Artificial intelligence does not predict the future; it creates it

Will the day come when artificial intelligence (AI) systems like ChatGPT and Claude will be able to predict our decisions, navigate our lives, and even control the future of human society? For decades, such a scenario was the domain of science fiction writers. From Yevgeny Zamyatin and Aldous Huxley to Philip K. Dick and modern series like "Black Mirror," "Westworld," and "Devs," culture has explored how algorithms might determine human behavior.

However, in recent years, the line between imagination and reality has been blurring. AI systems are capable of processing vast amounts of information and identifying complex patterns. Today, they assist in screening job candidates, assessing medical risks, and ranking the potential of young athletes. As these systems improve, there is a growing sense that we are heading toward a future where algorithms will not only predict our decisions but also shape our life path. This concept is called "digital determinism."

The Limits of Predictability

In a new study we published in the journal Culture, we offer a more complex picture. We argue that there are fundamental limitations of nature and human society that prevent perfect prediction of human behavior. In other words, not everything that can be measured can be predicted.

One of our arguments relies on the world of sports. Even free throws in basketball—a routine action—are not identical. Despite constant conditions, the world's best players cannot reproduce the exact same movement time after time. Neurophysiological studies show that the source is "motor noise"—tiny, inevitable fluctuations in the nervous system. This is not an error, but a fundamental feature of the biological system.

Emergence and Social Systems

The same principle operates outside the court: a slight deviation while driving or a typo can change the sequence of events. Here enters the concept of "emergence," where the connection between many parts creates a completely new phenomenon. A traffic jam, for example, often arises not from an accident, but from a chain reaction of braking between drivers. Human society operates similarly: countless small interactions create patterns that cannot be predicted in advance.

Does this mean there is nothing to fear from AI? Not at all. Here lies the paradox: an algorithm is unable to predict the future, but the moment we grant it the authority to make decisions—who gets hired or receives a loan—it shapes reality in practice. Instead of predicting the future, it simply creates it. The most important value-based and social decisions must remain in our hands.

Prof. Ilya Morgulev is the head of the MPE Physical Education and Sports program at Kaye College. Prof. Felix Lebed is a researcher and lecturer at Kaye College and an international expert in the theory of complex systems.