The Future Obsession: When prediction shapes the living

Prediction, selection, adaptation: a common thread

Anticipation as an evolutionary strategy

A living organism is a system that constantly fights against disorder. Contrary to a common belief, it does not consume energy in the strict sense — energy is neither created nor destroyed — but it draws from its environment energy in an ordered and usable form (food, light), which it then rejects in a degraded, disorganized form, mainly as heat. It is this flow of quality energy that allows it to maintain its internal order; without it, it degrades and dies. This fight against entropy has favored the emergence of anticipation mechanisms, at all levels of life. An isolated cell already knows how to react to chemical signals through gene regulation loops. But multicellular organisms went further by developing nervous systems capable of modeling the external world, that is, building an internal representation of it.

This need to reduce uncertainty is so fundamental that it has been formalized mathematically under the name of Bayesian inference. According to this approach, the brain would not be a simple receptor passively recording sensory information. It would rather function as a predictive engine, continuously generating hypotheses about the probable causes of what it perceives, then correcting them as new data arrives.

The free energy principle and allostasis

British neuroscientist Karl Friston (born 1959) popularized this vision through the free energy principle. According to this theoretical framework, any living system tends to minimize the discrepancy between its internal states and the external states it anticipates. Two strategies allow this minimization: acting on the environment to bring it closer to what was expected, or adjusting its own internal model to better match perceived reality.

Example: A simple example illustrates these two paths. At night, a suspicious noise in the house generates uncertainty that the brain seeks to reduce. Two strategies are possible: acting by getting up to check the source of the noise (the environment is explored until the prediction of danger is confirmed or refuted), or adjusting its interpretation by remembering that the house often creaks at night, making the innocuous explanation more probable than the worrying one (the internal model adjusts without any action being necessary). In both cases, the gap between what is perceived and what was expected is reduced.

This predictive logic underlies an essential physiological mechanism: allostasis. Unlike homeostasis, which corrects an imbalance once it has occurred, allostasis consists of proactively adjusting vital parameters (temperature, heart rate, hormone levels) before an imbalance even occurs. The body does not only react to the present: it actively prepares for what it predicts of the near future.

Example: Thus, heart rate and blood pressure begin to increase even before a runner's first step, as soon as the starting signal is perceived as imminent. The body does not correct for an oxygen deficit that has already occurred: it anticipates the coming effort and adjusts its parameters in advance, saving valuable reaction time.

The evolutionary cost of prediction

Evolution logically favored organisms with the most efficient internal models. This race for anticipation reaches its most sophisticated expression in primates, where two capacities have particularly developed: planning of future actions and social cognition, i.e., the ability to anticipate the intentions of others. Predicting what a conspecific will do has become, over the course of evolution, a major competitive advantage.

Example: In the laboratory, bonobos and orangutans chose and carried with them a tool with no immediate use, to use it up to fourteen hours later. This behavior reflects a form of planning oriented towards a future need, not an immediate reward. Social cognition in chimpanzees follows a similar logic: facing a dominant individual, a subordinate individual abandons the food the dominant can see and heads towards the one only he can see. Such a choice implies that he represents what his rival perceives, and anticipates his behavior accordingly.

But this ability comes at a price. A brain that constantly anticipates is also a brain that makes mistakes regularly: each gap between prediction and observed reality generates a prediction error. By repeating itself or being poorly regulated, these errors eventually foster a state of anxiety and chronic stress. This is the biological price to pay for a brain that has bet on predicting rather than simply reacting.

Example: Waiting for a medical test result clearly illustrates this cost. The body reacts not to a real and present danger, but to uncertainty itself: the simple fact of not knowing whether the prediction will be good or bad is enough to maintain a state of prolonged physiological vigilance (tension, sleep disorders, rumination). The stress felt is therefore not a response to a fact, but to an unresolved prediction.

The Tyranny of Prediction: when uncertainty becomes a threat

Faced with uncertainty, our brain uses cognitive biases to simplify things and avoid doubt

Psychologist Daniel Kahneman (1934-2024) showed that the human mind is allergic to uncertainty. Our heuristics, these mental shortcuts, are attempts to reduce the complexity of the world into reassuring narratives. Confirmation bias, pattern seeking, or the illusion of control are manifestations of this compulsion. Faced with a random situation, we often prefer an erroneous but reassuring explanation over a lack of explanation.

Example: Psychologist Ellen Langer showed in 1975 that players allowed to choose their own lottery ticket believed they were more likely to win than those who were randomly assigned a ticket, even though the objective probabilities were strictly identical in both cases. The simple fact of having "chosen" was enough to create a feeling of mastery over a purely random event, a classic illustration of the illusion of control.

The daily harassment of prediction

This cognitive tyranny "harasses" us daily: we check our phones to anticipate messages, we check the weather forecast to predict rain, and we develop future scenarios to ward off anxiety. In extreme cases, this hyper-prediction turns into generalized anxiety or obsessive-compulsive disorder, where the mind desperately tries to control the uncontrollable.

Example: The reflex of checking one's phone dozens of times a day is partly explained by a mechanism identified as early as the 1950s by psychologist B. F. Skinner: variable interval reinforcement. As with a slot machine, the uncertainty about when a notification will appear prompts checking the screen more often than if messages arrived at predictable intervals; the brain remains hooked on predicting the next signal.

Black swans: the ordeal of the unpredictable

Philosopher Nassim Nicholas Taleb (1960- ) has called unpredictable, large-scale events "black swans". Our inability to anticipate them, and our tendency to construct post-hoc narratives to explain them, perfectly illustrates the gap between our predictive biology and chaotic reality. The history of science is littered with such surprises, where our best models have collapsed in the face of the strangeness of reality.

Example: In 2008, the financial crisis took everyone by surprise. No serious economic model had predicted it. But as soon as it broke out, explanations multiplied to make it "logical" and "inevitable." The problem? These explanations did not exist beforehand. That is a black swan according to Taleb: an unforeseeable event that we can only explain after it has happened.

Technology and prediction: an impossible mastery

Tools to tame uncertainty

Faced with this anxiety of the uncertain, humanity has developed predictive prostheses of unprecedented power. From mathematical models to AI (artificial Intelligence) algorithms, via Big Data, we have built a technological edifice aimed at reducing informational entropy. Meteorology, finance, predictive medicine, or the simulation of complex systems are fields where we deploy colossal means to "tame" the future.

Example: In predictive medicine, genetic tests estimate your risk of developing certain diseases. But they only give a probability, never a certainty: a "low-risk" person can get sick, and a "high-risk" person may never develop the pathology, leaving uncertainty entirely intact.

The measurement paradox: quantum and chaos

But in doing so, we encounter a fundamental paradox. On the one hand, quantum mechanics, with its uncertainty principle, reminds us that reality, at the microscopic scale, is irreducibly probabilistic. On the other hand, chaos theory shows that deterministic systems can produce unpredictable behaviors in the long term (butterfly effect). Even our most sophisticated tools do not eliminate uncertainty: they frame it, at best.

Example: The butterfly effect illustrates how a tiny initial change, like the flapping of a butterfly's wings in Brazil, can trigger huge and unpredictable consequences, like a tornado in Texas, weeks later. Our mathematical models, however perfectly deterministic, are incapable of tracking such sensitivity to initial conditions.

Example: Long-term weather forecasting remains an intractable challenge. With supercomputers capable of processing billions of data points, we still cannot predict the weather in a month, because the slightest variation in atmospheric pressure or ocean temperature changes everything. Uncertainty grows over time, no matter what we do.

The trap of prediction: when evolutionary advantage becomes a handicap

An advantage selected by evolution

For millennia, the ability to anticipate the future has been a major asset for the survival of the human species. Our ancestors who could predict the arrival of winter, the movement of herds, or the approach of a predator were more likely to survive and pass on their genes. This tendency to project future scenarios was rewarded by natural selection. It is inscribed in our brains, shaped by evolution to keep us alive in a dangerous but simple world.

When anticipation becomes an addiction

But today, this same reflex works against us. In an environment radically different from that of our ancestors, our need to predict turns against us. We want to anticipate the economy, the weather, financial markets, health risks, social trends… and we regularly fail. Worse, this quest for perfect prediction makes us blind to the unexpected, makes us anxious about uncertainty, and pushes us to take senseless risks.

The cost of the predictive illusion

This inability to give up prediction has concrete consequences. In a world where uncertainty is the rule, our addiction to anticipation makes us more vulnerable, not stronger. We invest billions in models that are wrong, we make decisions based on fragile forecasts, and we neglect the weak signals that do not fit our frameworks.

The reversal of natural selection

The paradox is striking: what allowed us to dominate the planet is becoming a handicap to the point of reversing natural selection. It is no longer the best forecasters who survive, but those who know how to deal with the unpredictable. The ability not to know, to tolerate uncertainty, and to adapt in real-time becomes more valuable than the ability to predict. The most resilient startups are not those that predicted the market best, but those that were able to pivot quickly in the face of the unexpected. Airbnb, Slack, or Instagram emerged not because of a perfect vision of the future, but because their founders were able to adapt to situations no one had anticipated.

Towards an intelligence of the unknown

In a hyper-complex world, anticipation has become a trap. True intelligence may not be about predicting better, but about cultivating our ability to navigate the unknown.