There's a new Bruce Willis sci-fi/suspense movie coming out called Surrogates and the trailer looks like it could be very cool (and could also suck badly). The premise: in the near future, everyone remotely operates telepresence robots (surrogates), feeling everything they feel, seeing everything they see, and experiencing life without the risks of having your actual body out in the world. Plus you get to have the body you want, not the one you're stuck with. Sign me up!
There's certainly some similarities to Kiln People, but the differences are more telling: that you can only have one surrogate, and it offers no intelligence of its own, it's just a remote control.
But the surrogates serve as a perfect example of the kind of bad system design that comes from using backwards compatability as a means of advancement. No, really, bear with me. In the world of the movie, essentially everyone is always only interacting with the world through surrogates: complex, expensive, mechanical creations that go to a lot of effort to emulate an obsolete factor, the human presence. Take a step back and consider how you'd design that world, and you'll quickly realize that clunky, error-prone robots which only serve to fool other robots are an unnecessary complication. The whole world could be done far more flexibly, far more reliably, and far less expensively, as a virtual world. After all, the only bit you really need for that is the stuff that lets you experience the robot's sensations and control it; just hook that to software only, not to hardware, and you're already there. The robots add nothing apart from a complex, expensive, unreliable component.
And yet, the future of Surrogates is far more likely than one in which a virtual world, a cyberspace, is really a replacement for real-world interaction. Why is that? The simple reason that we can get from today's world to the world of Surrogates by a series of steps each of which can be done by individual people, companies, groups, etc. That's because each step is completely backwards-compatible with the world before it. That's why a million robots spend most of their time pretending to be people in front of other robots: they're all acting out long-obsolete backwards compatability.
But a move to cyberspace requires a single standard to emerge from the competing ones and gather enough momentum to gather the entire world, and to keep doing so despite challenges from new competitors. And everyone has to decide, when they get involved in one, which one to invest their time, money, and effort in.
The world of technology is constantly dogged by this kind of inefficiency that would disappear if we could agree on standards. In fact, this is one of the many flaws in the idea that the free market solves all problems, one that wasn't as clear in the time of Mills or even Keyes as it is now, when compatability is a much more important part of the things we do than it has ever been.