The Lifecycle of Software Objects
The Lifecycle of Software Objects is a novella by Ted Chiang about a group of people who become overly attached to their digital pets and try to raise them as their kids.
Setting and Plot #
The novella is set in the contemporary world. I am hesitant to call it a speculative fiction because the setting, the technology and the situations are dangerously close to our world. People inhabit virtual worlds on different software platforms. Entrepreneurs are experimenting with digital life and digital pets to entertain people. One such startup develops a software platform for people to buy, raise their own digital pets. These animats are called digients. The digients are special because the software company has reverse engineered genomes to develop them. These digients are close to actual organisms and might display emergent behavior unlike rule based artificial intelligence. Ann, who had been a zookeeper in her previous job, is an employee. She is hired to use her Pavlovian reinforcement learning methods to train, for lack of a better term, the toddler digients so that they are up for sale. During her testing, she becomes attached to one of the digients and buys one for herself. She, along with her other colleagues, raise digients on the company’s software platform. The software platform gives the digients a community to interact with other digients and evolve socially as well. The owners exchange notes on message boards to ask questions about raising digients.
Many other companies develop their own platforms for digient or digient-like software products. Various communities develop digients for their own research or economic gains like specialized digients for enterprises. Ann along with few other owners keep their pets alive after the original startup’s platform loses traction and the support ends. They do this by relying on third-party developers to write migration plugins for different platforms. The digients, in the meanwhile, have developed conscience of sorts and have begun to ask existential questions. The novella then focuses on larger questions about the nature vs nurture debate when the owners, who regard their digients not only sentient but also like people, have to face bigger challenges about how to prepare and let go of their children.
Characters #
Ann is one of the first users on the platform. She gets this job when one of her friends recommends her to the startup. She gets the job because her experience with actual animals might help her train new born digients using reinforcement learning. Ann develops infatuation with her child-like digient Jax. She trains Jax and is intrigued by Jax’s emergent behavior and is completely devoted to Jax. Ann suffers personal losses because of this. She strongly believes in “Nurture” vs nature in development. Despite numerous opportunities to shut down her pet, she does not shut it because of her mother-like devotion to Jax.
Derek is an animator with the same startup. He adopts two panda-like digients Marco and Polo. He suffers from personal losses as well. Derek is attracted to Ann because of their shared affection to digients. He likes digients to take their own course - become an adult, have autonomy etc. Derek represents the Nature part of the debate.
We follow Ann during the first half of the novella and sometimes Derek. The story progresses through news reports, message boards and Ann’s point of view.
Thoughts #
Ted Chiang is an amazing author.
I’ll say this again.
Ted Chiang is an amazing author.
He deals with a gamut of themes and philosophical debates with a simple narrative and brilliant anthropology. He describes the development of the software platform in vivid details. I have been a software developer for most of my working life. His ethnography of software workflow is spot on. The product ships in different stages of lifecycle - testing, debug and production release. Another interesting point raised is development of software platforms and their half cycles. How often have we seen a software platform losing traction of footfalls and the developers abandoning them. What happens to the content developed for the platform? What happens to the conversations and debates and comments and likes and uploads and captions and tags on the platform? The platforms with these software artifacts attract people and we value our time and commitment. When the software ceases to exist, the archives don’t carry the same nostalgia, the same emotions. Mutihoming of the platforms carries with it the same problems. Unless the migration is complete, or even when it is, the artifacts do not carry the same emotion or value. Something is always lost in translation.
Then there are the philosophical themes. When the digients migrate to shell, a hardware platform for digients, you cannot stop thinking of parallels with the concept Atman. Ann with her Pavlovian methods of teaching the digients is fascinated by the emergent behavior. I was reminded of Suzuki’s brilliant handbook about the nurture and caring of young children. As digients become sentient, they question the grand scheme of things, some of who face severe existential angst. At one point of time when Ann wants her digient to get employed, but they fail when compared with the advanced specialized AI, you tend to think of things like whether our own lives should be like a Homo Economicus specialized in Skilled labor or a Renaissance Man who empathizes with the human condition. Like the previous novel I reviewed, there are questions about Free will and our relation with our creations when we play God.