Stephenson’s process of creating his future world reminds me a bit of how I used to think when I was very little. I played with my building toys, and one time I figured out in my head how a perpetual motion machine could be made (but then I forgot how it had been done, and could never recall how exactly the device was hooked up). I thought of a system for transporting objects to any place in the house, using a network of ceiling fans, where the blades of each fan overlapped the blades of the fans next to it so an object could be shifted from one fan to another as necessary and easily be transported anywhere. I never thought much about the exact mechanics of all this (such as how the objects would be shifted), but it was an image I thought about a few times. Stephenson’s way of thinking reminds me of how I used to think, back then, and what I could have done if I had had thousands of little electric motors instead of just one, and hundreds of thousands of little rubber belts and pulleys, and much more complicated and capable little plastic interlocking pieces all spread out in a vast array. But sadly I didn’t have all this, and so I outgrew that way of thinking.

The book’s vehicle is a society in which nanotechnology is fully realized and just about any object can be manufactured freely in the home, in devices I pictured as looking like glorified microwave ovens called “matter compilers”. A matter compiler requires a “feed”, however, which is like a gas line, except that it carries little cubes and tetrahedrons (well, that’s how I imagined them at least) of “raw matter” to feed the microwave oven. Here’s where you have to stop thinking about the logic behind the story too much and just accept things as they are, because otherwise you might be tempted to start asking a whole bunch of questions: (1) Why couldn’t companies just nano-manufacture goods centrally and distribute them through normal retail channels? Why is there such a benefit to having microwave-style matter compilers in the home? (2) Couldn’t a home matter compiler use blocks of fuel which are bought centrally, and added into the machine as needed? I have a color inkjet printer, but it would be silly to have special “ink lines” carrying fresh ink to my printer from a central ink factory so that it’ll never run dry. Obviously there’s a balance where if my “printer” actually created all of the food that I eat, perhaps “ink” would be worth piping into houses, but in the book not too many people actually ate compiled food regularly, if they could afford it, because it wasn’t as good as the real thing. (3) If nanotechnology reaches its full potential, that is, we can create anything by manipulating individual atoms, then it’s hard to believe that as the book implies, there is no such thing as even halfway decent A.I. or speech generation such that a network of humans needs to be on call at all times and ready to act out parts from interactive (“ractive”, as the book terms) games. A main theme of The Diamond Age is that the protagonistic young lady’s instance of the Illustrated Primer is read to her by a certain ractor who is on call (seemingly at all times, ready on an instant’s notice to start giving audio to words from the Illustrated Primer‘s dynamically generated story), and that somehow the “words between the words” not in the word choice of the primer itself, but rather in the tone of voice of the human narrator, have a strong effect on the outcome of the character of the girl who is under the influence of this particular Primer. If a nano-computer is smart enough to dynamically tailor a story to a particular (protagonistic, named Nell) young lady’s intelligence, emotions, and environment from moment to moment, and to create such humanlike prose, in fact to create dynamic and animated illustrations on the fly representing this story, then it’s surprising that such a computer couldn’t provide just as plausible a voice, as well. It’s surprising that a human voice can cleanly introduce nuance into a heavily-nuanced text which isn’t the nuance intended by the creator of the text itself, or that if the text itself isn’t heavily nuanced that it can be compelling in the first place, even when read with the right nuances created and introduced using tone of voice by a human reader.

Stephenson has a love affair with Turing machines and in my favorite segment of the book-within-a-book uses them in a metaphoric way to teach Nell how computers work and eventually to show her what their limitations are, and what the limitations of the Primer itself are. But I think he misunderstands the scale and complexity of such things, (not that I have a perfect grasp of any of this), because metaphorically, for the purposes of illustration within the world of the Primer it all makes sense, but then take it to the scale of real computer programs like the one actually running the Primer (which can write solid prose and even seem to understand on a very deep level the mind and the world of a child), and although a program like that can be translated to a Turing machine (which in computer science is only used as a theoretical tool for analyzing algorithms and how long certain problems would take to solve, anyway), one has to wonder whether the human brain itself is a Turing machine and if not, why not. Diamond Age (through the Primer) strongly says “not” but doesn’t even imply there’s a question of “why not”, although perhaps it tries to convey that “too much complexity” is the answer, so that we really are Turing machines, just ones that can’t be reasonably emulated because we’re so complex. That’s why I say “misunderstanding of complexity” because if you can manipulate atoms individually and build anything, why not build a computer far more powerful than the brain? After all, our brain cells are huge and inefficient. One cell would be made of a number of atoms measured in the hundreds of trillions (I think that’s a conservative estimate, too), and that’s not even a number you can conceive of. Anyway, I’m rambling (and reaching the end of this annoying tether on my attention span) so let me say that I enjoyed The Diamond Age, that the book painted colorful, saturated images, but that I think Stephenson is still in the process of (at the point that he wrote this book) maturing as a writer. I’d be curious to see what he comes up with if he comes out of the past (Cryptonomocon) and goes back to the future (no pun intended). I’m beginning to realize that it’s so much harder to write (well) about the future, and that even when you do, no one takes you seriously. (Although I fault them, for lacking imagination.)

Aha, here’s an interesting note, via a comment on this page, which quotes from an interview with Neal Stephenson himself:

Is cyberpunk over?

The best I can muster is that for a while, information technology was incredibly important, yet it had been ignored or gotten wrong by science fiction. There was this vast terrain of virgin territory, and there was a land rush. Now the revolutionary nature of that technology has become familiar. To make the obligatory social criticism kind of comment here, the bursting of the Internet bubble has proven that information technology is just another technology.

…and so it’s going to be for every technology, and it’s true that we have to evaluate a novel in terms of the technological zeitgeist in which it was written, but still, it’s fun to look at things directly in the present instead of through the lens of the past. I don’t think cyberpunk is over, just that “modern cyberpunk” is different, and that it’s very very hard to write. And on the other hand you have books like Neuromancer (which reminded me of the computer game System Shock, by the way) which are much more poetic and don’t delve all that much into specific detail, and so are more able to stand the test of time. Both types of work have their place.