How AI is restructuring social life
There are several concerns I and many others have about AI: heightened bias and discrimination, pervasive misinformation, widespread job loss, and untenable demands on natural resources. Yet in her new book Predicted: How AI Is Restructuring Social Life, UVA professor and sociologist Dr. Mona Sloane proposes a new, more pressing concern—that AI is quietly, fundamentally reorganizing society by eliminating our ability to imagine more than one possible future. She also proposes a gracefully common-sense solution. On September 10, LawTech Center and UVA Digital Technology Lab hosted Dr. Sloane at the Law School to share these ideas. I will attempt, to the best of my AI-limited knowledge, to fill you in.
Sloane explains that AI’s massive impact is a natural result of its design. All AI systems, including generative AI, are based on mathematical predictions that allow the model to learn the connections between words in order to best predict (and generate) which words should come next. When an AI responds to your prompt, it isn’t understanding you. Rather, it has predicted what makes the most sense to say back, a result of its excellent data synthesis skills.
AI has gotten very good at predicting what to say. It’s so good at what it does that it has become a mandatory condition of participating in society, “mediating access to resources and fundamental institutions such as education, insurance, healthcare, [and] the labor market.” Law students feel its presence at school as well—using Westlaw’s big beautiful AI Deep Research box, debating how best to use Claude to study, being told by the administration that AI will assess the very papers they suspect us of having used AI to write. Ultimately, “we can see how AI prediction is literally wedged between us—between individuals, between individuals and institutions, and between institutions themselves.”
Sloane proposes this “AI-wedge” has reoriented society, instilling a “heightened or higher attention to an ideal prediction itself.” The current global AI arms race is driven by the notion that “there is one future, that this future is predictable, and everybody can get in on the prediction game.” In a world where “leadership is synonymous with geopolitical dominance, it comes in the power of prediction markets.” That is, geopolitical power comes in the form of AI.
Sloane analogizes AI’s “prominent societal position” to that of the Oracle of Delphi in ancient Greece. “The social status of Delphi was considered so high that Delphi was considered the omphalos, the navel or center of the world.” Unlike Greek oracles, though, “the divine does not play a central role” in our current AI context. Rather, how we construct time does (literally, how we construe and measure time, like clocks). “Historically, whoever dictates the time regime holds power. Time regimes centralize economic-political power and social control.”
What sort of time regime does AI impose? A linear one. “It solidifies the idea that time, and history for that matter, meet onto a line that has a singular before, and a singular after.” Just like weather forecasts, “a prediction it will rain tomorrow requires knowing the weather yesterday and today.” If we buy into prediction as a logic so much so that we allow it to dictate how we perceive reality, “a general orientation toward one, noble future” will inevitably emerge.
For Sloane, then, “the big challenge that AI brings . . . is as a social infrastructure.” This infrastructure “quietly unravels the social significance of holding onto the idea that the future is simply unknowable.” If this unraveling continues at its current rate, “deliberating over many futures looks inefficient.” The dangers implicated by this perspective are abundant. For example: if the future is knowable, what is the point of elections? Why engage in the process of voting for one of many futures when AI can just tell us which one will occur? If AI's “paradigm of predication” takes hold, people no longer feature as agents, but rather as subjects waiting to be told what future they are bound to live in.
As it stands, AI poses an existential threat to our democracy. The goal of AI is literally “to reduce the number of moments in which the individual human is making a decision,” but democracy is made “exactly of those moments that are inefficient, cumbersome, unpredictable.” We risk getting caught in a dangerous cycle in which AI diminishes the quality and frequency of human interactions, and, as those interactions diminish, we become progressively more dependent on AI as our primary communicator, as our primary decision maker. Eventually, human-made decisions become obsolete—that is, democracy becomes obsolete. In its place, we will rely on AI’s prediction-based decisions. Seeing as predictions are inherently tied to the past, AI “can only find old solutions to old problems.” We will moor ourselves to the past, having lost our belief that we create our own future.
Thankfully, Sloane proposes a solution. It requires treating AI “as both an instrument and infrastructure.” As an instrument, its services are abundant. It could massively improve “public health, scientific discovery, prosperous economics, a stable grid . . . access to decent education.” These services, though, “are social questions. They are questions for everyone.” They are a public utility, and they should be treated as such.
How does society usually deal with the utilities and infrastructures that support society? We regulate them. Sloane suggests we push for AI to be regulated the same way, “the way you regulate corporations that supply water, power, telephone lines.” She’s calling not for more litigation in the long line of antitrust suits against tech companies, but for AI to fall into the hands of government. We don’t need to break up AI companies because they are monopolistic; we need to put them into the hands of the democratic government, where we can regulate AI as the public utility it is.
Sloane left on this note: “People do not do the thing you think they will do. And that’s not a flaw. That’s the human condition. And I think that’s the thing worth preserving in the age of AI.”
Author: Virginia Gray