Dennis Gabor

Dennis Gabor (1900–1979) is the only person to win a Nobel Prize for an invention he made while waiting for a tennis court. Born Gábor Dénes in Budapest, he belonged to that improbable generation of Hungarian scientists — von Neumann, Szilárd, Wigner, Teller — so productive that colleagues joked they were Martians posing as Hungarians. Gabor was the engineer of the bunch: trained in Berlin, he built cathode-ray oscillographs, plasma lamps and electron optics, fled Germany in 1933 when the Nazis came to power, and settled into an industrial lab in Rugby, England — where, on an Easter morning in 1947, while queueing for his turn at tennis, he invented holography in his head, complete, before reaching the court.

A repair job that became a revolution

He was not trying to make three-dimensional images. He was trying to fix the electron microscope, whose aberrated lenses kept atomic resolution tantalisingly out of reach. Gabor's insight was gloriously sideways: don't fight the bad lens — record the whole wave it produces, aberrations and all, and fix it afterwards with good optics. "Why not take a bad electron picture, but one which contains the whole information, and correct it by optical means?" He needed a name for a recording that, for once, threw nothing away — amplitude and phase, captured as an interference pattern — and coined hologram, from the Greek for "whole writing". There was one small problem: the scheme demanded light far more coherent than anything on Earth in 1947. His demonstration holograms, made with a filtered mercury lamp, were smudgy little ghosts haunted by their own twin images. The idea was correct, complete, published in Nature — and useless. It stayed useless for thirteen years, until the laser arrived in 1960 and holography abruptly worked, gorgeously. The 1971 Nobel Prize in Physics went to Gabor alone; he had lived to see his elegant failure become a field. (The electron microscope, with fine irony, got fixed by other means.)

The other masterpiece nobody notices

A year before the tennis court, Gabor had already quietly rewritten another subject. His 1946 paper "Theory of Communication" asked how finely a signal can be pinned down in time and frequency at once, and answered with an uncertainty principle for information: you cannot sharpen both, and the optimal compromise is a little wave-packet — a sinusoid under a Gaussian envelope — now called a Gabor atom. The windowed Fourier transform built from them underlies every spectrogram, and "Gabor filters" turned out to describe uncannily well what neurons in the visual cortex compute — which is why they appear as the learned first-layer features of convolutional networks. One man, two fields' foundational wave-packets: light and information.

Later in life, as a professor at Imperial College and a consultant at CBS Laboratories, Gabor reinvented himself as a futurist — and a strikingly good one. His 1963 book Inventing the Future delivered the line he is most quoted for: "The future cannot be predicted, but futures can be invented." It was not a slogan but a method — the same method as the tennis-court epiphany: don't forecast the world, build the missing piece of it. He spent his last decades worrying, decades early, about automation displacing work, energy limits, and what a leisure society would do with itself, all delivered with émigré wit and an engineer's insistence that every problem statement conceal a design brief. He remained cheerfully hands-on to the end — among his sixty-odd patents are a flat-television scheme and contributions to a holographic microscope — and he liked to note that his most famous invention had spent a third of its life as a beautiful idea waiting for its hardware. Patience, it turns out, is also a research skill.