I feel like I should be fascinated, and I suppose I am, but mainly because I find it all so confusing. This is not an explainer-type article. This is me working things out for myself. Join me. I will not get far.
Heer’s post resonated with me. As a consumer technology professional with decades of experience, I usually understand most technologies I encounter, at least at a high level. But quantum computing baffles me, and grandiose claims that it will develop self-healing materials, clean up microplastics, and eradicate world hunger only deepen my skepticism. I’ve spent years watching tech companies make all sorts of claims that either never work out at all or have significant unforeseen consequences, so these pie-in-the-sky announcements always put me on edge.
Perhaps it was inevitable. Companies like Microsoft and Google invest billions of dollars in fundamental research (which is good!), but their marketing departments (whom I suspect don’t understand quantum computing either) can’t help but search for user benefits for their announcements. The result is cognitive dissonance—it’s hard to imagine these companies’ quantum products solving the world’s ills when Google users complain about the quality of search results and Microsoft users frequently swear at Teams and OneDrive. Meanwhile, university labs that make significant strides in quantum computing rarely make it into mainstream tech news.
I don’t want to suggest that quantum computing is science fiction—it isn’t—but it appears to be years or possibly decades away from practical applications, making it difficult not to view the field in that light. Even then, I suspect that quantum computers will be far from what we think of as computing today. They won’t sit on our desks or live in our pockets—they’ll be the mainframes of the future. Until we reach that point, the entire field will walk a fine line between science fiction and science fact.
The quick summary: It’s a breakthrough, but only of interest to other quantum computing researchers. Because in the final analysis, it is only 8 qbits. It’s better than previous quantum devices, but is nowhere near large enough to be delivering the things that marketing departments are promising.
I tend to think of quantum computing with the same caveat as all quantum-mechanics-related things: “If you think you understand quantum mechanics, you don’t.” Everything about quantum mechanics is so counter-intuitive to our regular everyday interactions with physics that it takes special effort to not incorporate inaccuracies when you try to explain things in any manner other than pure math. Mathematics is the only realm in which quantum mechanics makes sense, and even that is only partial.
As a ‘technology of the future’ it ranks right up there with nuclear fusion and its unlimited energy potential, as being so massively hyped yet always so illusive.
Both are ‘coming in the next decade’’ to solve all our problems. But that was 30 years ago.
What little I understand about quantum computing corresponds with what little I understand about physics. I did a few design projects with MEMC back in the 90s, back when dual layer silicon for chips was the big new thing. The scientist in charge had a Mac, and was full of interesting info about the chipmaking process. I saw these kilns where they grew the logs of silicon to be chopped into wafers and then printed with the microcircuitry. He impressed on me the difficulty of mass producing anything that was so small that it had to be perfect on an atomic scale. Back in the 90s!
Yes, this is pure marketing. What isn’t said, though, is how even with all these amazing breakthroughs in physics, they really don’t seem to be able to solve the problem of what human brains can do with speeds that approach the speed of light. Behind the marketing we have theoretical breakthroughs in quantum physics that mean incredible progress for science in general, paid for by businesses that can well afford it, much like Bell Labs used to do in the 60s.
I gave a presentation on quantum computers at a DOS users group meeting in Sydney in 1999:
I don’t claim to understand it but I think I got the gist across: https://www.vdrsyd.com/aoaug/qtm_comp.html
Quantum computing is certainly not science fiction but it’s still far from becoming mainstream.
In short: it offers huge computing power, as long as you’re using a programming language that is able to make use of this technology and you house the quantum chip in a place where it can function properly. Since the functionality of this chip is based on physics (altering the rotation direction of an electron) it can be easily disturbed by movement or electromagnetic fields. Thus it can’t just be packed into a computer/blade and installed in a data-centre cabinet. Needless to say it can’t be packed into a wearable or household device. At least not in it’s current state of development.
Unlike CPUs and GPUs quantum chips won’t be usable with existing code, nor will it be possible to house the chips in a run of the mill data center.
Alas - once (or rather if) you get them to work and you write proper code they’ll outperform any existing CPU/GPU and yield enormous computing power.
What will that computing power be used for?
Probably more AI, better simulation, safer driverless cars (if we’re able to house them in a car), better security as well as… better scams, deadlier weapons etc… pretty much anything the can make good or bad use of compute power.
Quantum Computing is real, and highly likely to have real world effects in the next decade. There are two main areas:
Shor’s algorithm provides an exponential speedup on a rather narrow set of problems. That set includes essentially all of our public key cryptography, so we will have to replace all our public key cryptography. NIST has already published the algorithms (FIPS 203, FIPS 204, FIPS 205).
Most estimates say that a quantum computer large enough to break our current systems will be built in the early 2030s. In 2019 IBM published a roadmap with that end state, and they have made all their milestones on schedule, so it seems likely that they will meet the 2033 end date.
There is also Grover’s algorithm which provides a quadratic speedup on a much larger set of problem. Because that speedup assumes everything else is equal (computer speed and size), it’s less clear when it will become better than current computers.
For computing about quantum behaviour (such as how molecules work) quantum computers are already useful. This will contribute to things like drug discovery and material science, as they are fundamentally questions about how the atoms in a molecule will interact.
There are obvious problems:
The first quantum computers will cost billions of dollars to build, so you can only use them on very valuable problems.
They have to be kept very cold (perhaps 15 milli-kelvin) which is both hard and expensive.
Like GPUs, they only can do part of the computation, so they have to be attached to the “main” computer.
In short, for some very important problems they are a tremendous breakthrough, but that set of problems may be small.
I started looking into Quantum Computing after seeing this story on 60 minutes last year. A good overview of where we are and what the possibilities may be for the future.
My understanding is that the methods use quantum superposition. the idea is that you can perform a huge number of different problems simultaneously, and then you can extract the most probable. I assume that is why they are worried about cryptography. The methods rely on people not being able to factor very large numbers in a reasonable time, and quantum computing allows that to be performed simultaneously. A former colleague did some work on quantum computing and optimisation which i don’t understand but think it was based around being able to perform a huge number of function evaluations. This could be useful in defence where they are trying to rapidly make decisions about targeting.
Now the latest from NPR RadioLab is that migratory birds like whooping cranes may use quantum “entanglement” to migrate, “seeing” the magnetic lines via photons striking their (immobile?) eyes. Beats me!
Roger Penrose posited years ago that consciousness might arise from quantum phenomena. There are structures in the brain called microtubules that seem small enough, but as far as I know, no one’s been able to describe how, even if one could get a quantum system in superposition in the brain, it wouldn’t decohere immediately at body temperature (originally quantum computers had to be cooled to micro-Kelvins. Recent ones can operate in low single digit Kelvins, but that’s a long way from the ~310K of a human body).
I don’t have a Nobel like Penrose, but I did take a course in Quantum Mechanics and later studied the behavior of microtubules in an academic laboratory. Whenever I hear about Penrose and Hameroff’s conjecture about quantum effects, microtubules, and consciousness, I imagine a dictionary where the definition of the British phrase “taking the piss” is illustrated by a photo of Penrose and Hameroff laughing over pints in an Oxford pub.
