The Illusion of Prediction: Why the Brain is Not an Oracle, and How to Apply Anticipation in Biosemiotic Reprogramming
Key Takeaways (Executive Summary):
- Predictive coding and active inference models are fundamentally flawed when explaining anticipation and goal-directed behavior.
- Romain Brette argues that neuronal "predictions" compute current states rather than forecast future temporal scenarios.
- True anticipation involves exploiting ecological regularities to achieve goals, not minimizing sensory surprise.
- In Biosemiotic Reprogramming, this means engineering strict, artificial ecological regularities for the nervous system to exploit, rather than attempting to generate false predictions.
In contemporary neuroscience, the concept of "predictive coding" and its derivative theory of "active inference" dominate. The popular narrative states that the brain is a prediction machine, constantly generating hypotheses about the future and striving to minimize prediction errors. However, recent fundamental criticism by Romain Brette forces us to reconsider this paradigm (Brette, 2026).
Inference Error and True Anticipation
In his article, Brette points to a fundamental substitution of concepts. The original predictive coding model was created to describe the neural coding of static images, not to anticipate future events. The "predictions" of neurons in this model are computational guesses about the current state, rather than the anticipation of scenarios unfolding in time.
The theory of active inference faces even more severe criticism. It suggests that we select actions so that our sensory predictions come true (surprise minimization). Brette provides an elegant counterexample: if you see dark clouds, you predict that you will get wet without an umbrella, and stay dry with an umbrella. You take the umbrella not to "confirm the prediction," but because you have the goal of staying dry. Equating goals with predictions strips the theory of its explanatory power.
Anticipation is not the constant computation of the next sensory frame. It is the organism's ability to utilize ecological regularities (just as cyanobacteria use the day-night cycle) to maximize outcomes and achieve goals.
Perspectives for Biosemiotic Reprogramming (BSR)
For the methodology of Biosemiotic Reprogramming, these conclusions hold immense practical value. If anticipation is built on the use of regularities, our task within BSR is not to try and "trick" the brain with false predictions, but to artificially construct rigid ecological regularities that the nervous system will inevitably begin to exploit.
Brette, R. (2026). Predictive coding is not a theory of anticipation. Frontiers in Human Neuroscience, 20, 1838057.