Last year, [redacted] and I built a proposal for a large L2 organization. We put together a vision and key questions on what it would take to build a DAO. It was expansive and successful. A side-effect of this process, and the design exercise since then, is that it has forced me to articulate the crucial parts of building Networked Enterprises: organizations that have oozified into a web of interdependent partnerships with mutual ownership and accountability.

Networked Enterprises are, in my opinion, the real outcome of the DAO-fad from the past 5 years. Similar to Gabe Shapiro’s BORG, it’s about entanglement and literal encoding of control. Like traditional holding companies, it’s about entanglement of assets. Like FAANG, it’s about digital capabilities. Unlike all of these, it’s about something more subtle: becoming interface agnostic by leveraging the blockchain.

Sometime between 2018 and 2019, I attended a lecture in London by Neil Shah that coined the term “interface agnostic.” I was never able to obtain the slides, but found the little-known medium article that hinted at the concept. The presentation, titled the same as the essay, was called “Machine Thinking, Conveyance, and the Future of Design” and outlined a prescient view of the changes we are experiencing today in how we approach design. In Neil’s words (emphasis mine):

In Machine Thinking, we are designing a set of interaction models for a machine to learn, output, and interact with a human (or other machines) with potentially infinite variations and outcomes.

As designers we naturally labor over the finest details of our work. But in an A.I. First model, we will not know many of the details. Machine Thinking places the emphasis less on the perfection of the design output, and more on the robustness of the design system.

I’ve encouraged teams to spend more of their cycles creating strong system maps, agnostic of interfaces, that outline the interaction model. In Machine Thinking this becomes even more valuable.

A simple Machine Thinking exercise: what are the fewest number of components (interaction or interface) that are capable of solving the greatest number of known transactions?

Neil Shah, Machine Thinking, Conveyance, and the Future of Design

This idea has persisted with me over the last five years. In these notes, I am trying to articulate what exactly this means in the context of emerginig organizations: swarms, memetic tribes, online communities, onchain organizations, and networked enterprises.

Years later, during the NFT bull run of 2021, there was a movement of onchain SVGs. These NFTs, such as Blitmap, are fully encoded into the smart contract. A variety of collectors and builders wrote about the innovation. Simon, an onchain veteran, wrote at length about the topic:

NFTs, as unique items on the blockchain, has a URI that points to data containing the metadata & the corresponding visuals. This URI can be an HTTP link, pointing to a video or image hosted on a normal server, or other services like IPFS (hash-based addresses), or Arweave (incentivized hosting of hash-based content).

There is another way, however, a format that's become increasingly popular: the usage of a data URI. These URIs contain all the information within it. There is thus no server at the other end. Using data URIs has allowed NFT creators to experiment with putting all the content related to an NFT 'on-chain'. It adds a vector of permanence to the art. If Ethereum continues, it does not need ancillary infrastructure to support it. A common format, currently, is to store the NFT visuals as SVG in the data URI, since most browsers are able to natively parse it.

Simon de la Rouviere, Flavours of On-Chain SVG NFTs on Ethereum

I’d like to take a moment here to understand the implications that Simon, long-term onchain builder, is describing: “There is thus no server at the other end… most browsers are able to natively parse it.”

What we’re talking about here is the same concept that Neil mentioned: being interface-agnostic. The code is the visual, and a plurality of interfaces (browsers) can interpret the data without a specialized application. The magic lies in adopting the SVG standard and base64 encoding. By investing into adding dense, complex, yet standard data into the code itself, the creators were able to minimize reliance on additional services: no server and no special app. This also means that anyone can reference the art itself and transform it. It is directly composable and interoperable; there are no software middlemen or dependencies.

That same year, Dom (the creator of the fully onchain Blitmap NFTs), also experimented with creating an organization fully onchain. The project was called Nouns. Each day, there would be a single auction for a fully onchain NFT (the Noun of the day). The organization was intended to grow steadily: “A Noun a day, forever.” Proceeds from the sale would go into a community treasury — governed by a smart-contract — which NFT owners could vote to disburse to “proliferate the meme” and therefore raise the demand for future Nouns. In this case, like the NFTs above, the organization was fully encoded onchain in smart contracts. Their main website, nouns.wtf, simply read the contracts and allowed users to interact with them. Others built other websites over time, such as nouns.camp.

Although they were initially successful and gathered millions of dollars in ETH, the community struggled in 2023; they were unable agree on a path forward. Their main digital gathering space, hosted on an “offchain” (i.e. web2 platform) called Discord, became increasingly hostile. A decision was made by the interface administrators, who were also the Noun’s founders, to close down the interface.

“If nuking the Discord kills Nouns, then this was never a neutral protocol to begin with and I prefer a swift death,”

Steve, owner of Noun 309

Shutting down the central mode of communication would kill most online communities. Even IRL, destroying a temple can be the end of its community. However, Nouns had an advantage: their organization was encoded on a blockchain protocol, not reliant on any specific interface. In other words, they were successfully interface-agnostic. Discord was one of many interfaces, even if it happened to be the primary gathering spot at that point in time. Having their organization encoded onchain meant that anyone could interact with the organization, independent of active interfaces.

Nouns has survived since then, and many argue they have flourished (I tend to agree).

These two examples are prescient to the coming changes in how we build and manage content and organizations.

Since the early 2000s, many enterprises became reliant and locked-in to specific interfaces; they became co-dependent with SaaS solutions. If Google, Slack, or Salesforce disappear, they would likely as well. They are fragile, relying on interface-specific file formats, private servers, and exclusivity agreements. As a result, the fate of these enterprises are often entangled with the fate of their service providers. Moreover, the incentives reinforce this vicious cycle: Enterprise SaaS wants lock-in. The higher the switching costs, the better.

Now, we’re entering a phase where organizations are built on data and standards that are independent of the interfaces servicing them. Like Nouns, we can expect the next generation of organizations — Networked Enterprises — to thrive on newfound freedoms, unchained from the interfaces competing to provide them value.