How a Quantum Computer Spotted the Arrow of Time express

Original: "Machine learning the arrow of time in solid-state spins"
arXiv:2603.10344 · 2026-03-11 · CC BY 4.0 · 2 min
Scientists trained a neural network to determine the direction of time from quantum processor data.

A film of a shattered cup played backward looks absurd: shards don't spontaneously reassemble. This intuitive sense of irreversibility is tied to the increase of entropy — a measure of disorder in a system. The second law of thermodynamics states that the entropy of an isolated system never decreases, setting time's direction from past to future. Yet the fundamental equations of quantum mechanics are time-symmetric — past and future are formally equivalent. So how does macroscopic time's arrow emerge from this microscopic symmetry?

To approach an answer, scientists used a quantum processor based on a carbon diamond, containing ten qubits — quantum analogs of classical bits that can exist in superpositions of states. On this device, they ran processes mimicking heat flow from hot to cold regions, along with their time-reversed versions. During the evolution, they introduced quantum measurements, which inevitably forced entropy to increase, injecting irreversibility. The resulting data, essentially 'traces' of the evolution, were fed to a neural network.

Without any explicit hints about time's direction, the neural network learned to distinguish forward heat flow from backward with 92% accuracy. It seemed to pick out subtle statistical signatures of irreversibility hidden in the measurements, much like spotting the faint wake on water to tell which way a boat moved, even without seeing the boat itself.

A generative diffusion model even reproduced key features of the directed energy flow, like an artist copying not just individual brushstrokes but the painting's very plot. This analogy, of course, simplifies: the model doesn't 'understand' the plot; it captures the probability distribution generated by the growth of entropy.

This isn't just an intellectual exercise. The ability to recognize time's arrow in quantum systems could bring us closer to building microscopic heat engines and help explain how our irreversible macroscopic world arises from symmetric physical laws. Remarkably, the diamond processor used operated at room temperature — a rare quality for quantum devices, which typically require extreme cooling.

Remove the measurements from the process, and symmetry was restored — the neural network could no longer tell time's direction. It's the quantum measurements that 'switch on' irreversibility.

The idea of recognizing time's arrow in quantum data calls to mind the film 'Tenet,' where inverted entropy makes time flow backward.

Scientists
Ludwig BoltzmannRudolf ClausiusJacob BekensteinStephen HawkingJames Clerk MaxwellEdward Witten
Concepts
arrow of time time evolution Machine Learning training atomic clock quantum clock quantum algorithm automation ensemble average learning thermal conductivity prediction automated analysis prediction accuracy pattern recognition ergodic theorem observation stochastic processes irreversibility thermodynamics reservoir computing