Open source and web3, simplified
An attempt at "dinner table" level conciseness for one of the most important arguments for web3
A friend recently asked me how web3 and crypto might solve some of the issues that make existing social media so harmful. I gave him a response, but tripped over my words — which I’m sure didn’t help him and also made me feel somehow guilty for how excited I am about this space. I’m confident it’s a feeling other people have when they are in similar situations, and it made me want to spend more time writing out as simple as possible arguments on specific topics, basically “dinner table” level logic. For me, one of the most important of these is around open source, so that’s where I’ll start. As a caveat, many people are far more qualified to write this than me - my intention here is just to try and clarify it for myself, and for you all in the process.
Open source technology provides many of the foundational bricks of basically every product you use. As a result, it is powerful to draw the analogy to crypto/web3 as an extension of open source — you likely have or will soon come across that analogy in arguments for why all of this work is worthwhile. The problem is just that open source is a niche topic, and for most people it doesn’t immediately give them a clear mental model to get excited about. It feels like nerd-speak that doesn’t really matter. I’ve been living at this intersection for a while, so I’m going to try and make this analogy as simple as I possibly can, such that you can either use it or at least understand it when you see it next.
The rise of Open Source
There are three things that you need to understand about open source as it applies to web3:
The what: open source and how does it impact our daily lives
The how: the trends that have made it possible
The “why make it better:” the public goods problem that plagues it
Once you understand those things, it’s clear that web3 is riding the same winds as open source as well as trying to improve on the model in ways that will help decentralized communities eat an even larger share of the economic pie — and spur the pie’s overall growth and inclusivity. I’ll go through each point, but in order to stick to my promise of a clean mental model I’m going to do a fair amount of simplification at every step. Apologies to the sticklers out there :)
The impact of open source
Open source is just as it sounds — software where the source code is open for anyone to see. The code for Linux, the dominant open source operating system, is here if you'd like to read, download, or adapt it for your own purposes. Think of open source like Wikipedia but where instead of writing an encyclopedia, groups of people band together to write valuable software. One of the things that makes open source very useful and trustworthy is that at any time you can always “fork” the project aka make a copy of it for your own purposes. As a result many of these open source projects become the dominant tool in an area since people don’t feel as nervous relying on a monopoly (this will be important later).
It’s hard to overstate the impact of open source. In a recent Forrester report, 96% of companies said “open source is important, very important, or mission critical to their business” and 98% said they plan to increase or maintain their reliance on open source in the coming year. Linux for example is the operating system of choice for modern servers - one report suggests that 96.3 percent of the top 1 million web servers are running Linux. Bottom line: open source is a critical piece of basically every technology currently built.
The trends that make open source possible
I would highly suggest reading “The Cathedral and the Bazaar” or “Coase’s Penguin” if you want a much more detailed view of the tailwinds that blow open source forward. Simply put, they are:
Decreasing cost of writing software
Near zero cost of distributing information online
Decreasing cost of coordination via technologies like Git or Wikis (version control that allows many people to work in parallel)
In essence, what these things mean is that diverse groups of people can all come together and produce software that is as good or better than the closed versions that traditional firms create and sell. They do so either just because they think it’s fun, or because they are themselves users of the software and don’t want to pay for it, or because they can earn a living doing so (more on this later). Bottom line: the trends that power open source are decades in the making and only getting started - this is not a flash in the pan.
Why make open source better?
For the most part, compared to the value they create or to how much they could make in “traditional” tech jobs, independent open source developers get paid like shit. The result is that open source projects are under resourced compared to their value, which leads to the other way you can feel the impact of open source — when things go wrong. The “Heartbleed bug” introduced a loophole into the software of 20% of active servers, leading to a huge list of attacks including 4.5 million stolen patient records from US hospital systems. A recent bug in a key protocol called “Log4j2” exposed everyone from Apple to Minecraft. The xkcd comic below is funny, scary, and very accurate. Bottom line: open source, while hugely valuable, is struggling to keep the lights on in many critical projects due to the public goods problem.
Web3 and the service model
Despite what its label might imply, there are viable business models within open source. In fact there are many multi-billion dollar companies built on top of open source projects. One of the biggest is Red Hat, a company that sells a service to enterprises that want to use Linux. Basically, enterprise customers would rather pay a fee to get a premium SaaS type model of the product that has the security and privacy functionality they need, rather than doing the work of using the open source version and building that functionality themselves. Red Hat was acquired by IBM in 2019 for 34 billion and generates billions a year in revenue.
One problem with these centralized business models is that they are essentially parasitic to their decentralized counterparts - what makes Linux so valuable is that with thousands of contributors it can offer a level of stability and integrations that other operating systems can’t. If Red Hat were to try to produce Linux on its own via its own employees, it would likely fail. It’s the same reason that Wikipedia entries outcompeted Britannica or other centralized encyclopedias - the swarm beats the hierarchy at creating certain types of goods. This is where the web3 argument comes in.
It should seem odd that many Linux developers are living on modest donations (from Red Hat and elsewhere) while Red Hat adds a layer directly on top and sells it for billions of dollars a year. Why don’t all of those Linux developers just build the enterprise layer themselves and use the profits to fund their work? To answer that question, consider the steps the Linux community would have to take to make that work:
Develop some sales/marketing team to deal with enterprise clients
Find developers that are willing to do the less sexy work of building an enterprise service business, likely via compensating them competitively with the private market
Figure out how to reward everyone proportional to their contribution if they do end up creating this multi-billion dollar company
These are all essentially coordination costs. And coordination is what companies do well, so you might set up a separate company that does these things. In fact, the challenge of distributing equity to casual contributors of Linux all over the world would be such that you’d probably have to exclude them from the upside and maintain this parasitic relationship with the open, decentralized project. In other words, you’d land exactly where you started.
Now consider the same scenario but where you have essentially infinitely flexible and inexpensive ways to coordinate around digital ownership. This is the promise of web3 — let’s run through those same steps above but via tokens:
Create a DAO that looks by all respects like a traditional organization except that it uses tokens for ownership and governance.
Use those tokens and the cash the DAO creates to pay developers of unsexy work and the sales/marketing team.
Reward contributors to the open source version with a token as well
Tie revenue from the enterprise version directly to the token so that everyone gets paid in proportion to the success of the overall project (open source + service model)
Yes the crypto space still needs to evolve, but all of these things are doable now for those willing to be pioneers. There are tools to create a DAO in one click, ways to tie revenue directly to a governance token, and systems to reward contributors according to their level of contribution as decided by the community.
DAOs and the beer business
Richard Stallman is as close as it comes to a godfather of open source. He founded the Free Software Movement and the GNU project, one of the original open source operating models. The reason the movement didn’t maintain its title of “free software” comes down to Stallman’s famous quote “free as in speech, not as in beer.” In essence, the point of open source is that it is permissionless and open, not that it’s free like “free beer.”
Despite Stallman’s quote, and despite open source’s continued growth, it has remained largely free as in beer as well. But that’s not because companies just refuse to pay for services and instead prefer to use source code themselves - in fact, the dominant model for software these days is “SaaS - literally, software as a service.” Instead of buying a permanent licensing Excel, companies pay a monthly service fee to use the current version of the product (usually in the cloud). Companies will pay for convenience, and that’s exactly the business model you can build at scale with open source.
To tie it all together, the reason that open source has remained free as in beer is because the cost of coordinating between an open source project and its service based counterpart has been prohibitively high. This is where tokenized, decentralized communities have such a huge opportunity. There are already big businesses using this model of tying a service business to a decentralized protocol, from Braintree, Parsiq, most of the centralized exchanges, to Ethereum itself. Those services business take many shapes - for example Ethereum the service is providing computation to a world of applications, whereas on Braintree the service is helping traditional organizations interact and route fiat currency through the protocol.
Not all of these platforms themselves are open source yet, but what’s important is that the revenue that gets tied to a service model feeds back directly on the value of the native token. In that setup, the incentives to open up the source code over time are huge, because it allows a more vibrant developer and service based ecosystem to exist and to trust that the project will continue to operate in the network’s best interests (because if it doesn’t, it can be “forked” aka copied easily).
Tokenized, decentralized communities will not be the only viable model going forward. Nor are they the only innovation possible via crypto. Bitcoin is digital gold, and NFTs are digital safes that you can’t put anything inside — those primitives have implications well beyond the future of open source. But the reason I feel this one particular area is so exciting is because decentralized communities are not just building open source, they are building so much of our modern lives in ways that could be vastly improved.
And it’s not just code. Consider Wikipedia, Facebook, Uber, Linux — all of these are projects where a few people work to define the rules of the platform and then a huge number of people participate in building the value. The difficulty of capturing and distributing value in those communities takes many forms, from the Facebook or Uber model where the few capture huge value to the Wikipedia or Linux model where hardly anyone does. The structure I describe above could apply to Facebook just the same as Linux, with users participating in the value creation and capture.
The reason I got into web3 was because of my lifelong connection and commitment to science, the world’s most important public good. The biggest issue I see is that public goods that rely on donations or grants that aren’t directly connected to the value of the goods will always be under-resourced compared to their impact. What web3 offers for me, via more flexible and dynamic forms of digital ownership, is business models where decentralized organizations and public goods can be tied directly to equity and revenue streams — the merger of Red Hat and Linux, via tokens. If we can do that, the foundation of open resources we can build will increase dramatically and power a much more innovative and equitable world.
I’d love to hear your feedback on this post, specifically whether I made this particular topic simpler for you to the point where you feel you could talk about it at dinner. I would be happy to treat this as a living document that I can make shorter and simpler over time. And if you’re looking for more thinking on web3/open source, check out Scott Moore, Spencer Graham, or David Phelps.