Legal scholar Paul W Kahn employs a surprisingly powerful metaphor in his book The Origins of Order when he note that the law can be seen as a system or a project. Projects have momentum, direction and are designed - we participate in them and create them. Systems are managed and often need more maintenance than creativity, they impact us and govern us - but also restrict our agency. If we see law as a project we become agents in creating law and writing rules, if we see it as a system then agency dissipates and we become recipients of legal decisions and follow rules.
This same distinction can be helpfully applied to other categories as well.
Companies, for example. Companies start out as projects, where everyone participates and builds the company together, but at a certain size, complexity and tenure they slowly, almost imperceptibly, slide into becoming systems. When we say that we want to make sure that we stay innovative, what we are really saying is that we want our company to remain a project and not shift into a slowly stagnating system.
This does not mean that projects are good and systems bad - not at all! Systems are necessary and often offer the only way to manage very complex problem spaces. Systems allow us predictability and stability to start new projects. The thing is that when we speak of a system as if it were a project, we create a dissonance that over time tends to become jarring both to the outside and inside audiences. So recognizing where you stand on that spectrum is helpful. With this shift you move from the momentum of the project to the accountability of the system - and also to what is often a critical role in society overall.
The perhaps most interesting application of this mental model, though, is the Internet overall. I think one way of understanding the tech lash and the way tech policy is developing is to analyze it as a shift from project to system. The Internet started out as a project and retained that status for a very long time - it was seen as a democratizing and equalizing force with the clear goal of creating something new - as articulated in John Perry Barlow’s bombastic declaration of independence for cyberspace in 1996:
”Governments of the Industrial World, you weary giants of flesh and steel, I come from Cyberspace, the new home of Mind. On behalf of the future, I ask you of the past to leave us alone. You are not welcome among us. You have no sovereignty where we gather.
We have no elected government, nor are we likely to have one, so I address you with no greater authority than that with which liberty itself always speaks. I declare the global social space we are building to be naturally independent of the tyrannies you seek to impose on us. You have no moral right to rule us nor do you possess any methods of enforcement we have true reason to fear.”
Barlow passed away in 2018 - but remains essential to understanding the Internet as project.
Barlow’s vision would be one of many deeply powerful visions that fuelled the growth and development of the Internet, and even if it was translated into more defensive visions with fore example Jonathan Zittrain’s and Lawrence Lessig’s work - it still had that project momentum that made all of us feel as agents.
This has changed radically, and in the last couple of years the way the Internet is described is as not just a system, but a complex, vast and manipulative system where we have lost literally all agency as users. We are being radicalized, manipulated and divided by the Internet. The Internet is polarizing and generates inequality.
The Internet went from being a project to becoming this faceless and opaque system in public discourse surprisingly fast. Why is that? What happened that made us all feel not as participating in Internet as a project, but rather being victims of the Internet as a system?
One answer is that the Internet project succeeded. We have global commerce, speech has been democratized in the sense that everyone can speak (and it still has good effects, but they are obscured by the focus on the state of our public sphere - a problem that has more to do with how we become citizens than ad technology) and we can now connect effortlessly with a global marketplace of both commerce and ideas.
But as Edward Luttwak warns us: it is in victory we craft our own defeat. In his seminal Strategy: The Logic of War and Peace Luttwak argues that the most common reason for defeat is overextension - that a victory is pushed beyond its culmination point and when that happens it paradoxically turns into defeat. It is easy to see how this applies to the Internet: the claim of victory and the overvaluing of the change in the Arab Spring was a part of that - and there are many other examples. The Internet did not on its own democratize the world, but could be used for surveillance - the claims that were made overextended the project to the point where it collapsed into a system.
And yes, for the Internet to already have collapsed into a system is a kind of defeat, I think. There is so much left to do, and so many things that the Internet can still improve that if we start to lose our sense of agency in the development of the Internet we really are squandering one of the most remarkable opportunities available to us to improve the human condition.
Can we turn back? Turn the Internet into a project once more? Probably not, at least not the entire Internet - but we can drive projects to improve the public sphere, to build new institutions for shared cognition and common baseline of facts. We can build projects on top of the system to ensure that we move further with the potnetial that actually comes with technological progress, and now seems to be largely forgotten.
That also requires accepting and operating within the conceptual framework of a system - building accountability and transparency, designing participation and openness. The natural evolution of a system is into complexity and opacity over time - and that kind of entropy will not fix itself. Where a project needs purpose and vision, a system needs governance.
*
In other news, the US election seems to be accelerating a tension in the intermediary liability debate that is very interesting — the question of where in the stack, and on what grounds, moderation needs to happen. The reactions to the shutting down social media accounts for president Trump and the subsequent actions taken against Parler are still unfolding - but just looking at what has happened so far suggests that we should expect moderation efforts to go deeper in the stack, to use the model suggested by Stratechery. This in turn gives rise to a counter-reaction that argues more or less for a ”must-carry” solution, limiting moderation to essentially determining obvious illegality. This tension between more responsibility lower in the stack, and a reaction to that pushing for narrowing the space for moderation radically has always been present in the debate, but it is now really coming to the fore. We have seen it play out in Europe between member states, but the decisions taken in the last couple of days are likely to echo in the debate for a long time.
One solution that has existed at the edges as a fringe idea, and that may be interesting to think through, is that of a structural or functional separation of the moderation function and the stack. Stratechery’s model assumes that moderation happens in the stack, but there is another possibility: that moderation decisions are essentially separated from the stack entirely and made in a separate institution - and then implemented at the level of the stack depending on the severity of the decision made.
Moderating the Stack
This idea seems far-fetched now, but with Facebooks recent moves into a content council and some of the academic writings here it is an idea worth having in mind for understanding how that would work. The downsides seem, frankly, to be mostly organizational and around efficiency. It is hard to argue that there is any competitive advantage in the formulation and application of content moderation rules when all of the companies essentially follow each-other, and the appetite for competing on free expression is close to zero. With Parler being isolated the competitive strategy here is also cut short - if your content moderation practices are radically different from those employed in the industry, you will suffer the consequences. (Though competition on content moderation should be distinguished from competition on encryption which is a more tricky thing to get into that analytical schema).
A good overview here on some of the models and thoughts around content moderation. Modelling functional separation here is an interesting exercise left to the reader.
*
Finally, then a few reads from the blog this week and a literature recommendation. On the blog this week:
Writing the History of This Moment. A short note on why archiving what has happened is key to learning from it, and how bit rot affects democratic evolution.
Using Skill Trees to Understand Organizations. Following my obsession with capabilities as a mental model for team development this short post outlines the use of skill trees with an example.
In terms of reading this week I strongly recommend re-reading Postman. Especially Amusing Ourselves to Death:
“I find it useful to think of the situation in this way: Changes in the symbolic environment are like changes in the natural environment; they are both gradual and additive at first, and then, all at once, a critical mass is achieved, as the physicists say. A river that has slowly been polluted suddenly becomes toxic; most of the fish perish; swimming becomes a danger to health. But even then, the river may look the same and one may still take a boat ride on it. In other words, even when life has been taken from it, the river does not disappear, nor do all of its uses, but its value has been seriously diminished and its degraded condition will have harmful effects throughout the landscape. It is this way with our symbolic environment. We have reached, I believe, a critical mass in that electronic media have decisively and irreversibly changed the character of our symbolic environment. We are now a culture whose information, ideas and epistemology are given form by television, not by the printed word.”
Postman’s analysis on TV feels curiously dated now, but his framework - borrowed from Ernst Cassirer - based on a symbolic environment is still interesting. Instead of Postman’s patterns of decline, though, we may want to try to understand these patterns as changing and evolving in ways that are straining the connection between our memetics and our biology. For more on that make sure to read the Julia Galef book referenced in the blogpost about writing history!
Thanks for reading, and help spread the word. If you know anyone else who should get this newsletter send me a note at nicklas.berildlundblad@gmail.com - I am still screen by invitation, but anyone on here already should feel they can openly recommend the note to others, of course!
Nicklas