Against Decentralization

In the free software/open web community, the notion that the web should be decentralized is more than a shared ideal, it is a piece of dogma.

It is key to all the characteristics of the web that we are most proud of: diversity, innovation, the competition of ideas rather than bank accounts. There are plenty of great reasons to strive for decentralization. Zittrain's The Future of the Internet and van Schewick's Internet Architecture and Innovation make brilliant arguments in this regard.

But it's obviously also a troubled model, and the majority of the web is moving towards more centralized services. Facebook, Google, Apple, Amazon or even GitHub and Slack. Most open web advocates stand in opposition to this. They opt to build small, open source alternative solutions that anyone can run on their own server.

If you squint really hard, these things kinda, sorta look like the real thing. But, fundamentally, this is the Ersatz web for the nerdy elite. Even worse, it puts open web developers - in a very literal sense - in a conservative, rather than in a progressive position.

Ad-based data centralization models and cloud services might not be the most sustainable thing on the planet, but for the moment they simply beat the shit out of the open web. Our formulaic response is to claim that this is just a "UX problem". This argument is in total, blissful disregard of the economics involved.

Making decentralization the only strategy for the open web is a big bet, with little evidence in it's favor. What are the alternatives?

Figuring out how to get better at making open centralized services is one. Pushing for innovative government regulation of monopolistic platforms is another.

Germany invests 7.6bn Euro a year into publicly financed news and entertainment media; to make sure that voters are well-informed and have access to diverse sources of information. While there's already the odd iPhone app and web site coming out of this, it's time to fundamentally renegotiate this system for the digital age. Using this money, we can create public spaces on the web - free software, open knowledge, digital public goods.

Even a percentage of this fund (and it's UK sibling) would provide enough resource for Europe to get out of the unhealthy pattern of trying to emulate Silicon Valley. Instead, we can create an alternative model. One in which the web is more than a way to connect ads and eyeballs.

The main benefit of decentralization over such alternatives is that it minimizes the extent to which we have to engage with real-world politics and the economics of operating large-scale web services. We are deeply infatuated with the notion of solving political problems through technology.

The poster child of this, of course, is Bitcoin and the blockchain. While ideas like smart contracts are genuinely innovative and interesting, the naive dream of digital currencies remains that they will catapult us into libertarian nirvana, a world in which all politics is protocol.

Many smart political choices are embedded in the core of internet architecture. But there's an equal number of risks - network effects, data silos - that technology just wasn't able to avoid. We shouldn't be stuck re-fighting those lost battles, but instead commit to using political process to address them.

This means working with government. You know, the thing our buddy Barlow said didn't exist in cyberspace but that somehow still got full take on our web traffic. It's no longer cool to engage with government only when they're about to screw things up - net censorship, surveillance, copyright policy - and to pretend that all political issues on the web can be solved given enough distributed hash tables.

So, what's the alternative to decentralization? Laws. Public service websites. Creating commons infrastructure. But let's just not pretend any more that the future of the web is all about better UX for our IRC clients.