Dr James Wootton

@decodoku

A superposition of quantum physicists

An example of the future of quantum computers, created out of their past.

Many people have contributed to quantum physics over the last century. Through their work, we now have an immense understanding of the quantum realm. And we can control it to make new and better technology.

One emerging example is quantum computing, which is currently going through a revolution in the sophistication and availability of prototype quantum processors. They now have enough quantum bits to play Battleships, and are even available on the cloud for anyone to experiment with. A hardware and software ecosystem is beginning to form, with big companies like IBM and Google as well as a host of startups.

I decided to use one of the cloud based devices to celebrate the giants on whose shoulders we stand. I took images of 16 Nobel prize winning quantum physicists, and I combined them.

I did this using quantum superposition. This is a phenomenon which effectively allows a quantum system to be doing multiple things at once. Specifically, I created a quantum superposition of the file names for the 16 images, creating something that simultaneously represented all the files.

This might sound quite complex, but the current crop of quantum SDK’s make it easy. I used the QISKit SDK, which is the one with native support for IBM’s devices.

With QISKit you control the quantum computers with simple Python. The code required to set up the superposition is just the following simple four lines.

program.h( qubit[b[0]] )
program.h( qubit[b[1]] )
program.h( qubit[b[2]] )
program.h( qubit[b[3]] )

Once this code is executed, the superposition exists in the quantum chip. But since this is a tiny device, sitting in one of the coldest points in the universe (created in IBM’s Yorktown Heights research lab) it is not very accessible. To make our creation into something more tangible, we first have to extract some sort of an output.

This is where the superposition stops being the kind of thing you’d find in sci-fi, and becomes something more mundane. Though superpositions are highly powerful within a quantum computation, allowing us to take faster routes from input to output, they aren’t so interesting at the end of one. In fact, at this point, they are little more than a fancy random generator of binary strings.

So that’s what happens to our superposition of filenames. We ask the quantum computer for a filename, and it just spits one of them out randomly. By repeating many times, we can get statistics on how often each one turns up.

This information was then used to construct our image. The alpha channel of each of the 16 images was chosen so that the strength of an image corresponded to how much the quantum computer liked it. Everything, from constructing the quantum program to processing the results, was done in this Jupyter notebook.

Ideally, all 16 images should be equally represented. But current quantum processors are not entirely free of noise. For one thing, there is a slight tendency for quantum bits to flip from 1 to 0. Filenames with more 0s will therefore be more likely to have their images turn up. This is probably why we get a good view of Max Planck’s mustache (0000.png) and Schrödinger’s glasses (0100.png). The devices will have to improve a bit before Serge Haroche (1111.png) gets fair treatment.

Even so, the average of all these physicists has a definite character: Pretty old, mostly white, and very male. This might be largely representative of the past of our quantum endeavours, but it won’t be the future. Cloud based quantum processors, and their accompanying SDK’s, put opportunities for quantum breakthroughs in the hands of everyone with an internet connection. Whether you want to do science or make a game, you can try it out with a quantum computer.

When we create a superposition of quantum programmers in a few years, hopefully it will look very different.

More by Dr James Wootton

Topics of interest

More Related Stories