So, I’ve been doing a bunch of talking in terms of decades in the last couple of posts, about the physics defining eras in the 20th century and the physics defining the last couple of decades. I’ll most likely do another decadal post in the near future, this one looking ahead to the 2020s, but the end of a decade by definition falls at the end of a year, so it’s worth taking a look at physics stories on a shorter time scale, as well.
You can, as always, find a good list of important physics stories in Physics World’s “Breakthrough of the Year” shortlist, and there are plenty of other “top science stories of 2019” lists out there. Speaking for myself, this is kind of an unusual year, and it’s tough to make a call as to the top story. Most of the time, these end-of-year things are either stupidly obvious because one story towers above all the others, or totally subjective because there are a whole bunch of stories of roughly equal importance, and the choice of a single one comes down to personal taste.
In 2019, though, I think there were two stories that are head-and-shoulders above everything else, but roughly equal to each other. Both are the culmination of many years of work, and both can also claim to be kicking off a new era for their respective subfields. And I’m really not sure how to choose between them.
The first of these is the more photogenic of the two, namely the release of the first image of a black hole by the Event Horizon Telescope collaboration back in April. This one made major news all over, and was one of the experiments that led me to call the 2010s the decade of black holes.
As I wrote around the time of the release, this was very much of a piece with the preceding hundred years of tests of general relativity: while many stories referred to the image as a “shadow” of the black hole, really it’s a ring produced by light bending around the event horizon. This is the same basic phenomenon that Eddington measured in 1919 looking at the shift in the apparent position of stars near the Sun, providing confirmation of Einstein’s prediction that gravity bends light. It’s just that scaling up the mass a few million times produces a far more dramatic bending of spacetime (and thus light) than the gentle curve produced by our Sun.
The other story, in very 2019 fashion, first emerged via a leak: someone at NASA accidentally posted a draft of the paper in which Google’s team claimed to have achieved “quantum supremacy.” They demonstrated reasonably convincingly that their machine took about three and a half minutes to generate a solution to a particular problem that would take vastly longer to solve with a classical computer.
The problem they were working with was very much in the “quantum simulation” mode that I talked about a year earlier, when I did a high-level overview of quantum computing in general, though a singularly useless version of that. Basically, they took a set of 50-odd qubits and performed a random series of operations on them to put them in a complicated state in which each qubit was in a superposition of multiple states and also entangled with other qubits in the system. Then they measured the probability of finding specific output states.
Finding the exact distribution of possible outcomes for such a large and entangled system is extremely computationally intensive if you’re using a classical computer to do the job, but it happens very naturally in the quantum computer. So they could get a good approximation of the distribution within minutes, while the classical version would take a lot more time, where “a lot more time” ranges from “thousands of years” (Google’s claim) down to “a few days” (the claim by a rival group at IBM using a different supercomputer algorithm to run the computation). If you’d like a lot more technical detail about what this did and didn’t do, see Scott Aaronson.
As with the EHT paper, this is the culmination of years of work by a large team of people. It’s also very much of a piece with past work— quantum computing as a distinct field is a recent development, but really, the fundamental equations used to do the calculations were pretty well set by 1935.
Both of these projects also have a solid claim to be at the forefront of something new. The EHT image is the first to be produced, but won’t be the last— they’re crunching numbers on the Sag A* black hole at the center of the Milky Way, and there’s room to improve their imaging in the future. Along with the LIGO discovery from a few years ago, this is the start of a new era of looking directly at black holes, rather than just using them as a playground for theory.
Google’s demonstration of quantum supremacy, meanwhile, is the first such result in a highly competitive field: IBM and Microsoft are also invested in similar machines, and there are smaller companies and academic labs exploring other technologies. The random-sampling problem they used is convenient for this sort of demonstration, but not really useful for anything else, but lots of people are hard at work on techniques to make a next generation of machines that will be able to do calculations where people care about the answer. There’s a good long way to go, yet, but a lot of activity in the field driving things forward.
So, in the head-to-head matchup for “Top Physics Story of 2019,” these two are remarkably evenly matched, and it could really go either way. The EHT result has a slightly deeper history, the Google quantum computer arguably has a brighter future. My inclination would be to split the award between them; if you put a gun to my head and made me pick one, I’d go with quantum supremacy, but I’d seriously question the life choices that led you to this place, because they’re both awesome accomplishments that deserve to be celebrated.