Coase's Penguin is learning to fly
Building the Wikipedia of the future
A penguin and a scientist walk into a bar...
In 1937, British economist Ronald Coase published “The Nature of the Firm” which outlined why companies exist and what limits their growth. It was and still is so important because it provided a clean answer to when firms are categorically more efficient than markets. Sixty five years later in 2002, Yale economist Yochai Benkler wrote “Coase’s Penguin” which noted that the web enables a new decentralized sort of firm, one that can be more efficient than traditional firms or markets at creating open projects like Wikipedia or Linux. Linux’s penguin mascot is the “penguin” in Benkler’s title — his point is that the waddling, flightless bird presents a clear need to update Coase’s otherwise clean framework.
Benkler’s key addition in “Coase’s Penguin” was a framework for understanding why and when groups of individuals motivated primarily by the cultural significance of their work can be more efficient than markets or firms in the allocation of creative effort. In other words, how the web creates an environment where massively valuable projects like Linux or Wikipedia can exist.
Benkler called this new mode of work “commons based peer production.” He noted that it occurs via “removing property and contract as the organizing principles of collaboration.” By property and contract, he essentially means ownership — if you create a Wikipedia page you don’t own it in any way. That connection is important later on, since ownership as a term is the buzzword in the digital creator economy.
So by removing the costly process of implementing ownership, commons based peer production reduces collaboration costs to a degree where large groups of individuals can efficiently interact with vast seas of information to produce new information. Critically, these are problems where firms and markets tend to be horribly inefficient. In Benkler’s words:
“What peer production does is provide a framework, within which individuals who have the best information available about their own fit for a task can self-identify for the task. This provides an information gain over firms and markets, but only if the system develops some mechanism to filter out mistaken judgments agents make about themselves.”
Take for example the Wikipedia page on Yochai Benkler. Quite likely, it would be difficult for firms or markets to create what is a pretty high quality and efficient publicly available summary of his life and work to this point. But somewhere out there on the web there’s someone who cares about Benkler, maybe a student or fan of his, or maybe even Yochai himself, that cares enough to self-select into the creative effort of creating this public knowledge. At the exact moment when they feel the motivation to create this Wikipedia page, whether at night or on a weekend, they can do so with virtually no impediments, no contracts or bureaucracy to get in their way. Andrew Lin, in his phenomenal book “The Wikipedia Revolution,” called that moment “Wikipedia’s magic,” the rare time when “the “socio-psychological” reward of interacting with others, and the “hedonic” personal gratification of the task” come together. And that, in essence, is why peer production can be so much more efficient at allocating creative effort.
In addition to describing the efficiency gains, Benkler’s essay expanded in detail on the otherwise difficult questions Andrew Lin referenced about why people actually dothis kind of open source or open knowledge work when they could surely be making more money or living an easier life in other ways. He usefully connected Wikipedia volunteers to academic researchers, noting that academic research is still the “most important” and largest type of commons based peer production. A common difference is that scientists have to earn enough money to survive, but in both cases they are creating knowledge primarily because they feel it is culturally significant work. The benefits of being able to earn a living — even a meager one — via this work is something we’ll come back to, because it is a key focus of the current buzz around digital creators.
So while it seems obvious, one of Benkler’s most important contributions is that he provided a neat framework and even some equations for thinking about why and when people contribute to the building of open knowledge. We enjoy creating and curating impactful knowledge. Anyone who has worked with scientists, but also with most writers or other knowledge creators, knows this to be true. Money is important, but mostly to the extent that it gives you freedom to follow your creative interests.
Acknowledging the cultural motivations to produce knowledge also underlies a frustration many people have with systems which are built to commoditize it. Benkler himself suggested that the most dangerous form of this challenge to commons based peer production is “unilateral appropriation.” In other words, when the few profit off of the work of many, cultural motivations can vanish. In that sense, he predicted the current backlash to big tech social platforms where creators create billions of dollars of value that is then monetized mostly without benefiting them. Facebook would be founded two years after his essay, and while he may have predicted the reasons for the ultimate backlash he certainly didn’t predict just how big social media would get. Looking at his framework though, we can see just how that backlash might now play out.
Proprietary peer production and the rise of the social giants
Now that we’ve laid out what we mean by ownership, we can ask what happens in the case where it is implemented for financial return? Is there a situation when peer production can still exist even if the value of the content is unilaterally appropriated?
While Benkler didn’t predict the rise of the social giants, his framework did lay out the conditions for when unilaterally appropriated peer production is possible. Specifically, he looked at instances where the value of peer produced content is such that it is worth implementing ownership, and where it can be done in a way that preserves the motivation of the contributors. He called this “proprietary peer production” and at the time noted just one interesting but relatively small example of Xerox’s Eureka project.
Eureka was created in the early 1990s as a way to try and solve the problem of fixing printers and copiers. Anyone who has seen Office Space remembers this iconic scene where they finally destroy the printer after a continuous stream of problems with no apparent logical fix. Eureka was meant to solve that problem.
In essence, Eureka was an open wiki in which the thousands of copy service engineers, who worked full time fixing printers and copiers, could collaboratively build a knowledge base. Presumably the most well developed solution pages were “Kick the Printer” and “Wiggle the Little Paper Tray Thingie.” In that sense, it wasn’t so different from Wikipedia except for one key difference that Xerox owned all of the IP and was using it to help build a big business.
Why were all of the copy service engineers willing to put in the work for free to build this peer produced knowledge without much top-down direction, when it was clearly being unilaterally appropriated by Xerox? The answer is that they as a community were benefiting from it. Their motivations weren’t cultural significance — they simply wanted better and surely less frustrating professions.
This has been the status quo for a while — either you have entirely open, non-commercial projects like Wikipedia or you have unilaterally owned projects like Eureka. But now, thanks to lower costs of building products and lower costs of implementing community ownership, we are starting to see a new mode: community owned peer production.
More than just copyright innovation
When we hear “lowering the cost of ownership of knowledge” most people immediately assume we’re talking about copyright. But as Stratechery writer Ben Thompson and ethereum founder Vitalik Buterin have both rightly pointed out, the scarce resource online isn’t copyrightable content itself - it’s people and their ability to bestow legitimacy on the value of content. This is something you feel intuitively in science - reputation, not objectively provable “truth,” is the most powerful force in building shared knowledge. So lowering the cost of implementing ownership is, actually, about allowing people to own their online identities and the ability for those identities to legitimize information. Thompson makes this point specifically through the lens of NFTs: “If the creator decides that their NFTs are important, they will have value; if they decide their show is worthless, it will not.”
So how does this apply to peer production? Writers and knowledge builders like Packy McCormick and Mario Gabrielle are experimenting with ways to reward peers who add legitimacy to your work. By minting an NFT for an article and giving portions to it to all of the authors cited, you reward the legitimacy added to the piece via those citations. If you squint hard enough, you can see that they’re driving towards a vision in which a Wikipedia type project could generate revenue proportional to its social value and then allocate that revenue to its contributors proportional to their investment.
Anyone in science will notice the inefficiency which Packy is commenting on - where an foundational academic paper is commercialized downstream. In that case, the commercial entity might make billions while the academic paper which enabled it gets nothing but a citation. The funder who funded that prescient foundational research gets no direct return either. The value of that commercial innovation is only returned to academia via taxes which then fund the NIH, but none of that connects back to the research or researcher who produced the knowledge. Systems thinkers will quickly realize how inefficient this feedback loop is for innovation in comparison to more direct connections, like Packy’s experiment. There are still some serious questions to answer before these feedback loops can work at scale in academia, but it’s incredibly exciting to see new experiments launching every day aimed at this critical inefficiency.
One of the reasons I’m particularly excited that experiments in digital ownership are tending towards communities and peer production is because knowledge creation is a team sport. That’s something I know from working in some amazing neuroscience labs, my career in product development, and the past few years of writing online in various places. Since legitimacy is the scarce resource, communities of trust are the most powerful for building knowledge. They’re also the most challenging in terms of the cost of implementing ownership, which is why most of the initial experiments in web3 such as NFTs have often tended towards single-player mode — transactional use cases between a single creator and a fan rather than community ownership. We still need to answer questions like how to reward editing and peer review, whether that’s via social tokens or some other solution. Those are the first experiments I plan to launch and write about.
In the very last of 70 pages in Coase’s Penguin, Benkler dreams of a “common property regime” that would allow peer production projects to start looking more like a “cooperative, managed and “owned” by its participants.” He noted that there were “no models for such cooperative appropriation on a large scale yet” but that it would essentially allow for projects to deliver monetary reward to their members in addition to the cultural significance of the output. No doubt, monetary reward could actually decrease the perception of cultural benefit — not everything should be commoditized. But one key benefit to such a model would be that it allows its members to actually build sustainable lives based on their contribution, rather than simply volunteering out of the drive to produce something valuable. Very specifically, consider how tragic it is that the retention of young scientists in academia is worse now than ever. Better community ownership means a shift of the proportional benefit from the big platforms to the community contributors, and the most powerful impact of that is that more of those creators can sustainably invest in the work full-time.
I believe the dream that Benkler mentioned offhandedly is starting to develop, and to be clear it’s not developing just via crypto. The digital world makes it easier than ever for knowledge creators to capture the unique value they create via a variety of mechanisms, and increasingly new infrastructure enables creators to build shared ownership via collaboration in peer-production communities and teams. There are versions of this where we set this up in ways that would go too far in commoditizing social value, for example where the cultural motivations are drowned out by the financial ones.
When you think about it on a people level, this danger becomes clear. If those who have been willing to participate in peer production purely for the societal benefit are no longer interested given its financial nature, then we lose the people who are most likely to pursue impact over hype. But if we can do it in a way that allows more of those people to build lives around culturally motivated pursuits, then we win enormously. My feeling from the past few months of this rabbit hole is that it’s decidedly the latter so far, but we need those same people to help build the systems that preserve cultural significance as the key value of this work. It’s hard to not be incredibly excited, at least, by the eruption of new smart, technologically oriented entrants to this space.
I hope you’ll join me in this new publication, Penguins and Rabbit Holes, as I highlight some of those people, as well as launch a few experiments of my own into how we can create better knowledge online. We have some pretty incredible new projects to launch over the coming months, and we’re actively looking for more people aligned to this mission to join us.
Thanks especially to Jess Sloss and the whole SeedClub community for their relentless effort to bring new people into the social token space, to my team at ResearchGate for helping build these insights into the system of science, and to the peer production team of writers and editors behind this new publication.