Google’s Quantum Supremacy Announcement Shouldn’t Be A Surprise

Google’s Quantum Supremacy Announcement Shouldn’t Be A Surprise

Headlines have boldly proclaimed that Google achieved “quantum supremacy” after the Financial Times on Friday leaked a draft of Google’s quantum supremacy research paper. I’m here to tell you that this is not a big surprise, nor is it a big deal — at least, not for you, yet.

But first, I’m going to ask you a favour: Before you read this article, go back and read this one that outlines the hunt for quantum supremacy, what quantum supremacy means, who should be excited about it (computer scientists, physicists, and Google), and who need not worry about it (most other people). Because everything that we said was going to happen is what’s now happening.

Last week, journalists with the Financial Times found a draft of a paper on a NASA website. It claimed that Google had achieved quantum supremacy. We also saw the paper, and it describes exactly what Google has publicly said it was going to do all along.

Google has declined to comment on the paper, which was clearly not meant for public release, but the gist is that Google says its quantum computer performed in 200 seconds a contrived problem that a classical computer would take 10,000 years to perform. The quantum processor probably isn’t “supreme” at any other computations; it only beat the supercomputer at this one, extremely specific task. That task, more or less, was to produce the output of a random set of operations that the quantum computer can do innately, but as far as researchers know, a supercomputer must simulate.

In short, the real announcement was that Google’s rudimentary quantum computer is better being at being a rudimentary quantum computer than a supercomputer is. The announcement is an early milestone, yes, but does not demonstrate any further abilities of the Google device.

Quantum computers are basically just computers built on a different architecture than the bits you’re familiar with. It’s an architecture that’s much harder to control and relies on the behaviour of subatomic particles, but the hope is that quantum computers might one day solve a certain subset of computer problems more efficiently than regular computers.

Among those problems are simulating the behaviour of molecules (which would be very useful in developing new pharmaceuticals) and solving certain maths problems, like factoring large numbers, that form the basis of modern cryptography. There are still plenty of problems (like the NP-complete problems) that, as far as scientists can tell, quantum computers can’t crack either.

There’s currently a lot of hype around quantum computers, and the quantum supremacy experiment is an important result demonstrating that they’re real and can actually do some things. But it does not demonstrate that they can solve problems with real-world implications. And scientists will long dispute whether the claim in Google’s paper is actually valid, because it’s an especially difficult calculation to check.

This did not stop the hype train, nor did it stop presidential candidate Andrew Yang from tweeting, “Google achieving quantum computing is a huge deal. It means, among many other things, that no code is uncrackable.” There is a lot wrong about this statement.

“Quantum computing” has been a thing for a while now, and as for the claim that “no code is uncrackable,” just, no. The National Institute of Standards and Technology is in the midst of running a contest for mathematicians to develop new codes that quantum computers can’t crack that will replace present-day encryption strategy, and per my last conversation with NIST, the search is going well.

Here’s another recent article I wrote that I’d really like you to read, about what the quantum internet will mean for you. As the headline states, you probably won’t even notice the change.

A quantum computer worth worrying about might be just on the horizon, or it might be decades away. But quantum computers are in a similar place today that computers were in during the 1950s: They’re large, inefficient, and expensive, with limited applications. Google’s Sycamore processor used for this experiment has only 53 qubits, the quantum-computing equivalent of bits, and those qubits lose their quantum behaviour after performing a handful of operations. Most quantum computer applications require thousands or even millions of these qubits working in tandem without losing their quantum behaviour.

Google’s supposed achievement of quantum supremacy isn’t a giant leap, and it certainly shouldn’t be shocking. It’s just another step forward on the long journey to a much more difficult-to-attain milestone: quantum usefulness.