Music DAO Deep Dives, Pt. 3: Mudd DAO’s generative music/Web3 infrastructure

This breakdown is part of an ongoing, members-only interview series focused on artist and label DAOs, as part of Season 1.5 of our ongoing collaborative research on music and Web3. The goal with this project is to make collective sense of the emerging, fast-moving ecosystem of music DAOs — not only giving structure to the current landscape and future possibility space for music DAOs from the perspectives of function, tech tooling and organizational design, but also identifying critical needs in the landscape that are still going unaddressed.

All interviews in these series are conducted and written by members of the Water & Music community, and break down a music DAO’s approach to community design and onboarding, tech tooling, governance, treasury management and more. Core contributors lead weekly, members-only research calls every Wednesday in our private Discord server to dive deeper into takeaways from these interviews.

Mudd DAO is a self-described “decentralized, community-run record label at the forefront of collaborative creation and generative music.” The long-term vision of the DAO is to help more collaborators make more music together than ever thought possible through generative creative techniques, and to give that music a platform via decentralized distribution infrastructure. The DAO is also innovating on how code can be licensed and utilized in Web3 environments.

For this project, we interviewed the DAO’s eponymous founder, i.e. independent artist, Julian Mudd.

Our main takeaways:

• Mudd takes a more maximalist approach to building out the DAO. The ultimate vision is to make the organization “fully self-functioning … If you build them right, they don’t need you.” This is in stark contrast to the approach of many other music DAOs, whose long-term plans rely, in part, on retaining a strong sense of leadership and direction from their founder(s).

• Mudd is building the entire tech stack for the DAO in-house. His main reason is what he refers to as “technical debt” — i.e. “the coding you must do tomorrow because you took a shortcut in order to deliver the software today.”

• The greatest pain point for Mudd DAO has been recruiting full-time employees. Web3 developers often have commitment issues and prefer to work on multiple projects at a time.

• The underlying code Mudd DAO develops is just as much a part of the “product” as the resulting artworks. The DAO is experimenting with new Web3-native frameworks for software licensing. The code behind Mudd DAO’s generative music creation can be licensed, in different tiers, by purchasing a certain number of Julian’s genesis NFTs. As long as the purchaser holds x amount of NFTs, in lieu of a traditional SaaS model, they can use the software.

Mudd DAO is a decentralized music organization at the forefront of collaborative creation and generative music, founded by eponymous independent artist and developer Julian Mudd. Julian — a 24-year-old recent college grad who recently quit his day job to focus on Web3 full-time — prefers to say “collaborative creation collective” over a more traditional, skeuomorphic term like “record label,” but will still toggle between the two terms.

In early December 2021, Julian launched a genesis NFT project around his debut song “Growing Pains,” utilizing the generative, Web3-native music tech that will ultimately power the Mudd DAO. Up to 1,000 one-of-one versions of “Growing Pains” can be minted as NFTs by collectors; each mint results in a unique combination of five primary musical elements, namely bass, chords, fills, percussion and vocals. The Mudd DAO will ultimately be home to many generative music projects, solo and collaborative, across many creators.

Mudd DAO’s in-house technology will empower its community members to come together and collaborate asynchronously on music, using a combination of Web3-native and generative techniques. The basic workflow envisioned is as follows:

1. An artist submits stems to the platform, all of which are attributed to that artist’s wallet. Anyone holding a Growing Pains NFT can experiment with the tech and upload stems, adding to the DAO’s collection.

2. Early Mudd supporters — including genesis NFT holders, DAO POAP holders and more — can earn Mudd tokens by validating submissions, which entails judging whether foundational stems meet a certain quality standard and whether or not stems from different submissions can go together. This quality assessment is inherently subjective, and therefore cannot be automated.

3. At the moment of mint, stems are combined to produce a totally novel and unique piece of music. All stems used are attributed back to their original creators on-chain.

Artists, either by purchasing a certain number of Julian’s genesis NFT collection or by being “inducted” into the DAO via a community vote, are able to create their own personal, generative NFT collections using Julian’s tech — in the spirit of other initiatives like Async Blueprints that give artists the off-the-shelf tools to create their own PFP projects.

By holding a certain number of the genesis NFTs, an artist will also be able to lease the DAO’s tech to use on a personal website, with the potential to get an API key or access to certain tools from the DAO. This roadmap demonstrates Julian’s ambitions to innovate in the realm of software licensing alongside that of music creation, and is a great example of how creators can monetize not only generative art NFTs, but also the underlying technology/algorithms used to make them (a dynamic that we covered in $STREAM S1).

A tooling-centric approach to music DAO development also paves the way for innovative approaches to DAO<>DAO partnerships. We’re seeing this with some other music DAOs like what Topshelf Records is building with their pay-what-you-want music streaming bot, Tone; Mudd DAO is exploring a similar initiative in its budding partnership with artist collective Songcamp. The value exchange in this case is crisp: Mudd DAO is looking for a vibrant, big, creative community to use and hopefully embrace their tech and Songcamp is looking to make music and experiment on the edge of what is technologically possible.

But let’s step back a bit — why does it make sense to structure the Mudd DAO’s core product as a DAO in the first place?

According to Julian, a DAO is “a structure and means where decision-making is in the hands of more than one person and facilitated by technology,” the execution of which is automated.In the context of Mudd DAO,where the means by which music is made is decentralized in of itself, it only makes sense for the decision-making structure around that creation to be decentralized as well. In fact, Julian has a very idyllic, and almost purest decentralized, vision for his DAO as a “fully self-functioning organization … If you build them right, they don’t need you.”

Everything in Mudd DAO will centralize around the Mudd governance token, which has yet to be minted. The Mudd DAO community is currently 50/50 when it comes to Web3 literacy — one part up-and-coming musicians who have heard about NFTs and crypto but are not onboarded, and one part people who are fully onboarded into Web3. To help drive early sales and membership, Julian is personally onboarding people, a process that includes everything from wallet setup to general “crypto lessons.”

Incentivizing artists/producers to upload and validate stems is top-of-mind for Julian, who has set up multiple mechanisms for contributing musicians to get paid by the DAO. Artists can earn the token by contributing work to the DAO (i.e. validating and potentially contributing stems), by generating yield via owning Mudd DAO NFTs and/or by purchasing the tokens on an exchange.

When a collector mints an NFT, the sale also gets distributed partially to the attributed musicians, and partially to a treasury that automatically purchases the Mudd governance token, essentially buying up the floor with every sale. Collectors will also have the ability to distribute the songs they mint through Mudd DAO, again resulting in automated payments to the original creators through onchain stem attribution. In addition, holders of NFTs will yield Mudd governance tokens, which helps to prevent dilution as the liquidity pool increases.

From a governance standpoint, Julian plans on implementing multiple different levels of voting classes — i.e. not only token holders, but also “super token holders” (exact mechanics TBD) — who will vote on large-scale organization and financial matters. Super token holders might also have the ability to withdraw funds from the DAO treasury (the equivalent of a signer on a multi-sig wallet), which an average token holder would not. The general voting model will be weighted or quadratic voting, with on-chain guardrails (e.g. restricting the number of tokens that can be purchased within a given timeframe), to prevent whales from exploiting a pay-to-play opportunity.

Julian is a highly technical founder, who plans to use in-house tech and a custom ERC-20 contract to spin up the Mudd DAO token. He also plans on makingcustom in-house web applicationsfor ancillary DAO tools,in lieu of using third-party software such as snapshot. It was illuminating to learn from Julian about the concept of “tech debt” — in his words, “the coding you must do tomorrow because you took a shortcut in order to deliver the software today.” In his view, building in-house gives the DAO more control over their infrastructure, and therefore makes the DAO less exposed to outside failure.

Julian is still debating which decisions truly need to be on-chain, such as which artists will be represented by the DAO or whether a smart contract clause should be edited. Proposing an artist to be supported by the DAO will require staking a certain number of tokens, as will validating stems. If a collector has a complaint about the song, resulting from a poor stem combination, there will be a review board. The validator who validated those stems could be penalized. There will likely be lots of redundancy in the beginning, meaning the same information will be stored both on- and off-chain. Mudd will store as much on chain as possible, but in the instances where data is not, everything will be stored off chain in the beginning and perhaps put on chain at a later date.

When asked about his greatest technical pain points, Julian actually brought up not 0s and 1s, but a rather people-centric problem: Recruiting. “It’s tough to hire full-time people,” he says. “This is a self-funded project, with no healthcare or anything like that.” Devs in the Web3 ecosystem seem to have commitment issues, and prefer to work on multiple projects at a time. As a runner-up to recruitment woes, it’s also expensive to launch a custom project, as deploying a smart contract can cost anywhere from $1,500 to $2,000.

All in all, Julian has taken some really innovative ideas on the journey from concept to creation, and has cultivated a nascent Discord server with support from several successful creators, entrepreneurs and collectors. The extent of which there is appetite for these experimental ideas is up next. At this point in time, Julian is the only person who has had the opportunity to use Mudd DAO’s technology so far. The real potential for that magic moment is when all his holders will have the opportunity to collaborate — not just directly with each other, but with the algorithms themselves, letting the code take the wheel.