The Evolution of Decentralized Autonomous Organizations (DAOs)

by piero scaruffi
Cognitive Science and Artificial Intelligence | My book on A.I. | My book on consciousness | Human 2.0 | Bibliography and book reviews | Contact/feedback/email
(Copyright © 2021 Piero Scaruffi | Terms of use )


Blockchain technology has spawned a chain of events (sorry for the pun) that was probably unintended by its inventor: the Bitcoin blockchain led to Ethereum's smart contracts and dApps and these led to "decentralized autonomous organizations" (DAOs).

The historical context helped make this happen. The end of the Cold War and the advent of the World-wide Web (both in 1991) generated a new mood among economic thinkers, beyond the binary contrast between communism and capitalism. The management consultant Peter Drucker discussed the rise of "the autonomous community organization" in "Post-Capitalist Society" (1993), arguing that decentralized organizations are more efficient that centralized ones and that there was a need for a third kind of economic sector besides the private sector (companies) and the public sector (government). That third sector would be the "autonomous" sector. Besides the economic factor, Drucker viewed the autonomous community organization as a form of "meaningful citizenship". A similar idea was embraced by the economist Jeremy Rifkin in "The End of Work" (1995), who argued that the community-based sector had always existed and had been crucial for shaping the USA. In 1995 the RAND Corporation produced a report titled "Tribes, Institutions, Markets, Networks" which argued that human civilization has progressed over the centuries through four organizational forms: the kinship-based tribes/clans, the hierarchical institutions, the markets and now the collaborative networks. By the age of Bitcoin, there was consensus that information technology had the power to forge new economic models and subvert the old capitalist order. Throughout most of the 20th century, communism was perceived as the only alternative to capitalism, but the potential of "social power" to represent a third pole distinct from economic power and state power was increasingly recognized. The problem for sociologists, economists and others was to come up with credible institutional alternatives to the existing structures of power.

The Great Recession of 2007 helped crystallize the feeling that something was fundamentally wrong with capitalism and therefore motivated a search for alternatives to the ruling capitalist models. Bitcoin and its blockchain technology debuted in 2009, just in time to provide an alternative to centralized institutions. In November 2013 the 19-year-old Bitcoin developer Vitalik Buterin published the Ethereum white paper, introducing the world to smart contracts that did not depend on a centralized authority.

A couple of months earlier Daniel Larimer had written about "decentralized autonomous corporations" (paper) and simultaneously Vitalik Buterin had written about Decentralized Autonomous Organizations (paper). Those were the papers that showed a way to realize the kind of "distributed autonomous organization" envisioned by Ori Brafman in "Starfish And The Spider" (2006) and other skeptics of the hierarchical corporation: by using smart contracts of the kind that Ethereum was about to make possible.

The key intuition was to use "secure multiparty computation", introduced by Chi-chih "Andrew" Yao at UC Berkeley in the 1980s ("Protocols for secure computations", 1982), in turn based on the "secret sharing algorithm" invented by the Israeli cryptographer Adi Shamir ("How to share a secret", 1979), to run the algorithm that controls the DAO (e.g. to generate Bitcoin addresses and sign transactions).

Such "decentralized autonomous organizations" or DAOs are unmanned organizations (no office, no staff) that run under the control of an incorruptible algorithm. The algorithm is, in turn, implemented in an open-source software that can be "audited" (verified, controlled) publicly. DAOs are autonomous: their functioning is not controlled by their creator; they are censorship-resistant: there is no central authority that can censor them; DAOs are self-enforcing: their operations follow the rules embedded in the code and only those rules; DAOs are eternal: they will continue operating as long as there is a computer running the code. Operational costs are low and the DAO can be located anywhere in the world.

As Larimer pointed out, Bitcoin itself was a proto-DAO that issued 21 million shares (the final number of bitcoins) to shareholders (the owners of those bitcoins) and employs miners and validators (who gets paid with new bitcoins for adding blocks to the blockchain). Bitcoin is not a real DAO because there is little decentralized governance: bitcoin owners cannot vote on the direction of the project. A real DAO has a more sophisticated algorithm of governance instantiated in a set of smart contracts. Anyone who buys its cryptocurrency has the right to vote on the DAO's future, proportional to the amount held.

A traditional corporation has a mission statement which is supposed to guide its business and is supposed to be enforced by the board of directors and be agreed upon by the shareholders. A DAO encodes that mission statement into a blockchain-based piece of software (a smart contract) that runs on the network, rewards people who perform some tasks, and also makes money for the people who join it. This is because the code, the smart contracts, and the transaction history of any decentralized protocol are visible by any node of the network.

The difference in transparency is colossal. The "shareholders" (or better token holders) of a DAO have visibility of its "balance sheet" all the time because it is stored on the blockchain, whereas a public corporation publishes financial statements every now and a private corporation doesn't even have to do that.

Not only can people vote on proposals, but they can even "fork" the protocol if they disagree with the decisions taken, i.e. literally spawn a separate DAO. This is much more "check and balance" than comes in the world of corporations.

Because the ones who contribute, work, discuss and invest are the ones who get rewarded and accumulate voting power, a DAO is a true meritocracy.

The legal status of DAOs that exist only on the Internet differs from jurisdiction to jurisdiction. For example, in 2021 only one state of the USA, Wyoming, recognized DAOs as legal entities (a form of limited liability corporation).

In 2020 OpenLaw (founded in 2017 by Aaron Wright and David Roon) launched the LAO, a project to create a new wave of DAOs as limited-liability autonomous organizations.

The first successful DAO was Larimer's BitShares, launched in 2014. The first DAO on Ethereum, called "The DAO", was coded by Christoph Jentzsch and launched in April 2016 but shut down a few months later due to a hack. "The Dao" was just one particular DAO, neither the first one nor the best one, and its widely publicized hack gave a bad reputation to the DAOs. It was meant to operate like a hedge fund for the blockchain world. It attracted a lot of publicity because it was the most successful investment crowdfunding in history until then, raising the equivalent of $150 millions (12 million ether).

Other decentralized autonomous organizations (DAOs) running on the Ethereum blockchain such as DigixDAO and MakerDAO (which was still in test mode) were not affected by the attack.

In January 2018 Vitalik Buterin also proposed the DAICO, a combination of DAO and ICO, to mitigate the risk of scam ICOs.

The core function of a DAO is its decentralized governance. While the set of smart contracts that coordinate its collective behavior is the foundational element of DAO governance, they are not the only one: smart contracts also define the creation and distribution of the native token, particularly its role in voting mechanisms and its role in incentivizing the behavior of participants and stakeholders. One's voting power is generally proportional to how many tokens one possesses. And one's motivation to improve the DAO is generally proportional to how many tokens one possesses. The DAO issues tokens based on participation, contribution and investment, and then token holders have the power to submit proposals and to vote on proposals.

For example, Uniswap DAO (September 2020) issued 1 billion governance tokens (60% to the Uniswap community member, 22% to employees and consultants and 18% to investors) and any Uni token holder was entitled to submit a proposal for changes or additions that was approved if it received enough Uni votes (40 million Uni votes). Compound (June 2020) issued 10 million governance tokens of which more than 5 million went to Compound users and 2,312 were given daily as rewards to lenders and borrowers. Again, all token holders were entitled to vote on proposals and there was a threshold for a proposal to be implemented (400,000 votes). Similarly, Aave DAO (December 2020) issued 16 million Aave tokens and a proposal required a certain threshold of votes to be accepted. MakerDAO differed in that the total supply of tokens was not fixed (initially, one million tokens were minted) and it used both on-chain and off-chain voting (in which any member of the community forum can vote, regardless of whether they owned governance tokens).

Several platforms emerged to help build DAOs, typically offering a libray of modular governance plug-ins (smart contracts) that could be combined to achieve the desired governance scheme: Luis Cuende's and Jorge Izquierdo's Aragon (2017), the quintessential Ethereum-based DAO for creating DAOs with basic governance features; Matan Field's DAOstack (2018), targeting large-scale DAOs; Ameen Soleiman's MolochDAO (2019), a "minimum viable" DAO with simple smart contracts that could be forked to create other DAOs; Dekan Brown's DAOhaus (2019), a platform to deploy MolochDAO clones (to "summon" a Moloch DAO); Colony (2019), developed in London by Alex Rea and Collin Vine; etc.

2020 was the year of the DAO.

In 2020 Jesse Walden suggested that projects should not start as DAOs but get to DAO state through a process of "progressive decentralization". At the start-up stage, the focus should be on developing a good product, and control should be in the hands of the key developers: no decentralization. Once a good product exists, the startup should open up to external developers, rewarding whose who contribute, so to create a community of developers that extends beyond the original team. Only at this point should the project become a DAO and distribute tokens to anyone willing to abide by the DAO's smart contract. Now the DAO is controlled by the community according to the governing smart contract. For example, Uniswap became a DAO only two years after its creation: for the first two years its founder Hayden Adams made all the important decisions, and token holders couldn't interfere in the governance of Uniswap. In 2020 Uniswap introduced a token that gave governance rights to all token holders.

MetaCartel (2019), a sort of venture-capital firm for dApps, was forked out of MolochDAO, but also opened a classic Delaware Limited Liability Company called MetaCartel Ventures.

A DAO can potentially have millions of members. "Naive" decentralized governance could rapidly become utopia as the number of members increase because of scalability issues. A large-scale DAO would be ungovernable. Several methods were proposed for efficient collective decision-making at scale. For example, Matan Field's DAOstack (2018) implemented a reputation-based governance algorithm called "holographic consensus" that consists in letting a small group make the decision (via a game of betting and forecasting) on behalf of everybody while ensuring that the decision truly reflects the "consensus" within the community (every part of a hologram contains all of the information of the whole).

The first test for DAOstack's holographic consensus was the dxDAO (2019), built by Martin Kueppelmann of Gnosis. It was meant to be an experiment in building the first large-scale DAO, and a "real" DAO. The original DAO ("The DAO") was in practice only a glorified crowdfunding operation. The goal of the dxDAO, instead, was to create a real organization, starting with 232 investors/owners (mostly anonymous people identified only by an Ethereum address), like a startup founded by 232 people. Each of these "founders" received full executive control. Furthermore the dxDAO was open to anyone who wanted to join, with any new member automatically becoming an owner and "executive". The dxDAO was setup to play the role of incubation platform for DeFi projects, initially the trading protocol DutchX and the automated market marker Swapr (derived from a fork of Uniswap and launched in 2021) that introduced an incentivization method called "conditional farming". The dxDAO was "decentralization maximalist": it was entirely run on the blockchain. One simple advantage was legal: while most decentralized exchanges were run by the developers, who could be prosecuted by regulators, the dxDAO was not run by any person: it ran itself. Someone built and released the code, but they were not operating it, and therefore nobody was liable for what the dxDAO was doing or not doing.

The dxDAO created a reference model for all wanna-be DAOs. it provided a token model, incentives, and governance, all on the blockchain, to facilitate self-running cross-jurisdictional cooperatives.

In 2020 a de-facto standard emerged for polling the user base of crypto projects: Snapshot, developed by Jorge Izquierdo of Aragon and Fabien Marino of Balancer, and based on IPFS, a decentralized storage system. The will of token holders was then implemented using a multi-signature wallet like Martin Kueppelmann's Gnosis Safe (2019), the de-facto standard for storing funds on Ethereum (the equivalent of joint bank accounts in traditional finance).

In 2021 a popular way to launch a DAO consisted in four steps: create a token on Denis Nazarov's Mirror.xyz (in theory a blogging platform on a blockchain, but in practice also a user-friendly tool to launch a new token); store tokens in a Gnosis Safe multi-sig wallet; setup a Snapshot space for governance; create a Discord channel for community discussion. In fact, both the token and the governance were optional: a "minimum viable" DAO consisted of a Discord server and a Gnosis Safe multi-sig. The token was the next step, representing co-ownership and governance rights.

Before and after the launch, there was also a need to manage the community, for example with Collab.Land (developed by MolochDAO's cofounder James Young and Raymond Feng in Los Angeles).

DAO governance faced the same issue that blockchain systems had faced from day 1: a small number of wealthy users controlling most of the computing power of miners could disrupt the blockchain; a small number of wealthy members acquiring control of most of the DAO's votes could disrupt the smart contracts at the core of the DAO. To avoid such a hidden "cryptocracy", Vitalik Buterin endorsed "quadratic voting". Quadratic voting, introduced by Glen Weyl with the paper "Nash Equilbria for Quadratic Voting" (2014), goes beyond the binary approve/oppose kind of voting and indirectly assesses how strongly voters feel about an issue: voters can vote as many times as they want, but they have a limited number of voting tokens and the cost of each token increases exponentially.

The traditional "exits" for startups are 1) the IPO and 2) acquisition by another company. In 2019 Nathan Schneider, director of the Media Enterprise Design Lab at the University of Colorado Boulder, coined the term "exit to community" to speak about the third option that opens up for blockchain-based startups: the option to transition from ownership by the investors to ownership by the users/ stakeholders via the issuance of crypto-tokens.

Silicon Valley startups established the norm that employees should be co-owners: nothing motivates a worker more than being a co-owner of the company. Hence startups issue stock options to employees. The ICO and the DAO expand that notion of collective vested ownership by making the user/stakeholder an additional co-owner, so that the user too will feel motivated to contribute to the success of the project (by providing feedback, by spreading the concept, by building communities, and in general by gifting time and work). This form of co-ownership ensures motivated cooperation. The user is also involved throughout the stages of the project, which ensures better alignment between product features and user needs. The ICO and the DAO are further steps towards "the ownership economy" (a term popularized in 2020 by Jesse Walden's essay) in which "interest" translates into "ownership" and ownership translates into participation.


See also Blockchain technology and A Timeline of Blockchain technology (with timelines about new media art)
Back to the index