Jun 21, 2010

Yeti Review: The Lifecycle of Software Objects - Ted Chiang


In a few words: Chiang shows signs of growing pains in his longest work to date; presenting interesting questions about artificial intelligence but refusing to provide answers.

My Rating: 3/5

Pros: Chiang is an extremely strong prose stylist, creating emotion impressively quickly; the ideas of artificial but immature intelligence are fascinating and beg for further exploration; the 144 pages read very quickly and smoothly

Cons: The sum of its parts are more than the whole; Feels like a long short story rather than a short novel; many of the most intriguing concepts are left hanging and unexplored.

The Review: Ted Chiang has earned more genre awards than he has published stories. With a reputation like that, it’s hard not to have heightened expectations for The Lifecycle of Software Objects. While only 144 pages, it also represents the the longest work of his career. Chiang's latest offering follows two human programmers and the digital sentient lifeforms they help to create. Known as digients, these programmed minds start off as a 21st century version of the Tamagotchi fad of the mid 90s. Provided sufficient empirical attention and a partially randomized “genetic” code base, these “creatures” are capable of learning not unlike a pet or small child. It is the gradual development of these digients and the way their passionate owners react to their evolution that drives the story forward.

As such, the structure of Lifecycle is basically a timeline of critical moments in those relationships. This works well at first, allowing Chiang to jump a few months or a year or two to the next interesting development. Trends change, technology advances, and the digients continue to learn, progressing gradually through a childhood of sorts. This keeps the story fresh, only touching on a concept briefly before moving on. Some highlights were the debates regarding trial and error parenting (you can roll back the digients to prior save points) and how the digients reacted to and interacted with the real world when they are provided robot bodies.

However, it’s this structure that ultimately weakens the overall success of Chiang’s novella. The story feels more like a series of interconnected short stories than a comprehensive whole. The ideas introduced at the front of the book get plenty of attention as Chiang depicts both their implementation and downstream effect but toward the end of the book there isn’t enough time for the same level of exploration. This is particularly frustrating as the digients are becoming more and more capable, reaching the point of young adulthood. So when Chiang throws open the doors on a whole new sets of issues, including independence, self realization, and sexuality, there are dozens of different avenues of exploration he could take. But rather than exploring any of these promising new elements, Chiang chooses to end the book abruptly. The story needs either more words or less words, but I couldn’t tell you which.

That’s not to say the book is without merit. Having not read Chiang before, I was very impressed with his prose. It reads very cleanly and the pages fly by effortlessly. 144 pages, while short for a novel, is still a sizeable piece of fiction. It never seemed to drag and unfortunately this may have contributed to the sudden feeling of the ending. The emotional weight Chiang imbues in his writing was equally surprising, especially how quickly I became invested in not just imaginary humans but in imaginary, hypothetical, virtual creatures.

A great example of this occurs barely twenty pages into the book. One moment the digients are exploring their digital playground with all of the boundless optimism you would expect from a child. The next they are exposed to vulgarity for the first time by a careless technician, cementing a few choice words in their limited vocabularies. Because they are being developed to be sold as virtual pets for the greater online community, they have to be rolled back, erasing their childlike experience of amazement and discovery in a heartbreaking instant. While that recollection fails to do Chiang’s work justice, the amount of innocence he is able to instill his writing is very impressive.

Much like the adolescent digients described within, The Lifecycle of Software Objects feels like something stuck in the middle. There is no denying that Chiang's prose surpasses that of more prolific writers and the story teems with emotion and speculation. But it also feels like an overly long short story and lacks the conclusions that reward readers after investing time and emotion in a longer work. Instead the readers are presented with an abrupt conclusion that resolves little and feels like a trademark short story ending placed where one doesn't belong. Chiang’s writing screams with potential and it’s only a matter of time before he overcomes his growing pains and delivers a complete work. In the meantime, the intriguing concepts and prose laden with emotion make The Lifecycle of Software Objects worthwhile provided you don’t expect to find answers to the questions Chiang poses.

3 comments:

  1. Sounds interesting. Perhaps he deliberatly ended it in such a way to leave the possibilities open to interpretation?

    Either way, still sounds like something worth checking out.

    ReplyDelete
  2. @Anonymous - I felt that was the case but at the same time, it felt sort of like a cop out. I might just be subconsciously vindictive because I wanted more and it wasn't available

    ReplyDelete
  3. I'm going back through your blog as I only just discovered it, so I apologize for the "6 months later" comment.

    If you really have not read any other Chiang before, please, please go out and get a copy of his collection. It is probably the most influential collection of SFF I have ever read. Ever.

    ReplyDelete

Related Posts Plugin for WordPress, Blogger...