Derek's Reviews > The Adolescence Of P-1

First, let me say that I last read this in a pirated e-copy. Somewhere, I lost my paperback copy, and it's out of print, and the only way I could re-read it was to download an illegal copy from the net.

So sue me! Tom Ryan, if you contact me, I'll send you twice your legitimate royalty payment. And the epub I downloaded, which is MUCH better formatted than most books of a similar vintage that I get from legitimate publishing houses. It was a labour of love for somebody, while most of the traditional publishers who put out e-copies of backlists put no effort into it at all.

So, to the story. This is an extremely flawed book. For the technical reasons, see Rob Sawyer's review. There's a lot of hand-waving about how a computer actually achieves consciousness. On a more personal level, having been at the University of Waterloo just shortly after the events set there, I'm ticked that he sets UW in Kitchener (an adjacent city) and clearly knows nothing about the area (though I'd say he was in the Math faculty computer center)—London is not six miles south of Kitchener, that would be Cambridge (or maybe, at the time, Preston). London is more like 60 miles west.

On the other hand, there are moments that are sublime. When Burgess is first talking about hacking an IBM mainframe, his roommate says it can't be done.

"Have you ever tried?" Gregory asked
"No. And I don't have to try to know that I can't fly without an airplane."

That's brilliant, as it reflects the belief, less than a century earlier, that heavier-than-air flight was equally impossible.

Sawyer's own books about emerging consciousness in a networked computer (WWW: Wake & sequels) are technically better, but this tale is absolutely seminal! For all its faults, it's the progenitor of the computer-gains-consciousness-takes-control-of-the world genre. Other authors had had intelligent computers, but Ryan was the one who actually tried to explain how it could happen (he was wrong, but at least he tried!). Programming an AI is far more difficult than Ryan would have us believe (which is why there still isn't one, 40 years later), but at the time of writing it sounded believable. I suspect Ryan never got as far in Computer Science as Moore's Law, because he just doesn't think nearly big enough. P-1 is planning a new crystalline storage medium that can contain 4 billion "addresses" in a 150mm x 400mm cylinder. Even assuming an address points to a 64-bit word, that would only be 256GB of memory. In something that's actually much larger than my current laptop, which in one tiny portion of the entire machine has a terabyte hard drive.

But even knowing how badly he got some things wrong, and cringing at his American understanding of Canadian geography, the scope of his vision, and the things he got right make this utterly compelling. And he leaves us completely hanging! These days (or even a few years earlier with Colossus) there'd be a trilogy. A must read!