The Lifecycle of Software Objects (book) by Ted Chiang

The Lifecycle of Software Objects is science fiction. Although I labeled it as a "book," it's very short, only 150 pages. If you don't mind reading something that length in a browser, Subterranean Press has made it available for free online. As far as I can tell, the entire text is included, although the illustrations are not.

Review:
 
The Lifecycle of Software Objects is an exploration of what might happen if AIs capable of learning, and possessing an unknown level of potential, were released as a toy. The closest real-life equivalents I can think of at the moment are Furbies and Tamagotchis, but 1) digients have far greater flexibility and potential and 2) digients have a primarily virtual existence, although they do acquire the option of real-world bodies later on.

The story follows two human characters. Ana Alvarado used to work at a zoo until she was hired by a company called Blue Gamma to train and test its digients. Derek Brooks is one of Blue Gamma's animators, charged with giving the digients avatars that are endearing and cute but not cartoonish. Over the course of several years, the digients develop personalities and new abilities, copies of their “infant” selves are released to the public for purchase, and some people adopt them wholeheartedly while others grow frustrated with the difficulty of training them. Eventually, new companies with different philosophies pop up, Blue Gamma folds, and Ana and Derek adopt their favorite digients (Ana's is Jax, a digient with a robot avatar, and Derek adopts both Marco, whose avatar is panda-like, and Polo, Marco's younger copy/sibling). A small community of digient enthusiasts tries to keep things going.

Reading The Lifecycle of Software Objects really did feel like watching a possible history of artificial intelligence. At first, I found myself wishing that I could be in that story, training a digient of my own. However, Chiang tried to make the story as realistic as possible, which included the issues involved with companies going out of business, shrinking user groups, aging technology, and a lack of funding. Ana and Derek were dedicated to taking care of their digients, which had an effect on their romantic relationships, employment choices, and more. In the end, they were faced with making some incredibly tough decisions.

The weakest thing about this story was its characters, which were bare-bones at best. Derek's marriage became rockier as his interest in digients grew and his friendship with Ana developed. Ana's devotion to digients meant that she had problems finding a boyfriend who'd stick around for more than six months. Readers were given more of a peek at Derek's private life and thoughts about romantic relationships than Ana's, but even that felt somewhat distant. The digients could have used more depth, as well. There were a few great moments, such as Marco's efforts to get Derek to give him more independence, but they always felt too brief.

This story excelled most as a thought experiment. It drew on ideas about personhood, child education, and free will. When you have a being that was created artificially, at what point might it be considered capable of making its own choices? Ana and Derek struggled with that question and with deciding which path would be the best one to take, for their digients and for themselves. On the one hand, it bugged me that, as far as I could tell, Ana never even asked Jax his opinion about the different options they were faced with. On the other hand, I felt that Derek let his feelings for Ana color his decision too much, especially considering that she'd never indicated that she had any romantic feelings for him. I was really glad that Chiang didn't turn Ana into nothing more than Derek's “reward.”

I should mention that this story touches on some of the darker possibilities for digients. Many digients were abandoned, much like unwanted pets. Mention was made of digients who'd had their pain receptors turned on for the amusement of sadists. The Sophonce digients, which had been programmed to have an obsessive interest in one thing (for example, puzzle-solving, which would make them helpful in virtual gaming environments), disturbed me – they almost seemed like addicts.

I was worried when the possibility of the Neuroblast digients becoming sex bots was introduced. This was the part that surprised me the most, because the “why it would be okay for digients to become sex bots” marketing spiel was so convincing even I found myself thinking it might not be a bad idea. I liked that the corporation approached it from more of a “sex education” direction and wanted the digients to be willing participants in whatever they did...but it was still a for-profit corporation. What if the digients proved to be terrible sex bots and weren't sufficiently profitable? What would happen if a digient genuinely wanted to get out of their contract? Would they even be allowed to want something like that? And what does “willing” mean, exactly, when you're dealing with a virtual being that can be reprogrammed to a certain extent?

This was an interesting and thought-provoking read. Despite some worrisome aspects, the ending felt hopeful, and the final illustration of a woman (Ana?) and a robot (Jax?) walking hand in hand reinforced that.

Extras:

There are various illustrations throughout the book. A couple of them may be considered objectionable by some readers: one shows a man destroying a robot with a hammer, while another shows a robot giving a human a blowjob.

Read-alikes:

• I, Robot (book) by Isaac Asimov - It's been years since I last read this, but I think those who liked Chiang's brief snapshots of digient development might like these short stories.
• Silently and Very Fast (e-short story) by Catherynne M. Valente - Valente's writing is more poetic than Chiang's, but this story may still appeal to those who'd like something else that's short and focused on artificial intelligence and its growth over time. It can be read or listened to for free online.
• Avogadro Corp: The Singularity Is Closer Than It Appears (e-book) by William Hertling - Book 1 of Hertling's Singularity series. I've read Book 3, The Last Firewall. This might be a good series to try if you'd like something else that looks at a possible history of artificial intelligence over the course of several years - each book focuses on a different point in AI history, from its beginnings to a more stable society that has accepted AIs as part of daily life. A word of warning: based on reviews and the one book I read, this series contains a lot of dangerous and out of control AI, plus a lot of AI and robot destruction.