The Datafication of People
Upriver Press Book Review
The Last Human Job: The Work of Connecting in a Disconnected World
By Allison Pugh
Princeton University Press, 2024
While economists, computer scientists, and Wall Street analysts focus on how artificial intelligence might improve efficiency and productivity, Alison Pugh is thinking about how it might further depersonalize our society.
Pugh, a professor of sociology at the University of Virginia, argues that we are increasingly choosing to construct a “social architecture” around the values of technological efficiency rather than human well-being.
“Social architectures do not appear on their own,” she writes in The Last Human Job. “They are constructed of the choices of organizations and their stewards, bending to the twin pressures of market and state, prioritizing efficiency over relationship, and hoping that it won’t matter in the end” (p. 159).
Her concern is not new, but now the relentless drive for even more efficiency, by deploying artificial intelligence, has the potential to further neglect the formation of meaningful communities and social bonds. Specifically, Pugh says that AI could lead to dramatic declines in human interaction, potentially with devastating consequences for our overall well-being and economy.
Pugh positions the rise of artificial intelligence within a broader history: the gradual datafication of every aspect of work, economics, and social life. The belief in data, she says, has created “measurement regimes” even in highly human fields like medicine and education (p. 6). Everyone is forced to gather data and measure outcomes, which changes the dynamics of human interaction.
“. . . modern capitalism and modern bureaucracies converge on the priorities of data, accountability, and standardization in service to imperatives of efficiency and productivity,” Pugh writes, adding that these changes squeeze people into productivity molds, turning them into “industrialized objects” (pp. 8-9).
Some tech enthusiasts are attempting to push AI technology into the most relational types of work, such as using chatbots for psychotherapy. But Pugh demonstrates that these efforts will fall short, reducing the quality of the outcomes and degrading humans’ sense of dignity.
Her book’s central claim is that most work requires “connective labor” that is best accomplished by people. According to her definition, connective labor is composed of five elements: 1) using the body as an instrument; 2) reading and deploying emotions; 3) collaborating; 4) responding to spontaneity; and 5) making and managing mistakes (p. 128).
Connective labor is often difficult to monetize, but it is essential to almost all jobs. Some types of work—teaching, therapy, medicine, sales—tend to directly rely on the uniquely human ability to connect in authentic ways, but Pugh argues that almost all work requires human connectivity.
The Last Human Job pushes back against the view that AI will be good for the economy or for people. Pugh gives objective consideration to the capacities of artificial intelligence, but she demonstrates that humans have abilities that are essential and irreplaceable in nearly every vocation.
For example, the human body is able to do far more than manual work, which robots can sometimes do. The body also expresses and perceives emotions. It can give and receive affection. Most communication is nonverbal, as conveyed through facial expressions, hand gestures, and eye movements. Because of our amazing bodies, we can perceive when people are lying, anxious, authentically concerned, sad, joyful, and enthusiastic.
By contrast, artificial intelligence reduces the rich dynamics of communication to a bodiless transfer of digitized data delivered through an electronic device. This makes it much more difficult to establish trust or forge bonds in work contexts, such as business negotiations, teaching students, or serving clients.
Artificial intelligence programs force us to interact with impersonal algorithms. However, most work requires collaboration between people who work in teams. Humans have a remarkable capacity to collaborate with instantaneous and creative improvisation, as Jazz musicians demonstrate during a live performance. AI does not facilitate spontaneous, improvisational human-to-human collaboration at all. By switching us to interactions with machines, AI depersonalizes our lives and diminishes the quality of our work.
Pugh makes this point by using the example of teacher-student collaborations. Education quality declines when it’s based on mere data transfer. As everyone knows, learning happens best through interaction and relationships.
“Connective labor is a mutually constituted event, a reciprocal, momentary performance captured by the well-worn teacher’s adage that ‘education is more like lighting a fire than filling a pail.’ If connective labor were just about filling a pail, these workers would never have to worry about how to center the other, how to help them feel safe, how much to let them in; it is in its very interactivity that connective labor is both powerful and unpredictable,” she writes (p. 129).
In a business context, any successful company needs far more than high-speed information processors. Companies need leaders, mentors, teams, inventors, and a vibrant workplace culture. They need relational bonds, both among employees and between staff and clients.
In light of these facts, it seems likely that our trust in the wonders of datafication and artificial intelligence is overblown. Because so much depends on connective labor that only humans can provide, an over-reliance on AI will most likely reduce workplace effectiveness and business outcomes. AI tools can increase the speed of some tasks, but speed does not necessarily equal improved outcomes, or even better productivity.
Beyond the economics, the main outcome of both datafication and AI could be a loss in the quality and meaning of our lives. Pugh points out that heavy reliance on digital technologies can erode communities and make it harder for us to build them, the result being increased rates of loneliness.
We might fatalistically assume that AI is coming and there’s nothing we can do about it, but Pugh joins economists like Daron Acemoglu of MIT to remind us that we must choose how to use technology, and we must design the social architectures in which we live and work. All consumers have the freedom to say no to what is harmful and yes to what is good and beautiful.
So, what is the “last human job”? After reading this book, I suspect that the last human job is the first human job: to be fully human.
Quote to Ponder
“When machines and software control more and more of our lives, people will seek out more human-to-human connections—all the things you can’t download but have to upload the old-fashioned way, one human to another.” — Thomas Friedman, columnist for The New York Times