Decentralization is the process of ensuring no ownership or control is limited to a single entity. The issue of decentralization is important in politics, economics and technology to limit the amount of impact a single, centralized body can have over the ecosystem.
In politics, we see this in the form of federal governments, where control of execution is handled down to the lowest governance levels. This distributes the responsibility and decision rights from autocratic or pseudo-democratic agents to the citizens. Important tasks are divided and spread across roles and regions to avoid single point of failure or corruption, in case those in authoritative position fail to do their tasks. This may not have worked out ideally so far across the world, but this gives maximum choice to citizens to express their concerns and take action.
In economics, the concept of decentralization is broad, but it boils down to efforts for overall economic prosperity by ensuring that no single entity controls the flow of money, distributing economic policy control and distributing flow and creation of currency.
In technology, this means creating a system with no centralized control. Decentralization in technology can promote open access to information, provide improved privacy and create more resilient systems.
The internet is our backbone today for creating decentralized systems. Internally, it is built through multiple layers of technology — at the lowest level we have hardware cables that relay signals across the globe, then we have protocols that convert these signals to data packets that our computers can understand, then we have software that we use to create, serve and consume those data, like web servers and web browsers. At its core, the internet as technology embodies all the principles that builds successful decentralized systems.
The “World Wide Web”, the one that we fondly call “the web”, is built on top of this internet by utilizing servers that serve data, a globally distributed domain name system (DNS) that translates names to IP addresses and web browsers that can convert the data received from those servers and render them into interfaces for human interaction. The power of the web comes from the ability to link to other web resources that do not reside on the same server, thus creating a globally traversable pool of information. So, all the information in the world does not need to remain on one huge mega-computer. It can be decentralized; spread out across millions of computer across the globe that form part of the network and exchange information with each other.
Regression to centralization
The technology stack of the internet actually consists of multiple eventually-centralized processes.
Like, the DNS works by having a record of domain names that map to IP addresses. We can host our own DNS servers, but on public networks and the internet we use daily, we are dependent on centralized name servers that are called “root name servers”. DNS queries follow a chain of name resolution up to the root name servers and try to resolve a domain name to its IP address. Even though this removes the requirement of every computer having to remember every other computers’ IP, the root name servers still form a centralized part of the domain name resolution service.
Also, as we take our internet subscriptions from internet service providers (ISPs), all of our internet traffic passes through their hardware through to the undersea cables. This means, if a government or any agency for that matter can control the ISPs or damage the undersea cables, they can take down the internet. We see these happening in the form of governments asking ISPs to restrict access to certain websites or by trying to censor the internet usage of the entire country or by banning internet completely.
The trouble does not end there. If we move one level higher, we see the internet traffic today is becoming increasingly homogeneous. Five or six corporations account for the majority of internet traffic today. This has happened over the years by creating services which users of the internet have learned to depend on, be it their emails or documents or videos or chatting with friends. The truth is, these corporations do offer services of significant value. But, at the same time, they subvert the decentralized nature of the internet by either amassing user data thereby establishing eventual dependency or by trying to be the gateway to the internet or, worse, by pretending to be the internet itself.
And then, we have the “cloud”. Cutting across the various (mis)interpretations, a “cloud” is simply another computer always connected to the internet where you can store your data or offload your computations. The cloud has proven to be an effective measure to silo users and create centralized services where an entity can control what happens with the data. Once again, it is not about whether this is “good” or “evil”. It is about whether it is a resilient solution that pays off in the long term. Once a data is in a cloud, it is dependent on the resiliency of the systems in the cloud to ensure no data leaks happen or no data is corrupted. Even if we assume that such a cloud is safe enough, we run the risk of the cloud system wiping our data and terminating our account or leaking out data to surveillance agents under pressure from governments. In short, it is not a good enough solution, yet it has met with huge success.
It seems that in the process of making successful businesses around the internet, we have made a regression to centralization and almost removed the idea of an open playground.
Bringing back decentralization of the internet
There has been multiple commendable efforts to promote decentralization of the internet. There are two major obstacles, however:
- Technological - It takes a lot of research and efforts to build such systems. I’m talking about decades worth of effort.
- Awareness - People have become too comfortable with having one account that gives them all facilities from emails to document storage to chat. It is hard to ask people to change, unless you have a better deal.
This is why the people behind MaidSafe have been researching for over nine years to create a better version of the internet, one that is decentralized, secure and privacy-aware. Then, there are the minds behind Matrix who have been working on creating a standard for secure, decentralized communication. There are joint group efforts like Redecentralize who, as part of advocating for a decentralized internet, maintain a directory of projects working on decentralization. There are crypto-currencies like Bitcoin, Safecoin and others. This is not an exhaustive list, but an indication that the technological challenge has not scared people away. Those who believe in that idea, have eventually come forward and taken up the hard work. These are the people who in turn motivate my work on projects like unicorn.
We may be close in achieving the technological goals, yet the awareness part is still a big problem. We need to focus on building an ecosystem of technologies that talk to each other without as much centralization as possible. From a user’s perspective, those things need to just work. Otherwise, there is little incentive for a user to even give it a try. Also, there is the part about advocacy and letting people know that alternatives exist which can be way better than the technology people use today. This awareness needs to be built from the ground-up. We cannot just say that “cloud services are evil”. We need to show people real advantages; we need to show people that they do not have to exchange comfort, usability and performance for privacy and security.
We need to make decentralized systems work.
Suggested readings
- The mission to decentralize the internet - Joshua Kopstein
- The future of the open internet is decentralized - Joseph Cox
- A Declaration of the Independence of Cyberspace - John Perry Barlow
- The future computer utility - Paul Baran
- Why decentralize power in a democracy? - Larry Diamond
- What economic structure for socialism? - David M. Kotz