Plot Twist! Quantum Physics Prevents Free Will, Physicists Show

A new argument links quantum physics to a hard limit on “agency,” claiming that if the brain and its environment operate purely under quantum rules, then quantum mechanics blocks the internal world-modeling needed for genuine decision-making. The core claim matters because it challenges a popular route to free will: the idea that quantum indeterminacy supplies the “multiple possible actions” from which a conscious choice selects one outcome.

The discussion starts by contrasting classical physics—where objects have definite positions and velocities—with quantum physics, where properties can be indefinite and outcomes are only predictable probabilistically. In that quantum picture, measurement is what turns possibilities into a single observed result. That indeterminism has long appealed to thinkers such as Roger Penrose and to proponents of the Copenhagen interpretation, including Niels Bohr, as a place where free will could “hide.” The intuition is straightforward: if several actions are available in a quantum superposition, then choosing one could correspond to a measurement-like collapse.

Most physicists, however, reject the idea of two separate regimes (quantum for the microscopic, classical for the everyday). Instead, they treat everyday classical behavior as the effective result of quantum physics becoming too subtle to notice at macroscopic scales. The new paper’s twist is that this “everything is quantum, but we can’t see it” view still doesn’t rescue agency. The authors argue that agency requires more than feeling like multiple options exist; it requires the ability to internally simulate the environment and compare the consequences of different actions.

That simulation demands repeated mapping of environmental possibilities into the brain—essentially building and updating a model of “what would happen if I did X instead of Y.” Here quantum mechanics introduces a barrier: the no-cloning theorem. Quantum theory forbids copying an arbitrary unknown quantum state without destroying the original. If the environment is fundamentally quantum and the brain’s internal representations are also quantum, then the brain cannot generate multiple accurate copies of the relevant quantum states to run parallel “what-if” simulations.

The authors further argue that even approximate modeling fails quickly. If the pizza, for example, is treated as a quantum state, then trying to create multiple approximate models of that state leads to rapid degradation in fidelity—the model stops resembling the real thing. From that, they conclude that a quantum computer cannot develop true agency, and, by extension, that human agency would require quantum physics to break down in some way.

The reaction is skeptical. Sabine Hossenfelder calls the no-cloning-based restriction “genuinely new” but doubts the leap from no-cloning to the impossibility of even approximately correct world models. She rates the paper six out of 10 and expects someone may overturn the argument soon. Still, the line of thought is framed as potentially productive—at minimum offering a new way to attribute bad decisions to quantum mechanics—while also pointing viewers to Brilliant for interactive explanations of math and science concepts.