Decentralization is a pretty vague buzzword. Vitalik considered its meaning a year ago. In my estimation, it can mean a couple of things:

  1. Abstract principle when analyzing general power structures of any kind: "Political decentralization" means spreading political power among differing entities. "Market decentralization" refers to outcomes being produced without being coordinated by a central authority. It's a philosophical idea that can be interpreted broadly in a lot of different contexts.
  2. Bitcoin, mostly. Lots of credit for the buzzword's current popularity traces back to cryptocurrencies and blockchains, and I think the term "decentralization" without context is rightfully claimed by the yescoiners and defer to Vitalik's interpretation for its meaning. I call this "financial decentralization" in contexts where my definition is dominant.
  3. A second, specific implementation of (1) that I want to talk about.

I'll admit it -- I've been riding the buzzword. Lots of people often say "decentralization" and conflate various ideas with it. The umbrella term for the thing I want to talk about is called web decentralization, and it refers to a specific set of ideas and initiatives to distribute technical power in the social networks that run our digital lives.

Fundamentally, the problem with the web ecosystem is that consumer choice is limited. Facebook, Twitter, Google, and other tech giants "own" a large part of the social graph that both powers the core digital connection goodness and sustains the momentum that they will keep owning it, due to something called Metcalfe's law. If you want to connect to people on the internet, you have to play by their rules.

What does it mean to "own" a social graph? Largely, it happens automatically. Some people get a Facebook in 2004, others join in 2007, and finally in 2012 your granddad. They built a superior product to others at the time (RIP, Myspace). But Facebook is now worth 875,400% what Myspace was when it sold. Could a new Facebook survive in 2018 and unseat the incumbent? That seems impossible. The Metcalfe momentum has become a gravity black hole. A small brigade of the most competent engineers around the world are employed to psychologically optimize how much time you spend on site. Direct competitors don't stand a chance.

The end result is that social innovation is dead. When is the last time you saw a significant change to how your Facebook or Twitter timelines worked? There's infinite potential to improve your experience on both platforms. What if you had a version of Facebook centered around your family & genealogy? You go to Fambook.com, and it's just warm updates and photos and memories from people you love, extended family you barely know, and your data was backed up in your whizbang cousin's basement.

Facebook would buy it immediately. It would redirect to family.facebook.com, they would mine your interactions and pictures to power god-knows-what AI systems, and if 20 years from now whoever Facebook decides to impede your access to historical family photos...

We've seen this a million times with social platforms. Anything with significant traction gets gobbled up.

On the other end, these platforms are facing vastly increased scrutiny from the rest of society; political forces concerned with foreign electoral manipulation, or social disharmony caused by algorithm-powered alternate truth universes. Every decision about how news, research, and political discussion happens online is directed entirely by the business models that power distribution channels like Facebook, reflected in the priorities of publishers who make content. Since publishers are rewarded by maximizing any emotion fruitful in generating clicks, negative results like outrage spirals and social discord are inevitable.

There are technical solutions to these problems. These solutions are also huge opportunities for a next generation protocols and applications that restore digital power back to individual engineers and users.

A "web decentralized" system looks like thus. You start with bare-bones replicas of social networking, publishing, microblogging, and chatting. You build a small social graph of your friends. This time, the data structures powering these applications live on your computer and are in a format you can easily grok and extend (Sorry, normies, it will be engineers-only for the next year or two).

Now you start building. I've always wanted a social network powered by movie, book, and music recommendations. I'll have a personal "library" of media I've watched/read/listened to, and a wishlist of things I'm excited about (and upcoming releases). I log into Bookbook.com, and the UI is tailored to me. The main feed is a list of progress updates from books that close friends are reading. Their Kindle automatically sends updates to their personal website as they read. On the sidebar is my top To Read list, which I built using a generic algorithm and some custom extra weighting on people whose recommendations I value highly. Click through to my profile and the prominent element is "Books that have made me who I am", a UI I designed entirely custom showcasing my intellectual genealogy and my top favorite authors I want to pay tribute to (Michael Crichton, Albert Camus, and Stephen King, if you're curious).

I actually want a lot more of my life online than is currently possible. But I don't want to shoehorn it all into Facebook's already very crowded UI (have you clicked through all your Facebook subpages lately? it's... a lot). And every time I see a similar service or want to build one myself, I'm hit with the dreaded wall of "are people actually going to use this?", or "am I entering in all this data for nothing?", or "are corporations going to spy on my deeply personal belongings?".

The solution is technological standardization. Individuals, mostly engineers, need to expend a lot more effort contributing to the protocols and processes that drive inter-application communication. Your core Facebook identity -- your username, your connections, your chat history -- should be a universally standardized protocol with a Democracy-scale process for updating and extending it. Crucially, that process needs to be directed outside the direct control of tech companies, who are capitalistically bound to monopolize and direct control back to their domains.

Users can help as well. The open source world also doubles as an army of free support engineers. If your decentralized-powered application bakes in ways for individual users to help contribute feedback, bug reports, user research, you've unlocked even more quality power, bringing the cost of maintaining the system down by orders of magnitude, leveling competition with the Facebooks.

The crazy thing about all of this is that it's all possible without huge companies getting in the way. Individual programmers and maintainers need funding mechanisms, but a lot of this can be provided through crowdfunding, institutional support, and direct donations. This is where the core work needs to be done in the web decentralization movement, and I will talk about this at length in future blogposts.

There are many avenues of action to enact the desired outcomes of web decentralization. Tech leaders could simply become more benevolent; Mark Zuckerberg has pledged his 2018 to thinking about decentralization and Facebook's role in society. I sincerely welcome Mark's effort and reflections. I would love to trust Facebook and use it as my digital home, but there's a long way to go to match my vision.

Increasing talk of federal regulation of Facebook and other tech giants has also become a popular topic of late. If this avenue seems appealing, I highly recommend watching the congressional hearings last November with representatives from Facebook, Twitter, and Google. In my opinion, the inexperience and bewilderment of the average congressman to digital affairs was palpable. An effective government could in theory be a good digital regulator, but I have a strong lack of faith in the US government in particular to craft regulation with end user values in heart.

Ultimately, decentralization is about shaping the the balance of power in digital domains. I for one would not like to wait around while the Tech overlords and Crusty regulators decide what happens with our digital lives. There's no reason for us to keep listening to either of them. A handful of dedicated engineers, designers, a organizers could implement the alternative today. And that's what web decentralization is all about.