Decentralization Matters: But Why?
About a year ago, the Filecoin Foundation for the Decentralized Web asked if I would help edit and compile a magazine" talking about decentralization and why it's important. It was a fun and interesting challenge, and now the final product is out, the D-Web Digest.
There are a bunch of fascinating articles in there, which I summarize in my editor's letter to open up the issue. I'm not going to go over all of the articles (you can see that in the editor's letter) but I would like to highlight a few that I think are particularly worth checking out:
I think, by far, the piece in the collection that made me think the most is Danny O'Brien's piece on terminal values and cognitive liberty." You need to read the whole thing, but it really gets at the question of why decentralization is so important. I've been citing and quoting the piece to people since last Spring, but now it's finally out for people to see.
The key to Danny's piece is understanding what fundamental, core, terminal" values are so important, and whether or not decentralization is one of those, or rather a means to reach one of them. His conclusion (though you should read the process by which he gets there) is that cognitive liberty is the terminal value, and the only real way to get there is through decentralization. Here's just a snippet:
Our own consciousness cannot be rented from others, or temporarily conceded to us, with built-in police or backdoors or hidden ad men. We need to seize the means of computation, and that means ejecting all of these interlopers, and relocating it back into the personal domain we control: whether that's physically, or by using tools like encryption and zero-knowledge proofs to preserve our control when our data and processing power sits on others' hardware.
That's the pyramid of digital rights for me: a firm foundation of decentralized, user-controlled technology, giving us broader cognitive liberty, internal privacy, freedom of self-expression, and freedom of self-determination.
A second piece I want to call out is Mai Ishikawa Sutton's piece that finally addresses the whole DWeb/Web3 split that has been so frustrating to me over the past few years. Effectively, there are two camps within the decentralized tech world: one that believes it will be powered by crypto and blockchains, and one that... most certainly does not. They sometimes use the same language, and talk about the same goals, but often do not. And the two groups don't seem to trust each other much. Mai moves the debate significantly forward not by digging into what the two groups agree or disagree on, but looking into a core element of the technology (which is similar to Danny's piece above) regarding who is designing the technology, who controls it, and who stands to benefit from it. A snippet:
Building new and shinier tools out of the same political and economic conditions will do nothing to fundamentally change the world. But decentralization" is also not a value in itself, and it's not enough to build the kinds of technological networks we need to confront intensifying global crises. People need to continue to discuss, shape, and re-shape collective values. With those values held at their core, communities with common interests can collectively coordinate and embody them, first and foremost through how they govern themselves and treat one another. After all, the current internet is a reflection of shared values; the question is how we embed justice and mutual care into the Webs to come. Instead of expecting some technology to save us, we need to organize and save ourselves.
Farzaneh Badiei and Holmes Wilson both tackle important points regarding centralization being unavoidable, even in decentralized systems. These systems always have certain elements of centralization to make them work, and being aware of what these are and how they work is critical.
Badiei highlights the role of DNS, and how it has become increasingly centralized, and the potential risks that creates.
One such space is the operation of DNS resolvers, which is increasingly centralized. Imagine if we provided the capability for thousands of operators to run efficient and secure resolvers. Imagine there were so many of these operators that they were not restricted to a single jurisdiction, ensuring that any intellectual property lawsuit would not affect access to web services as a result of the resolvers being forced to block them.
Wilson, meanwhile, highlights that for all the benefits found in open source software for decentralized systems, centralization often rears its head in various cloud services, as necessitated by scale:
Cloud services such as AWS, Microsoft Azure, and Google Cloud enjoy such significant marketshare not because they control the code for technologies they deploy (tools such as Linux, Postgres, and Kubernetes are free and open source software) but because they control the ability to use it at scale. Most engineers don't even learn these tools directly anymore; they simply learn to turn to a cloud services company, purchase the correct service, and make it work for their needs. The power to operationalize code is creating the same lock-in and barriers to entry that software licenses once did.
This leads well into my own contribution to the series, which tries to highlight how to think about decentralization vs. centralization, recognizing that there are certain tradeoffs, benefits, and costs to each, but suggesting a framework for considering where each makes more sense, and how to implement them in a manner that minimizes the problems of centralization, while enabling more of the benefits of decentralization:
This lesson is important: having centralized infrastructure that is open and on which others can build in a decentralized manner can open up tremendous possibilities.
And we see that same pattern in the internet.
In some ways the internet is an even better example than the U.S. interstate highway system, because the internet did not require a huge centralized planning system to build the infrastructure, nor is the upkeep of the internet reliant on the same centralized system. Instead, it was built and created in a distributed manner, as an open system that anyone could build on, adapt, and contribute to.
As a centralized open protocol, it enabled amazing decentralized benefits. The protocol allowed anyone to build on it and experiment. And out of that grew tremendous benefits, through open innovation. A consistent, standardized protocol allowed for widespread innovation through competition, a standardized infrastructure basis on which to build, and a singular ability to communicate across the different experiments.
Out of this comes the best benefits of both centralization (efficiency, economies of scale, enabling infrastructure) and decentralization: distributed power, adaptive and rapid innovation, and the ability to be more nimble and responsive to opportunities.
The final piece I'll post a snippet of is one by Cory Doctorow, that highlights one of the biggest dangers of centralization: twiddling. That is, how centralized companies, moving down Cory's famed enshitification curve, start tweaking" the knobs to benefit themselves over their users (and partners).
There's a bitter irony in enshittification: the internet's great promise was disintermediation, but the calcified, monopolized internet of five giant websites, each filled with screenshots of the other four" is a place where intermediaries have taken over the entire supply chain.
As Douglas Ruskoff puts it, the platforms have gone meta" - rather than providing goods or services, they have devoted themselves to sitting between people who provide goods and services and people who want to consume them. It's chokepoint capitalism, a market where the intermediaries have ceased serving as facilitators and now run the show.
The double irony is how the platforms seized power: by installing so many sliders and knobs in the back-end of their services that they can twiddle away any temporary advantage that business customers, advertisers or end users take for themselves.
Again, that's not all of the pieces, but I need to leave some room for surprise. Adam Rose and Basile Simon provide very real (and extremely important) examples of where decentralized storage are hugely important. Kurt Opsahl writes about how recognizing code as speech is central to making this all work. Naomi Brockwell highlights how privacy as a core (or terminal?) value is a part of why decentralization is key. Chris Riley talks about the too frequently underexplored issue of how the data flows (not just where it's stored). And Kristin Smith looks at the regulatory landscape in which all of this plays.
Not all of the pieces in the collection agree with each other (and I certainly don't agree with everything said in them either), but taken together, they represent a really interesting exploration of decentralization, why it's so important, and what challenges and tradeoffs there are to building out a decentralized world.
As we seriously consider how to make the world more decentralized, I think this collection of articles should be considered important primary reading, even if only to get everyone thinking.