The Lifecycle of Software Objects by Ted Chiang

Tamagotchis: pure 90's nostalgia.

I usually don’t mind the length of Ted Chiang’s short stories. More often than not, the extra pages are there to support the propulsive storyline or he's setting up for some kind of mind-blowing payoff at the end. The Lifecycle of Software Objects, though, feels a bit different. It’s closer to a novella and I’m not entirely convinced the story earns all those extra pages. Don’t get me wrong, it's still worth reading because this is still Ted Chiang, one of the best science-fiction short-story writers working today. He has an uncanny ability to cram a novel’s worth of ideas into a relatively compact form. Here, however, the seams start to show. The story occasionally spins its wheels, revisiting the same ideas, before drifting toward an ending that feels less like a conclusion and more like the opening chapter of a sequel.

So what’s it actually about? Imagine living inside an MMORPG, something like World of Warcraft or Skyrim. In this virtual world, there are creatures called “digients” that players can raise and train like pets. Powered by increasingly sophisticated AI, these digients start off performing simple tasks, but over time they learn new skills, develop personalities, and eventually communicate in meaningful ways with their owners. As the technology improves, they go through multiple iterations, growing more complex and self-aware, until their consciousness can be transferred into robotic bodies in the real world.

I’m glossing over a lot of the plot because Chiang packs in a ton of background detail and long stretches of technological and social evolution. The problem, for me, is that the emotional stakes never quite rise to match all that world-building. The story spends so much time mapping out the gradual progress of AI that it sometimes feels more like a case study than a narrative and I never fully connected with the characters, who came across as a bit stiff and underdeveloped despite the extra length.

The central theme of free will for artificial intelligence is at the core of this story. Chiang is less interested in flashy sci-fi spectacle and more concerned with the moral responsibility humans have toward the intelligences they create. At what point do the digients stop being property and start being something closer to people? If they can learn, change, and make decisions, do they deserve autonomy? Moreover, if granting them free will means exposing them to suffering, neglect, or exploitation, is it ethical to do so at all? The story raises uncomfortable questions about ownership, consent, and control, especially in a capitalist system that treats technology as disposable once it’s no longer profitable.

Ultimately, I’m a bit torn. There’s a lot here to admire, especially in how thoughtfully Chiang explores the idea of AI free will over long stretches of time. But the execution didn’t fully land for me and the ending left me unsatisfied. If you’re already a fan of his work, you’ll probably find plenty to chew on. If you’re new to Ted Chiang, though, this might not be the best place to start.