« Ray Ozzie's Clipboard for the Web (ETech 2006) | Main | Dick Hardt on Identity 2.0 (ETech 2006) »
Bruce Sterling on the Internet of Things (ETech 2006)
The evening keynote (last night) was by Bruce Sterling on the Internet of Things. This was one of those talks that is impossible to blog. Even a word-for-word transcript wouldn’t do it justice because Bruce’s delivery is as much a part of the content as what he says. I’m sure it will be on IT Conversations soon and I encourage you to listen to it there.
Bruce’s message was about language and the power of naming. He said, that when it comes to remote technology, you don’t want to freeze your language too early. It limits the ability of people to find the intrinsic value. Artificial intelligence is an example. Computers are not intelligence. They boxes of circuitry. They just sit there “ordinating” as the French say.
Bruce gave a great example about whether you’d want to have Google or Alan Turing in a Box (a reference to the famous Turing test). It brings into stark contrast the difference between AI and what we’ve built without AI. He makes the claim that a tech world that talked about computers as ordinators or sorters or card shufflers probably would have invented Google in the 1980’s. The new stuff, people interacting with technology, is what’s important. On demand social application are unimagined by Computer Science.
The idea behind the Internet of Things is laid out in Bruce’s book Shaping Things. He envisions the Internet of Things doing its work in six interesting ways:
- interactive chips
- local and precise geolocative systems that sort out where you and things are
- powerful search engines
- sustainability through auto recycling
- 3D virtual models of physical objects
- rapid prototyping of object
Bruce calls these Internet things “spimes” because they’re tractable in space and time. Manufactured objects that are material instantiations of an immaterial system. They are virtual object first and actual object second. The primary advantage is that you no longer inventory your possessions inside your own head. That work is done far below your notice by a host of machines. You just ask where things are and where you bought them and what you paid, and so on. You no longer wonder where your shoes are, you just google them.
Why create a new word? Because it creates a cloud of associated commentary and data online. This is an example of what Tara McPherson calls a theory object. Theory objects accrete attention and generate trackable trails of attention. Any real theory object has trackbacks and pictures and a FAQ and so on.
Bruce isn’t talking about some smart object with a processor inside, he’s talking about everyday objects that have unique digital identity. When they work well, new words like spime are like good brooms that get rid of old words. This is different than hype. Hype is a system call on your attention. Hype is aimed at your wallet.
Bruce continues with a riff about hype in technology and politics. In technology, the opposite of hype isn’t the truth (like it is in politics), it’s argo—technojargon. Techjargon is a superspecialized geek cult language. It’s deliberately hermetic. But, a small clique doesn’t have enough people to successfully name things. It takes a whole lot of people to manage a popular language.
Bruce doesn’t think that ubiquitous computation will be either ubiquitous or computation. It will be patchy and limited like cell phones. It’s not a smooth supported product. As we move in that direction, you don’t want to avoid the contentions of the literary struggles. The words are signifiers of the places where things aren’t yet hammered out.
Let’s take “Web 2.0” He quote’s Tim’s definition. Bruce says that this definition is not a clean, sanitary things, rather it’s about people Tim thinks are important and the causes and values they hold as a class. He then quotes Alan Liu on the idea of Web 2.0 who is highly “skeptical.” There’s too much of the “high priesthood” of administrators who work behind the scenes to make it work.
Bruce goes on about this because it matters. Naming can make completely new concepts and make them everyday. That’s the goal.
The presentation was very entertaining (which is saying something given that it was starting at 9pm). Of course, with Bruce Sterling, you’d expect nothing less. Like I said earlier, you’re going to have to listen to it to get the fine nuance. I’ve ordered a copy of his book because I’m interested in the identity aspects of what he calls the Internet of Things.
Posted by windley on March 7, 2006 10:31 AM
Comment from Simon Wardley at March 9, 2006 2:53 PM
Fantastic overview, thanks. I've a personal viewpoint to add.
Bruce’s general thesis is on tagging and sorting of physical objects and how these objects interact with the digital world.
The potential of the physical / digital collide goes far beyond tagging as Bruce hinted at, and his creation of a new name - spimes - for these future nodes of collisions between the physical / digital world means at least we can talk about this.
Even if as Bruce says, the name may not be right - it’s certainly good enough.
That objects can and will become combined with metadata, as Bruce proposes, seems likely, however objects themselves and their properties can be defined by instructions sets (the process of CAD diagram to physical object through “printing”) and interact with metadata or physical interaction to change physical function or form (as per concepts of origami electronics and shape memory etc).
That software compiles to an instruction set that runs on hardware which can be defined by another instruction set compiled from CAD diagrams and that both instructions set interact to create the required result, means there are two ways of achieving this.
Traditionally the easier way is to change the software, however as printing electronics and objects becomes more advanced and common this may change.
To complicate matters, the concepts of origami electronics and other such ideas is that the physical functions on a object may change with interaction with the physical world. An “electronic” sheet of paper which folded one way act as a TV set, folded another way becomes a calculator and if you tear of a piece of the paper you have an alarm clock.
Such interactions could be from metadata, and are not limited to just user interaction (tearing, folding) but could include shape memory type affects (where the object changes shape itself due to some other interaction).
A jacket that changes it’s properties according to the weather or fluoresces bright colours when you are trying to find it in the mountain of coats left at a party, or bright red when someone else takes it - calling your phone at the same time and telling you where it is and who’s got it.
The jacket, could have been one designed by yourself - printed by you, with the customisation you want and the interactions with meta data and other physical events you require.
Tagging and sorting? Yes to begin with, but the collision between the physical and digital extends far beyond this.
Do developers of the future need to worry about writing in software and CAD diagrams at the same time or do we have new languages some form of “Spime” Script which is compiled into both physical and digital instruction sets, such decisions of where the boundary lies being made by the compiler?
These are the things we can start to discuss and understand, for the last ten years I have talked with many others on this potential change, often finding it incredibly difficult to communicate the impact of what was going on and how this would change.
It wasn’t until Bruce’s talk that I realised how important a name was, and what an obstacle not having a name had become or that using older names such as objects was.
Spimes it is and now we can talk about it.