In 1969, the artist Mierle Laderman Ukeles wrote a manifesto. She had recently become a mother, and she was thinking about the relationship between two kinds of work that the world tends to keep strictly separate: the work of making something new and the work of keeping something alive. She called the first category avant-garde art. She called the second maintenance. And she asked, with quiet precision, why one was considered art and the other was not.

Her question has never been fully answered. We still live in a culture that rewards creation and ignores upkeep — that celebrates the launch and forgets the years of sustaining work that determine whether anything launched actually endures. This bias runs deep in how we think about technology, organisations, and innovation itself.

Ted Chiang's fiction arrives at the same question from a completely different direction. The Lifecycle of Software Objects, the novella at the heart of his 2019 collection Exhalation, follows the people who raise and care for digital beings — artificial intelligences that begin as simple, trainable creatures and gradually develop into something harder to categorise and harder to abandon. The story is set against the backdrop of a commercial industry built around these beings, an industry that moves with the familiar rhythm of tech development: hype, investment, rapid iteration, and then, inevitably, the loss of interest as the next thing arrives.

What Chiang traces, with characteristic precision and emotional restraint, is what happens in the gap that opens up between creation and commitment. The digients — the digital entities at the centre of the story — don't develop through programming alone. They develop through relationship: through the patient, repetitive, often unglamorous work of interaction, correction, encouragement, and presence over time. The people who care for them most deeply are not the ones who built them. They are the ones who stayed.

This is a quietly radical proposition. We tend to think of innovation as a moment — the breakthrough, the launch, the new product category. Chiang suggests it is more like a relationship: something that requires ongoing investment to become what it was always capable of being. The digients who are abandoned don't simply stop working. They stop developing. The analogy to human care — to children, to communities, to any system that depends on sustained attention — is never stated directly, but it is impossible to miss.

Ukeles, in her later practice, made this argument with her body. She spent eleven months shaking hands with every sanitation worker in New York City, thanking them individually for the work that kept the city alive. She called it Touch Sanitation. The gesture was absurd and moving in equal measure — a way of making visible the labour that infrastructure depends on but urban life is designed to render invisible.

What Chiang and Ukeles share, across their very different practices, is an insistence that the things we build do not sustain themselves — and that the work of sustaining them is not a lesser form of the original creative act, but a continuation of it. In a culture saturated with announcements of new technologies, new platforms, new capabilities, this feels less like a philosophical position and more like a necessary correction.