How quantum computing will impact the future of Cities
Joe Kornik: [Music] Welcome to the VISION by Protiviti interview. I’m Joe Kornik, Director of Brand Publishing and Editor-in-Chief of VISION by Protiviti, our new quarterly content initiative where we put megatrends under the microscope and look far into the future to examine the strategic implications of big topics that will impact the C-Suite and executive boardrooms worldwide. In this, our first topic, The Future of Cities, we’re exploring the evolution urban areas are undergoing post-COVID and how those changes will alter cities over the next decade and beyond. We’ve got a good one for you today as I’m joined by Konstantinos Karagiannis, Associate Director of Quantum Computing Services for Protiviti. When we talk about how technology will impact the future of cities, we certainly need to include quantum computing in that equation. Simply put, quantum would enable problem-solving capabilities, involving a number of variables and potential outcomes at amazing speeds not attainable with conventional computers. Quantum-inspired algorithms could bring innovative solutions and approaches to product development, reduce time to market, optimize customer delivery, and speed up data transfers, but what impact could it have on the future of cities around the world? For that, I’m bringing in our expert, Konstantinos Karagiannis. Konstantinos also hosts The Post-Quantum World Podcast. Thanks so much for being here, Konstantinos.
Konstantinos Karagiannis: Yes, thanks for having me.
Joe Kornik: Okay, so this is a big topic. Let’s level set for just a minute if we could. We’ve sort of been hearing a lot about quantum computing for a few years now, so what is it? Where are we in sort of the maturation stage and how far have we come to this point?
Konstantinos Karagiannis: Yes, so quantum computing uses the mysterious nature of quantum mechanics to deal with a type of computing that was never possible before. Quantum computing is not just about building a computer that’s faster, but it’s also about building a computer that does things in an entirely different way. Most people are familiar with classical computing where you have binary, that’s zeroes and ones, that allow computation. In quantum computing, we have something known as a quantum bit, which, thanks to superposition, can be both zero and one at the same time, and it sounds weird. It doesn’t sound like a big deal, but with the right algorithms written to take advantage that, you can do things that have been impossible in the past or will become impossible just because of size. So dealing with a lot of data is a lot easier for a quantum computer. A classical computer, every time you try to improve it, it’s a very linear progression. People now look at Moore’s law. Classical computers get more powerful every couple of years or so. They double every couple of years or so. Quantum computers though, if you just added one of those quantum bits I told you about, you’ve just doubled the power of the machine, just to give you an idea how those things get explosively powerful over time. We’re still a little bit in the early days as far as maturation. We haven’t been able to prove a usable quantum advantage just yet, but we’re on the cusp. In fact, it could come as soon as this year that we’ll be able to prove some practical application that is just done better by a quantum computer even today, even with the machines we have now. What’s next would be taking these machines and making them even better, and we have plenty to discuss about that today, I’m sure. [Laughter]
Joe Kornik: Okay, so when we talk about cities specifically, where do you the see the biggest benefits of quantum computing as it relates to cities?
Konstantinos Karagiannis: Yes. So, those algorithms I told you about, that if you have them, just take advantage of the machine, they’re going to allow some pretty impressive things to happen on the space of cities. The big three with quantum computing is basically anything involving optimization, anything involving machine learning, or anything involving simulation, and all three of these things apply to the cities. The biggest one we’re going to see first would be optimization, because there’s a type a quantum computer called an annealer, and there are certain types of problems already that if you apply an annealer to them with a little bit of classical assistance, so you sort of have like classical computers doing some of it and an annealer doing the quantum part, we’re already starting to see things that are approaching advantage. For example, you can do an experiment where let’s say you have a city that has had some kind of catastrophe and they need to distribute goods to multiple locations in the most efficient way, we’ve already seen experimentation done on D-Wave’s hybrid system where a classical machine learning algorithm would try and find the best approach to deliver those goods, and it would take, let’s say, 27 miles of driving to get them to all the locations, and then the quantum computer was able to figure out paths, timing and everything to do it in only 20 miles of driving. So, from 27 miles to 20 miles, that’s pretty decent, and you can imagine how that would start to explode if you add in more variables, like weather or avoid this street because of crime or whatever. Quantum computers just do really, really well at handling larger and larger amounts of data, so we expect this type of thing to hit all avenues of smart cities, all types of planning with optimization. Just regular delivery routes, the classical traveling salesman problem people probably have heard about. It’s this idea of true efficiency, and to me, a smart city is not a smart city if it doesn’t have a perfect efficiency, right, and it just becomes a complicated mess with traffic jams and all sorts of things like that. Then, you move on to the pure machine learning side. Anywhere the machine learning is sort of touch a smart city, it will just be better with quantum given time. We’ve already seen the extrapolation of this happening, so anytime you’re trying to make intelligent decisions or automate processes, you’ll find that a quantum computer will be able to do things in thousandths of the time that it would take a classical machine, and then eventually gets even more exaggerated. What’s fascinating is this isn’t going to be like a neck-and-neck arms race with classical and quantum. Whenever quantum proves advantage in an area, that will be it. It will be almost like a straight line up if you think about graphing its performance. It’ll never be like classical could catch up again. Just add one qubit and you’ve doubled the system, et cetera, so over a course of a few years, quantum computers are going to seem to be practically [Laughter] magical at the things that they excel at. Then, on the simulation side too, we see massive impacts of that affecting cities. You’ll be able to do simulations about risks, simulations about how things will literally go, building materials can be improved by simulating which types of molecules handle stresses better. Basically, anything where you want to try and make a guess about how a certain flow on any given day will be, and in the numerous intricacies in the city, quantum machine learning algorithm would just be much, much better.
Joe Kornik: I’ve read quite a little bit about this quantum apocalypse that’s coming. I guess that means the day when quantum computers will be able to crack encryption, which certainly sounds bad. I know on the smart city side, there’s quite a few concerns around cybersecurity, but I would imagine that this is sort of on steroids. So, tell us a little bit about that, the quantum apocalypse, and tell us how bad could that be and what could it mean for cities, and what can be done about it.
Konstantinos Karagiannis: Yes. Way back in 1994, Peter Shor, a mathematician, came up with an algorithm that showed that we’d be able to reverse two very large numbers that were used to make an even larger number, which is basically the secret behind RSA, another type of public key encryption, like elliptic curve. So, once a quantum computer gets powerful enough, which could be five years away—it also depends on, like, if we can link them together, sometimes it might be able to do work more efficiently in the future, it remains to be seen. So, when that happens, all the old ciphers, because of this algorithm, will be able to be sliced apart. This will affect Bitcoin too, and other cryptocurrencies. So, for a smart city to excel, it would have to build on what we call post-quantum cryptography. NIST is currently working on new math-based ciphers that could replace RSA. Within three years or so, we expect to have fully approved and vetted new ciphers that can be put in place to avoid that quantum apocalypse. So, building a smart city, it would be smart to already consider from now having the crypto agility needed to build those ciphers in.
Joe Kornik: Yes, interesting. A couple of timeframes there, you mentioned a few years out from that. The speed of all of this—I mean, it seems a little mind-blowing, to me at least, a mere mortal. [Laughter] Can you put together sort of a timeline short-term of where we’ll be maybe in a few years in terms of quantum computing in cities? I mean, I know this is a little bit of a riskier proposition when talking about something that moves at the speed of quantum, but can you even hazard to guess as to what could be possible, say, a decade from now?
Konstantinos Karagiannis: Yes, absolutely. I talk about quantum timelines all the [Laughter] time with our customers. Some of them are concerned just about the crypto apocalypse we were talking about, so for them, the timeline is like just from now start looking at being able to adapt to have these new post-quantum solutions in place, but in general, they’re also concerned about how soon they can see real benefit with quantum. So, this is kind of how I see the timeline from what—these aren’t even just guesses. This is sort of like based on roadmaps and things from the people making the machines themselves. Within this year to next year, we’re going to actually see true quantum advantage in areas of optimization because of the amazing work being done with annealers, so that type of quantum computers are already getting pretty mature. It can only solve certain types of problems, but when mapped correctly, we’re already seeing—we’re just about there. For example, like we could do portfolio optimization now where if we were picking how to invest money, it might not be as accurate as classical, but it will be literally 1,000 times faster, so once we tweak the accuracy, we have the advantage. There’s all sorts of little tweaks and things like that that have to be made, so we’re going to see that with smart cities, too. If anyone wants to apply optimization back to any kind of NP hard type of problem, like a traveling salesman going everywhere without repeating your steps, we’re going to see advantage in that area within a year. So, this is the beginnings of the benefit to a smart city, and it’s already here. About two years from now in 2023, we expect to see multiple 1,000 qubit and up machines, and those are going to start to do really amazing things. The other type of quantum computing are gate-based, so that’s where you’re going to see explosive benefits to doing quantum machine learning, quantum simulation like I talked about. That’s really exciting, how close that is. Around 2023, I imagine there’s going to be a headline-generating paper almost every week, [Laughter] talking about how now we can do things so much better than we did before. That’s just 2023. Then by 2024, we’ll start seeing more interconnects and links, and the ability to connect quantum computers together to have them behave as if they were one machine. So, it’s not like when you’re trying to do grid computing now that you’re doing a little piece of a problem, and then sending the answer. You’ll almost be able to think of all of them as one machine, kind of sending calculations and progress over special connections for this quantum sort of internet. That’s really, really soon. In 2024 and 2025, we should start seeing explosive growth in the power of the machines because of that interconnect capability. Then, as far as a decade goes, I would say—I mean, Google has already announced that they plan on having fully error-free quantum computers by 2029, so that’s definitely within the decade. Once that happens, quantum computing, there aren’t even have to be any technical hurdles really to making it work. It’ll sort of be like, “Is it a type of problem that quantum is better at?” You just effortlessly encode it and really make it run. It should be quite an amazing time. The interfaces for programming are getting better too, so I can see that within a couple of years, it will be possible to just apply quantum in multiple areas without having to build it from the ground up [Laughter] every single time. This is really the most exciting time. That’s why we’re putting so much focus on it here at Protiviti.
Joe Kornik: Right. Will we see like actual—like when we move around a city in 2030, will we have noticeable differences or will it be so baked in to sort of everything that’s happening around us that we won’t be able to see it in action?
Konstantinos Karagiannis: Yes, because of the areas that it touches are going to make the big changes themselves, I don’t know if people will be aware of what’s happening. I mean, we all see how machine learning, AI, in general, has made life a lot easier. They’re going to notice that AI, all of a sudden, magically got a whole lot better, [Laughter] practically overnight in some areas. I don’t know how aware people are of this all the time, but anything that they thought that AI was doing a decent job at, all of a sudden it’s going to be doing a really great job at, so that could impact everything from like autonomous vehicles —they’ll just have that better sort of grid back-up or brain that is giving them the extra information they need to make better decisions. A city will be more hyperaware of these cars that are moving through it, for example. It will be able to give more real-time feedback and make predictions about where a car will be in a few moments and plan for it. So, quantum might make possible those terrifying images you see of cars shooting through intersections without any traffic lights because they’re all autonomous and slicing the neck out of each other. As much as I love this technology, I don’t know if I want to be in one of those cars. [Laughter]
Joe Kornik: Yes, it’s going to be interesting to say the least. Konstantinos, before I let you go, can you tell us a little bit more about the Post-Quantum podcast that you recently launched?
Konstantinos Karagiannis: Yes.
Joe Kornik: Where can people find it and what can they expect from it?
Konstantinos Karagiannis: Yes, it’s on all the platforms. You could just search The Post-Quantum World. Every episode, I talk to a company, a representative, a researcher, someone who is having a big impact on quantum today. We’ve already had some of the big machine makers, like Honeywell, we’ve had Microsoft who has their Azure Quantum platform where people could code, so it has a business focus too. Anyone who listens to it is going to hear about technology, what’s emerging and what's new, but also what it means to them, what it’s going to mean to—in the case of Bitcoin, like what it would mean to basically—and what if cryptocurrency was cracked by quantum computer and everything to, “Hey, what are the use cases? What are the things that my company can start getting benefits from immediately or in the near term? We try to cover it all and make it somewhat understandable. [Laughter]
Joe Kornik: Yes, I’ve given it a listen. It’s fascinating stuff. Thank you so much for your time today. Konstantinos Karagiannis, our Associate Director of Quantum Computing Services for Protiviti where he’s out there fighting the fight and helping our clients on a day-in, day-out basis to face the future with confidence.
Konstantinos Karagiannis: Thanks.
Joe Kornik: Thanks, Konstantinos. [Music]