Supergirl meets physics

kw: book reviews, science fiction, androids

Killswitch is the third novel of Joel Shepher's "Cassandra Kresnov" series. The series is approaching the limit of action-packing, on a level with books based on films like "Matrix" and "Terminator". I've commented on the author's skills and characterization before. He's gone a bit farther in the explicit sex department than before, to no good purpose that I can see...just pandering.

The prior novels established Ms Kresnov as a "best of the best" android superwarrior, tactician, cook, lover, and all-around great gal...an order of magnitude beyond "Ivan Skavinsky Skivar" in the old Crimean War ballad.

This time she has a handicap: She finds out that her brain contains a self-destruct mechanism, the killswitch, that can be activated over the network to which she is almost always wirelessly connected. So she is conveniently almost blinded at crucial times. She is also now only one of four androids at her level, and she has to fight—or ambush—a couple of them.

This time, I found myself wondering, what are the physical limits to performance of a human-size mechanism? I'll defer pondering brain power. Just how strong can her "myomer"-based muscle be, how fast can she run, and how high can she really jump?

In one scene, the author has a virtual sidebar that she has power enough to jump after low-flying aircraft, but too short a stroke to "realistically" jump higher than fifteen meters (~50 ft). If she weighs around 50kg, the gravitational potential to overcome is 7,350J. A one-meter stroke in her legs means an average 15 G's (height-to-stroke ratio), and probably double that in mid-stroke.

The leg bones of a healthy, young person can typically support about ten times one's weight, so if the muscles could supply the energy, a five meter leap (~16 ft) would be possible. Chuck Yeager brags about jumping off 15-foot fences as a youngster, and that's probably about the limit. Taking off is a much slower proposition, which is why high jumpers don't raise their center of mass more than about four feet (1.2 m).

Thus, Ms Kresnov's bones need to be at least twice the strength of a human's. Not out of the question. But the muscles? Mr. Shepherd's "myomer" must have really amazing energy storage capacity and release efficiency.

Figures are lacking for how fast she can run; it is just remarked here and there that people find her hard to follow (in a visual sense); she seems a blur in combat mode. The shoe isn't made that will allow a runner to accelerate with more than one G. That's pretty good, though: 9.8 m/s², so if her top running speed is of the order of 20 m/s (~45 mph), she can hit that in less than 3 seconds. Faster takeoffs require a wall to push off of...a very strong wall.

But how long can she sustain high energy consumption? A young woman at rest has a basal metabolism of about sixty watts. A sprinter who peaks at 30 mph (13 m/s) in a 100-meter dash must sweat about a quart of water during and after that 9-second run, because of expending a couple kilowatts for that time.

The kinetic energy reaches about 4 kJ, and much more energy goes into staying off the ground, but most of the total energy expended goes to heating the body, perhaps half a million Joules, enough to raise body temperature a couple degrees C (~4ºF). So it seems to me that a modest improvement in muscular efficiency is the main thing needed to give an android the means to run long and hard without excessive heating.

If "myomer" could be 50% efficient (compared to ~5% for muscle), the androids could run a few miles before breaking a sweat. This is more plausible than 15-m leaps. I wonder what genes we'd have to tweak to double the efficiency of actual, mammalian muscle...?

Well, it is fun to speculate. Nearly as much fun as the reading.

On the boundary between cyborg and human

kw: book reviews, science fiction, far future fiction, space opera, cyborgs, golems, androids, terraforming

Chris Moriarty is prone to leaving as much as he can to a reader's imagination and inference. For me, it is on the boundary between refreshing and irritating. Usually, when I can't figure out what his novel's characters just figured out, I get it within a page or two, sometimes because Moriarty explains, sometimes because my slow brain catches on.

Spin Control follows his successful novel "Spin State" with an equally readable, even more idea-packed story. The genre is hard science, far enough in the future that it verges on fantasy, but Moriarty has done his homework, and the technicalities are not-too-wild extrapolations of current trends and speculations.

Some four centuries hence, the human race has taken charge of its own evolution, and taken computing close to quantum limits, to produce a number of "near relatives" with various amounts of integrated machinery, from AIs of several kinds, to Golems (cyborgs with no childhood), to the all-biological but engineered products of several "syndicates", and at least one "Emergent AI" who has "lived" on the delicate edge of growth and destruction for a few centuries, partly due to its being "shunted" to a series of rented human bodies. Other AIs, that control groups of shunted human "volunteers" such as military squads, are found to be most stable when kept ignorant of what they are really doing, by being deceived that they are playing an elaborate game with really good virtuality; once one of these figures out what it is really doing, it promptly suicides...or is that self-destructs?

While many recent books have scene shifts nearly every page, thus resembling episodes of modern TV soap operas, each chapter in Spin Control has a venue and a cast, usually all centered on Arkady, the protagonist here. The chapters' scenes hang together better than those in most novels. Arkady is an A-level clone produced by the Rostov Syndicate. He is more victim than protagonist, only finding out what he was really sent to do in the closing pages of the novel.

Arkady begins by defecting from the Ring (space colonies in the synchronous orbit) to Earth. He'd previously been on a terraforming team to the well-kept-secret planet Novalis. Early survey and seeding probes had visited, and ten clones form four syndicates were sent to see how well the process is going. Seems it was going much too well, and some trouble ensued that Arkady doesn't know how to parse...yet. One result was a new kind of virus that enhances fertility, if that's all it does. Earth really needs that, because the population has crashed to the point that most people are walking ghosts—infertile and unlikely ever to reproduce.

Of course, there is more to it than that, and Arkady becomes a much coveted, much abused pawn in a high-stakes bidding war. It takes the long-lived Emergent AI Cohen and an old memory in a young commando to bring about the denouement: one really nasty character is eliminated, and the long-deceived military AIs learn not only what they are doing, but enough of what has kept Cohen alive, to take appropriate control of their own fates.

I like a hopeful ending...even if it is a setup for more sequels!