I don't know anything about Quantum Computing, but that doesn't seem to stop anyone else from writing about it, so strap in.

All the time I read bla blah bla about AI from people who know little or nothing about it.

You know, almost everyone who writes anything about AI (apart from me, obviously).

Well, that's life, people have to make a living, and it's also clear that the outsider perspective is important when it comes to imposing technology on the community regardless of the desparate economic, social or geopolitical drivers that are pushing that technology to be so imposed. And also the rest of the population of AI writers (outside the almost everyone) have got better and better about writing about it in recent years and there are lots and lots of them (Michael Kearns, Melanie Mitchell, Gary Marcus, Mike Wooldridge and many more) so I am happy to smirk at the bullshit and learn from the thinkers.

So it's fun now to see a new wave of articles and (semi)learned reports about a different topic breaching out of the depths of other consulting companies and into my feeds. Every shyster, charletan and monteblanc in town is writing about Quantum Computing (QC - fed up with typing it) as if they know about it. So, why not me?

I've actually been very privileged to attend some meetings on QC in the past, once at MIT where I got the impression that the main thing about QC was that there were two types - a mystical type that was wholly unobtainable and held untold promise, and another type promoted by a startup company that was fictional and prompted head shaking and some good semi-good natured teasing ("so it'd work if you refrigerated it to -172?", "oh yes", "have you done that", "oh, no, that would be too difficult, it's as big as a wardrobe you see","well you can do it in my lab, it's as big as a truck and right down the hall". Cue foot shuffling and nervous giggles all round).

The other meeting was at Oxford, where I was part of a group that had a day long tutorial on the approach to Quantum Computing underway there. I was deeply challenged by the maths that was on show (I don't think that anyone intended that it be understood, it was just part of the air supply for the academics in the room), I was also staggered, shocked and deeply, deeply impressed, by the technolgy, scale and ambition of the work.

My view on the field is shaped by these experiences and the exposure to pop-sci reporting that everyone gets, and the occassional focused investigation to work out if cryptography really is buggered and what could be done if it was. I've tried programming the IBM QC online, also the University of Bristol one, I don't understand Shore's algorithm (in the way that I understand some other algorithms, like bubblesort). I know that there was a QC supremacy announcement a few months ago, I also understand why that was good (it was a demonstration that they really, really do work) and why it doesn't mean that our new QC overlords have arrived (it was a made up problem - the kind of made up problem that I wish I was clever enough to make up, but made up). So, that's where my thinking comes from, you are warned. Here we go.

QC theory is a boon for computer science - see Ewin Tang's work for example; it's exploited the tools that have been developed in other subfields and it's created novel algorithmic approaches that change how we think about computing. QC technology is a boon for application and development of technologies like superconducting magnets, photonics and cryogenics. The upsum of these requirements is that QC research doesn't just push individual technologies forward, but it's serving as a melting pot for multi-disciplinary collaboration at a fundamental level. People who are doing basic research in many fields are coming together to work on QC. Because of this I think that funding and pursuing QC is one of the best investments in fundamental research that's on the table at the moment, and will drive a lot of progress in unexpected ways.

Let's be clear - for what it's worth I think that QC is one of the rare fields of fundamental work (with no immediate application whatsoever on the table) that it is worth focusing large amounts of money on. Money that could be spent building particle accelerators... particle accelerators that will be used to zap kid's cancers, rather than the rather dubious ones that are used to prove that the Standard Theory is still true and that the worlds goverment are suckers for powerpoints using MS Comic Sans.

But I also think that economically viable application of quantum computers for anything at all is much more than five years away. I think that in five years there might be a demonstration of a "quantum eclipse" where a quantum computer beats the best classical computer for a real problem, maybe. But there are a host of problems in terms of practically running such a device (like the need to pour liquid nitrogen into it, and to mine the seabed for the minerals that are needed to wire it up) and I expect that the first quantum eclipse demonstrations will be just that - lab experiments that are run a handful of times as the performance of the scientific method.

Just as the groups that are now demonstrating QC supremacy are genuine scientific hero's the folks doing QC eclipsing will be the same. It's a great quest - it should be done, I want my tax dollars to be spent on it, but no one who is concerned about their kids schools should be concerned about QC's impact on their careers (unless they are research scientists). There will be a very hard and long road from QC eclipse to application, very long, very hard.

I think that when I am worried about my pension running out (if I am so lucky) then I can expect to see a significant impact of QC on our economy. Maybe it will be significant enough to mean that I won't have to worry about my pension, or the climate, or my childs future.

But by then I will be old, and so will you. Despite what everyone who knows nothing about it says.