Last Tuesday the Nobel committee handed out the Physics prize, and unless you follow this stuff closely you probably scrolled past it. John Clarke, Michel Devoret, and John Martinis got the nod for showing that quantum tunneling, the weird thing where a particle just sort of appears on the other side of a barrier it shouldn't be able to cross, happens at a scale you can actually build a circuit around. They did the original experiments back in 1984 and 85 at Berkeley, using superconducting loops cooled down near absolute zero. Forty years is a long wait for a phone call from Stockholm.
Here's why I actually care: that circuit, the Josephson junction setup they were poking at, is the same basic building block sitting inside every superconducting quantum computer being built right now. Googles chips, IBMs chips, the stuff Martinis himself worked on later when he ran the team that built Sycamore and claimed "quantum supremacy" back in 2019. Devoret left academia for Google too, hes been Chief Scientist over there for a few years now. Clarke stuck with Berkeley the whole time, which I respect honestly, not everyone needs to chase the industry paycheck.
What I like about this pick is it's almost a rebuke to how quantum computing gets talked about in the press. Every few months theres a headline about some company achieving a "breakthrough" thats going to break encryption or cure cancer by Thursday, and none of it ever quite lands the way the press release implied it would. The Nobel committee, meanwhile, went and rewarded the unglamorous 1980s groundwork that actually made the whole field possible in the first place, not the flashy recent demo. Its a good reminder that the hype cycle and the actual science run on completely different clocks. The stuff that matters is usually decades old by the time anyone outside the lab notices it.
I'll admit I have a complicated relationship with quantum computing coverage generally. I've read maybe two hundred articles over the years that open with some version of "imagine a computer that tries every answer at once," and I think that framing has done more to confuse people than help them. Thats not really whats happening, and it papers over how genuinely strange and hard the actual physics is. The tunneling effect these three guys demonstrated is stranger and more interesting than the "parallel universes" pop-sci version, honestly, and it got basically none of the same attention this week. Feels backwards.
Anyway. If you want the slightly more useful version of this story: macroscopic quantum tunneling means a system big enough to see and touch (well, a system built on a chip, not a single atom) can still tunnel through an energy barrier the way a single particle would, and that it does so in discrete, quantized jumps rather than smoothly. That single result is a big part of why physicists eventually believed you could build a working qubit out of a superconducting circuit at all, instead of needing to trap individual ions or atoms one at a time. Trapped-ion approaches exist too and have their own fans, IonQ being the obvious example, but the superconducting path is the one Google and IBM bet the farm on, and this prize is basically the origin story for that bet.
Not sure how many people are going to read a Nobel Prize writeup on a random Monday in October, but I've been following the quantum computing beat on and off for years now, mostly out of stubbornness, and it was nice to see actual foundational physicists get the trophy instead of whichever startup put out the loudest press release this quarter. That doesn't happen every year.