—the purpose of a system is what it does. There is no point in claiming that the purpose of a system is to do what it constantly fails to do. So when researchers start looking under the hood of DAOs—going all the way through to the smart contract layer, looking at all the forms, examining all the communication channels—they can see what is actually happening inside versus the projected narrative about the system.
Let’s move on to the idea of “progressive decentralization.” What does this term mean?
One of the first mentions of the term was in 2018 by Dapper Labs in Toronto. They had a big hit with their NFT game Crypto Kiddies, and they wanted outside people to contribute to it. But Dapper didn’t want to own the whole thing; instead, they wanted to progressively give ownership and/or responsibility for different aspects of the project to the community so that the project could “scale horizontally.” Again, they were using the term “decentralization” without specifying whether everyone participating would be running their own node and getting collective ownership or if the founding team would maintain power through concentrated token-holding. In general, the term “progressive decentralization” refers to a gradual process that begins with founders who have complete ownership, and then slowly, over time, the entity is opened up to a token-based model or a DAO.
This image comes from BlockScience’s advisory work with Gitcoin DAO and simply shows the gradual process of progressive decentralization.
Eventually, the meme “Just DAO It” emerged. Some people took it very seriously:
Others who had experience trying to “hand things over to a community” may have read the slogan more ironically:
That broad dichotomy was also expressed in the slogans “DAO first” versus “Exit to community.” The question arose: How much sense does it make to dump all that responsibility on an amorphous community that is just getting started? Would it make more sense to work out the operational and functional components of a project in advance before burdening the community with all that?
There’s a lot of genuine desire to achieve horizontal scaling and co-creation, harnessing and coordinating collective intelligence toward some aim. Very often, in web3 projects, there’s a sincere political idea to expand and have more collective ownership. This also goes back to the open-source ethos, which has given much to the world. And it’s also true that for many purposes, collective intelligence is superior to that of the individual or even the small group. With the challenges that humanity is facing globally right now, there’s a real need to tap into that and let people work bottom-up on some of those problems.
At the same time, some people are implicitly using these ideas to avoid regulation. They may have a multi-sig with millions of dollars; to avoid getting in trouble with the SEC, they quickly do an airdrop to “decentralize” their project. That’s obviously not in the same spirit.
Following on your statement that some people are leaping to decentralization just to avoid regulation, ShapeShift is one interesting example. Erik Voorhees, the founder, has explicitly said that part of their logic was to avoid regulation. Do you have any concrete examples of progressive decentralization that you can share?
It’s interesting to watch how these trends come and go. A lot of it is pure mimetics. People say a certain word, and it’s almost like waving a flag or showing a “summoning” card. In a sense, this is a big part of BlockScience’s work as well, getting past the limitations of language and terminology to reach other legitimate kinds of mental modeling that people have about these issues.
BlockScience didn’t work directly with ShapeShift, but their case is very interesting. They did one of the largest token airdrops in crypto history. The wall they hit was around intellectual property. There wasn’t a way, at least not one they were comfortable with, for handing over the token holdings to the DAO or community at large. On the intellectual property issue, they realized that there has to be a legal entity that “owns” those rights. This is something that’s still being experimented with in web3. How does a DAO have collective ownership of intellectual property?
MakerDAO and Uniswap are both fascinating, but again, BlockScience didn’t work directly with either of them. MakerDAO went through a “decentralizing” process, then centralized into a foundation, then decentralized again. Some people called into question the voting power and engagement there. There’s a ton of research that is worth looking into. Regarding Uniswap, some highly critical articles called into question the concentration of token holdings, saying that the infrastructure and the proposal process weren’t that accessible and that the whole project wasn’t as decentralized as the narrative made it out to be.
Bear in mind there’s no absolute right or wrong here. Going back to the POSIWID concept (i.e., the purpose of a system is what it does), if a system is a purely financial service or product, it doesn’t necessarily have to be politically decentralized because the goal of the system is just to provide a financial service above all else. You wouldn’t want everyone involved voting on how to do operations; that wouldn’t be a stable system. In many web3 projects, there’s a lot of mixing of political and social goals with financial goals, and blending these things doesn’t necessarily serve everyone.
Gitcoin is a project that BlockScience was deeply involved in. It was a project that involved IP, token holding, voting power, infrastructure, and other issues. BlockScience came on to support their fraud detection and defense, specifically designing and iterating a machine-learning algorithm to detect fake accounts. Because Gitcoin was experiencing a huge amount of fraud, and there wasn’t a known solution then, both BlockScience and Token Engineering Academy jumped in to try to solve it.
The algorithm was hugely technical, not something that just anyone could do. BlockScience went through the process of trying to hand it off to the community, and it took a full year to map the entire fraud detection and defense system, all of its functions, roles, and stakeholders, not to mention the technical documentation. Saying “hand it over to the community” sounds simple, but in the case of technical projects, it can be a very complex process.
In addition, there were no incentives or salaries in place. When it comes to progressive decentralization, operations and budgeting are the biggest areas that DAOs struggle with. Do you really want everyone voting on budgets, especially when there are things that the community at large doesn’t necessarily have any expertise in? It may be basic stuff for corporations or non-profits, but with DAOs there’s necessarily some reinventing of the wheel. There’s also separating day-to-day operations from what the purpose of the system is, and this is where we get into “governance surface” and governance minimization, rabbit holes that we can’t go down in this conversation.
But the point is that you don’t necessarily want all aspects of an organization to be decentralized; oftentimes, that just isn’t functional. This is a high-dimensional space of trade-offs. If you insist on having everyone voting on your budget, you probably won’t get a lot done. Still, at the end of the day, the real point is that all of this is very contextual, and it’s hard to make firm pronouncements.
What broadly needs to exist in the space to help people work through the complexities of progressive decentralization?
Investing in educational infrastructure is incredibly important right now. One of the biggest issues is information asymmetry. What are the ethical responsibilities of people launching tokens or building projects? We need to know that because this is new public infrastructure. The best thing we can do to have agency in these systems is to educate ourselves. There will always be an information asymmetry, particularly with technical subjects. Some very intelligent people whose voices we want to hear will never be able to read code, for example. But in this incredibly important and evolving space, it’s vital that we advocate for ethics and sharing knowledge.