Friday, 3 June 2011
It may be Ballantyne's 2004 debut, but it looks certain that his later work will never progress beyond amateur hour in comparison to (for example) Ian M Bank's fecundly imaginative narrative, or Reynold's brooding atmospherics. The scant few unfamiliar technologies presented in Recursion are named entirely literally, like the pivotal "Von Neumann machines" (or "VNMs"), for example. The number of raw ideas in Recurusion would barely sustain a couple of chapters of a Stross book. Also, either of my aforementioned favourite writers usually leave me with a post-it note full of fun new vocabulary that required Googling (if I'm in a literary mood), I almost feel I might have done a better job myself.
The 2051 (chronologically earliest) storyline was the most compelling in terms of character, and managed to build tension, terminating with a modicum of surprise (although it may as well have been set in 2020 in terms of future-fantastic).
The spy thriller thread (in 2119) was unconvincing and the coherence of the protagonists falters badly when he tries to write a frantic fleeing scene, that incidentally felt like it owed much to 'Eternal sunshine of the spotless mind'. In fact the cumbersome way details are often spelled out, some needlessly arbitrary events and the odd gaping continuity hole leave the whole thing feeling like the synopsis for a Hollywood movie. Which seems to be where much of the author's inspiration comes from.
My impression from the opening page, with 'Herb' in 2210, is that it could have been written in 1964, not 2004. Incidentally, his impression of a black skinned antagonist in this opening chapter left me somewhat uncomfortable (embarrassed on the author's behalf). Herb is the least believable character, supposedly smart (for a human) he is made to act overly retarded, attempting to paint the robot character as implacably smart, and allow many details to be spelled out to the reader through dialogue.
+ Nuts, bolts and spoilers:
We are presented with cliche Star Trek space travel: unaugmented humans in penthouse luxury apartment interiors with magical artificial gravity and convenient "warp" travel. Then into this blandness Ballantyne shoehorns in the caveat that all modern technology can self replicate, magically fission into 2 identical copies, each with half the original mass, a whole spaceship in a matter of minutes... No suspension of disbelief there for me!
The depiction of VNMs was overly simplistic for my tastes too (and their abilities too unphysical). He glosses over how they could self assemble equally easily from any rocky planetary matter, regardless of element (avoiding talk of gaseous planets altogether). Each batch of VNMs are are monoculture of clones, I would have liked to have seen specialised body plans, totally different shapes and sizes all working together like the 250 odd diverse cell types in a human body.
The three converging lines did indeed tie up on pat, after several rounds of very evenly sized chapters each. But the final chapter felt like he had run out of writing steam altogether and just wanted to finish off communicating all his thoughts quickly. He wrestles, like a man with no arms, with the idea of what the existence of an omniscient AI should mean for humans. It starts to look like he's about to go with a tired, pseudo-religious, movie industry style, meaningless moral (of things are just best off with humans being as them are, etc), but then avoided pissing me off entirely (like 'Calculating God' did, with a similarly engaging writing style too). It just left me 'meh'.
The core flaw in it's underlying philosophy are it's lack of conception of where ideas and technological change come from and the unstoppably disruptive effect this progress (increasingly) has. There is clearly no memetics at play here, he talks of all the advanced technology being impossible for humans to understand, requiring the omniscient/Omnipotent world brain AI to drip feed such things to humanity on a whim, arbitrarily decades ahead of the generally available state of the art. Formulaic, lazy, nonsense. Idea's can not be conjured into existence as if intelligence were 'the force', ideas evolve from a pool of existing ideas that continually grows, whether in human brains or virtual environments.
His super-intelligent AI are also able to predict deep into the future, apparently oblivious to (or ignorant of) the presence of computationally irreducibility, particularly in the chaos of reality.
He doesn't explicitly deal with the lack-of-technological-singularity-problem, but it could be argued that the supreme AI guided things away from such explosive changes, rather than herald them as might typically be expected. And anyway, I'm used to this oversight.
This was a fairly harsh review. To be fair, I was sufficiently absorbed to finish the book, and it did present a few ideas to think on. If nothing else it inspiring in that it left me feeling hope that a geeks with merely nominal literacy can write and publish full length sci-fi too.