When Vinton Cerf engineered the internet, he devised a system with no core: computers connected to computers over a vast web. No one computer is that much more important than another. Internet traffic–and the various protocols of messages (emails, web requests, streaming video, and more)–all bounce around the internet in a beautiful, decentralized way. Emails from one email server are sent to another email server. Videos are sent from streaming servers all over, distributing the load. Even text messages bounce from our servers to carriers, and then out the air waves to you.
Why is this so beautiful, you ask? Why are you writing about this? It’s beautiful for two reasons: scalability and democracy.
Let’s first address scalability. The internet has no “single point of failure”. With no node being any more important than any other node, if you remove a node (or blow one up in the course of a war), nothing much bad happens. Sure, a few messages might not be delivered, but the internet lives on. This is why the internet will never “go down”. It can’t. It can really only get better and stronger, with more and more redundant connections, increasing its speed and durability.
Now, let’s talk democracy. The internet–in all its beauty–is a very democratic thing. As messages are distributed across this vast web, it becomes difficult (though not impossible) to control the messages being sent and received. The owners of each of these hundreds of thousands of email servers (or IM servers, or video streamers, etc.) have control over the messages stored. Each one would have to work in harmony with the others to control the flow of information in a meaningful way, which is very difficult to do. The internet, in many ways, keeps us honest, and exemplifies the nature of the United States economy: decentralized and free.
Recently, however, there have been some interesting shifts. The biggest of which is Twitter. Many people seem to think of Twitter as another standard protocol, like email or FTP. But with a standard protocol, there is no real “owner”. Programmers write servers that send email to other email servers, and the protocol serves as…well, just that: a standard, common language for interaction between the machines. With Twitter, however, there is only one Twitter. There’s one monstrous computer system, called “Twitter”, run by a bunch of people at a company called “Twitter”, in a building (or a series of buildings) leased by Twitter and it’s investors. You won’t find a computer system called “Email”, run by a company called “Email”. You can’t “blow up” email. But could you “blow up” Twitter? Sure. Twitter is a giant step back to a time of centralized systems, single points of failure, and massive (and I mean incredibly massive) amounts of control over the information it stores.
“Massive control over information?” you ask. “Really, David? Get back to writing about mobile technology and stop this nonsense about Twitter. “
Ok, fine. I will. I promise to get back to my normal rants, but please just hear me out on this off-topic subject for just a little longer. I’m almost done, I promise.
When you go to Twitter.com, you’re immediately confronted by a search page. Unlike Google, or Yahoo, or Bing, or whatever, with Twitter, you’re searching…. TWITTER! You’re searching its own, stored, centralized database of “tweets”. On Google, you’re not searching Google! You’re searching the vast web of decentralized information known as the internet.
I know Twitter is not an evil organization. And I’m sure their founders play ping-pong in the rec room like every other Silicon Valley start up. But if it were an evil organization, it wouldn’t have to do much to shape the view of the world around it: just do a “search/replace” on the billions of tweets it stores, and presto: new news. All the news and information in the hands of a few leads to opportunities for control over the democracy.
A more likely scenario is the following: massive (and again, we’re talking truly, truly massive) amounts of data, piling onto servers at an exponential rate, eventually bringing the Twitter servers to their knees, blowing out Twitter entirely. Buh-bye, data!
Or, an even more likely scenario: massive (and again, we’re talking truly, truly, TRULY massive) amounts of data, costing an even more massive fortune to maintain in server farms. As there’s no real revenue model for Twitter. (Have you ever seen an ad on Twitter or your “Tweet Deck”?) Twitter eventually goes bankrupt. So long, tweets!
So, I’m sure by now you’re thinking I’ve finally lost it. David’s totally paranoid and has some crazy anti-Twitter paranoia.
But the real scare here isn’t Twitter. It’s the trend and precedence Twitter sets. It’s the change in course from distributed to centralized, from egalitarian to monarchy, from unbreakable to a single point of failure. Is this the correct direction for so-called “social networking”? There really doesn’t seem to be anything “social” or “networking” about it.
Of course, Twitter is not alone. Facebook is just as bad from a single point of failure or control perspective. If Facebook were to go down (which is less likely due to the fact there’s a fantastic, ad-supported revenue model associated with it), we’d lose massive amounts of information as well; many would say more valuable information, such as family photos. However, the difference is that Twitter portrays itself as a protocol, and Twitter portrays itself as a “search engine” for “tweets”.
I believe the paradigm must shift again, as we look for ways to create truly distributed social networks. New actual protocols will need to be developed, allowing the individual to host their social networking server and connect it to other social networking servers over the web. Text message “un-tweets” will be attached to photos. Photos will be attached to profiles. Profiles will be attached to blog posts like this one. And each one will be stored across the internet in a truly distributed, democratic, and unbreakable way…just like Vinton intended.
