Chaos in a Crystal Learns to Think ✎ express

Original: "Responsive Disorder in a Metal-Organic Framework Enables Solid-State Reservoir Computing"
arXiv:2602.18335 · 2026-02-20 · arXiv non-exclusive · independent analysis · 1 min
Disorder in a porous material turns out to be capable of remembering and recognizing patterns.

A crystal that inside resembles a sponge with a labyrinth of pores turns out to be capable of performing computations. The material, named DUT-8, consists of metal nodes and organic linkers forming an intricate lattice. Under the influence of different gases, this lattice slightly changes shape, like a house of cards that rearranges with a gentle breath.

Scientists directed X-rays at the sample and monitored their scattering. Each gas left its unique 'fingerprint' in the scattering pattern, and from these fingerprints the system learned to distinguish input signals. It turned out that the material not only responds to the current stimulus but also remembers previous ones, like a person who flinches at a loud sound and then remains wary for a while.

Unlike a conventional computer, where information is stored in bits, here computations happen directly in the physical structure of the material. This saves energy and allows faster signal processing.

The system built from such material handled pattern recognition and sequence prediction as well as larger physical reservoirs. The authors called this phenomenon 'responsive disorder': the chaos inside the crystal turned out to be not a hindrance but a computational resource. Perhaps in the future, such materials will become the basis for tiny, energy-efficient devices that learn on the job.

Metal-organic frameworks are so porous that a single gram of such material can have an internal surface area the size of a football field.

The idea of a material that learns from external stimuli on its own evokes the 'living' ship from the series 'Firefly'—technology that adapts to its surroundings.

Scientists
William Lawrence Bragg
Concepts
reservoir computing quantum order disorder disorder induced phenomena neural network memory architecture neural network data reconstruction stochastic self organization density functional theory entropy driven self assembly data processing drug design redundancy dataset qubit diamond quantum memory logical qubit disorder robust machine learning neutral atom quantum computer solidification solid state qubit