March 26, 2006

The Internets of Things

People are making tiny computers and sticking them in everything, this much is pretty clear. So is the fact these things are starting to talk to each other. Just what to call these everythings though is a whole other story, but from the rather sub-academic linguistic mess comes one compelling figure of phrase, the internet of things. And well this is the internet and this is a blog, so let me just do my duty and add my own iteration to the word slop and suggest we need to get a bit more Dubya Bush on that phrase and start talking instead about the the internets of things.

The world may have laughed painfully when George Bush started talking about the internets, but the very fact that we can talk about the internet in the singular is a rather remarkable historical fact. A fact of protocol. Whether or not that the wireless objects increasingly percolating around us will all talk the same language, speak the same protocol, is in geek terminology a non-trivial problem. That is to say it is a resolutely political issue, and one with ramifications that could well effect us all.

Who writes the protocol(s) for the internet(s) of things. I first asked that question over in the comments on Adam Greenfield's An quite accurate and rather dismissive response came in from internet of things coiner Bruce Sterling: "Geeks are writing the protocol. Geeks are the class of people who write protocols." Which is sort of like saying "monks are the people who write books" back in 1300. Completely accurate and completely irrelevant to the issues of the future. We quite likely are a generation and change away from a world where being able to write code is as second grade as being able read and write. Now of course writing protocol is a rather special class of programing, the same way say writing novels or legal statues is a specialized class of writing. But that just begs the question, are the protocol(s) of the internet(s) or things getting written by the lawyers or novelists? And if it's both and more as it may well be well you can kiss that internet singular goodbye.

The unity of the internet comes two protocols embedded in the five letters of TCP/IP. We'll save you the details and histories and just say that that birthing process occurred in a rarefied embryonic stage of networked cultured. So rarefied that a surprising percentage of the main players went to the same California high school. More importantly perhaps though, it was only through luck and a naive disinterest that institutions like AT&T, Xerox and the US Department of Defense never sunk their claws into the processes going on in their nether regions. You can bet that the NSAs, Sonys, Microsofts, Nokias and Deutsche Telekoms of the now are not going to let similar opportunities to mar the protocols of tomorrow slip past unmolested.

Yes the geeks are writing the protocol today, but which geeks? The Chinese government geeks? the Redmond option whore geeks? Russian mafia geeks? American dwarf linux geeks? Sony DRM geeks? favela internet cafe geeks? If its an internet of things than only one can really win. If multiplication prevails we have not an internet, but internets of things. And the difference between the two is quite literally of Tower of Babel proportions. Are these things in this together or do they take sides in the networks around us?

Posted by Abe at March 26, 2006 06:41 PM


It appears we are indeed headed towards a multiple Internet - an "internets" - model.

DRM efforts alone - designed with the ambition of turning all sound and image manipulating devices into copyright enforcing rental properties - will create a force field protected zone of digital communication existing within the global Internet.

There are other examples.

Now that I think of it, I believe Sterling's "Snow Crash" depicts - perhaps accidentally - just such a paradoxically connected world of disconnects.

Snow Crash was written by Neal Stephenson, but it took Bruce Sterling to make it a Gizmo.

Wasn't the "inter" the whole point of IP in the first place? There were already networks, the Internet was originally the InterNetwork, the network of networks.

We already have networks of things, similar to the military and educational networks that were linked by the internet. I think it's implicit in Greenfield and Sterling that we need one protocol for those networks of things to communicate with - an internet of things.

But now that the Internet itself is on the verge of fragmenting, it does seem less likely that any new internetwork protocol will become standard.

It's a really good point on the "inter" but I've never seen any indication that anyone writing about the "internet of things" has particularly taken it into account. Certainly from that little exchange on v-2 it seemed pretty clear that Sterling never took it into account at all. Adam does write about protocol a bit in _Everywear_ but he never makes any case for how there might be some unification of protocol occurring. And that's precisely the point of this post.

Not so sure about the internet fragmenting bit. Any fragmentation that occurs seems to be happening above the level of TCP/IP (and UDP/IP). I'm sure there are some minor exceptions (maybe some of the on demand digital cable services), but when you are talking about wired two way communications IP is pretty damn ubiquitous, and unless you want to run your own wires you pretty much need to run on top of it. Wireless devices just don't have that limitation. Some company in China can build a new device with it's own protocol and if it sells there is a new network right there. You and me can make our own device and if it sell boom we have our own network. No inter so yeah maybe it's just a network, but the internets of things works better in the PR machine...

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)