Quantum Computing will radically change not only the sorts of problems we can take on with computers but also the way in which we think about computing and the way the the building block of the universe work. A fascinating journey through the minds of those building tomorrows quantum computers. Bob Sutor, VP for IBM Q strategy and Ecosystem at IBM Research joins Tom for part one of this two part podcast series on quantum computers.
VP for IBM Q strategy and Ecosystem at IBM Research. Learn more about Bob on his website
TK: This week on Foresight Radio, Quantum Computing: How the Next Era of Computing will Harness the Sub-atomic Properties of Quantum Bits to Perform Tasks that Classical Computers will never be Able to Take On. Joining me is Bob TK, Vice-President for IBM Q Strategy and Ecosystems at IBM Research. I’m your host Tom Koulopoulos, and Foresight Radio is brought to you by our good friends at Wasabi Technologies, the leader in the next generation of cloud based data storage. You can learn more about them at wasabi.com. Here’s my conversation with Bob TK.
Try introducing the term quantum computing into just about any conversation and it’s always a good way to figure out exactly how esoteric of a crowd you hang with. Quantum computing is wrapped up in all sorts of mystique and misunderstanding and yet it’s a term that we’re hearing more and more. The best way to think of quantum computing is to liken it to where we were with our understanding of today’s computers back in the 1960’s when computer scientists of that era tried to explain the simple principles of binary operations and programming. There were a few ways to actually make them understandable and yet today we’re surrounded by 10 billion classical digital computers from our wristwatches to our automobiles to laptops, we take digital computers entirely for granted. Let’s face it. Most of us don’t really know how even a classical digital computer works. It’s not necessarily to understand that in order for it to create value and to do its job but classical computers are reaching their limits. Moore’s Law which has predicted that doubling of the density of transistors every 18 months is reaching its limits. The pathways on intermittent circuits are becoming so small that they’re nearing the point where they simply cannot pass any more electrons through without encountering this strange rule of quantum behavior. As with pretty much any technology limit that we encounter, at some point the answer to progress is not found in the same tools and technologies that created the problem but in an entirely new way of addressing it, and that’s where quantum computing steps in. To better understand what they are and how they work, I interviewed a group of engineers and scientists who are working in this spooky new world of quantum computing and quantum mechanics.
It was in a recent event that was held at the Boston Museum of Science called NanoDays with a Quantum Leap. I’ll warn you ahead of time that understanding quantum computing is part science, part philosophy and part letting go of so much of what you already know about computing, and it’s that last piece that maybe the hardest obstacle of all to overcome. To make it easier to follow along, I’ve broken up my interviews into two podcasts. We’ll air the second podcast at the beginning of August. Here is part one of my interviews on quantum computing.
One of the big challenges of quantum computing is trying to describe what it actually is. So, we went to Bob TK who is the VP of IBM’s Q Strategy and Ecosystem at IBM Research to find out from him how he would describe a quantum computer. Here’s what Bob said:
Bob: Sometimes I toss that back at people and say, “What do you think a classical computer is? How does a regular computer work?” Because a lot of times when we try to answer this question, what’s a quantum computer, it’s nice to be able to contrast it and say, “Well, a classical computer works this way, a quantum computer works that way.” So, just a little bit on classical and then how I understand it for quantum is classical computer architecture, these are the chips that are in your phone, your laptop, the servers that run the internet and so forth. This architecture really goes back to about the 1940’s and it’s built on bits, zeroes, and ones and you put lots of bits together and you get bytes, and kilobytes, the megabytes, and gigabytes and so forth, things like that. Of course we’ve been very spoiled through the decades. We’ve had Moore’s Law which of course isn’t a law, it’s just kind of an observation, which meant that we could kind of say that if you waited long enough, computers - these classical computers would get powerful enough to do what we wanted. A very valid strategy was have patience, we’ll get there.
People started asking these questions in the early ‘80s and saying, “Are there really some problems that classical computers just are not designed to do?” Is there something inherent in their architecture that says, “This problem, you can express it. We could solve it but it would only take 10 million years. Is that okay?” Of course it wasn’t okay. So, the first observations were of that nature, and so here by nature we mean electrons, atoms, molecules, the things that make up you and me and everything around us, obey the laws of what’s called quantum mechanics which is a part of physics which is strange and wonderful and confusing. As you might guess from the name quantum computing is related to quantum mechanics, but the observation was the same. Instead of using only these zeroes and ones, could we build an analog? Could we build quantum bits that obey the same laws of quantum mechanics? In fact, it’s part of the way you program these machines. That’s where it really kicked off and so really down at the lowest level. We don’t use basic logic gates; ands, and ors, and nots, and things like there’s much - as classical computers, we use the quantum properties because we have created essentially qubits which are artificial quantum particles and amazingly enough, we can put these together and we can program with these, and that what’s quantum computing is about.
TK: By now, most of us are aware of the fact that classical computers, as Bob refers to them, use what are called bits and these can be either ones or zeroes, and if you string the bits together, you get bytes, and from bytes you can construct all the information that makes up this data set that classical computers use. With quantum computers however, we’re not dealing with ones and zeroes anymore. We’re dealing with these funky things called qubits which have what are called superposition. They’re never a one or a zero until you actually ask them what they are. Then they reveal their true identity and that sounds pretty obtuse, so we asked Bob to give us some clarification on what a qubit is and how you capture these and actually marshal their resources to create quantum computers. Here’s what Bob had to say about qubits.
Bob: The remarkable thing about quantum computing and the fundamental unit, if you will, of quantum computing as I said before is the quantum bit or what we called the qubit. Every time you add another qubit into your system, and I’m not saying it’s trivial to do so. It’s not just manufacturing one more qubit and sticking it in. It has to fit within an architecture, but every time you add one more qubit, you have the potential of doubling the amount of memory and the amount of computational power that that device has. For small numbers, I have one qubit that has two pieces of information. I have two qubits that has four pieces of information, but then it starts to get interesting when they have three qubits, I doubled in so I have eight pieces, and I go to 16, and 32, and 64, 128. By the time I have just 10 qubits, I’m already up to 1,024 pieces of information. This is truly exponential and I will tell you, I don’t know if it’s an American thing but I certainly see it’s commonly used. People use this word exponential just to be really fast, just really fast. Lo and behold, exponential involves exponents and so it’s two to the number of qubit.
TK: Okay. So, at this point you’re asking yourself, what in the world does that mean? I’ve got a lot of qubits. They exponentially increase in terms of their power and the amount of data they can store, but how does that help me as opposed to a classical computer because we still have the ability with classical computers to store enormous amounts of data. Here’s the analogy Bob used about how a simple caffeine molecule helps you understand the power, the true power of quantum computing.
Bob: If we were to model caffeine inside a computer, and so by modeling, I’m saying we have a representation of this molecule that is inside the computer that we can manipulate just as accurately as we might in a test tube, in a beaker in a lab, so it’s complete. How much information would that take? Caffeine is not a very large molecule. It turns out, even for part of this information, the energy that’s used to kind of keep all the electrons together and the atoms from flying apart is tremendous amount of information, and that number is roughly 10 to the 48 bits, zeroes and ones, so that’s a one with 48 zeroes. Now that sounds like a big number but how big a number? A scientist estimates that the number of atoms in the earth is between 10 to the 49th and 10 to the 50th. This means that the number of bits, the number of zeroes and ones just to represent one caffeine molecule at one instant could be comparable to one to 10% of all the atoms in our planet.
TK: If you’re like me, at this point you’re asking, how is that even possible? Bob had an answer for that as well, and you know when you think about it, it makes sense.
Bob: How does nature do that? How does this little caffeine mole— one molecule at one instant. It does it all the time. I mean in your coffee cup or tea, there are trillions of these and they’re all using this much information. Some nature in practice harnesses much, much more information that we can possibly imagine. The answer as to how, it’s more philosophical than it is physical. There are different interpretations that have gone back a hundred years. In fact, people like Albert Einstein used to argue with other physicists on the front pages of international newspapers - this was the interest in it. You’d never see this today, but this argument of how this would work. We know what the math says, right? We know how the math says and then we try to translate that into physical computers and that’s what we mean by quantum computers, which are physical devices that come close to behaving like these artificial quantum particles individually and then working together and it’s when you have multiple qubits that you get this extraordinary growth and this power to address these exponential problems.
TK: If you follow that last bit of commentary from Bob, it becomes somewhat obvious that if nature had a computer, it would be a quantum computer because fundamentally, the real world, the world that we live in not the digital world, is in many ways already dealing with enormous amounts of data at that atomic and sub-atomic level. It’s interesting to think about some of Bob’s comments though because Elon Musk at one point had postulated that perhaps, we live in a simulation and a lot of folks are sort of toyed with that idea. I would submit that if we do live in a simulation, it probably is a quantum computer that is simulating what’s going on today but let’s bring it back down to earth. We asked Bob to give us some practical applications of quantum computing today. Here’s what he had to say about that.
Bob: These quantum computers have to get sufficiently powerful enough to tackle problems and ways that we can’t already do classically to the extent that they are only doing what we can already do. We say, “That’s interesting,” and we can say they’re evolving, they’re getting better but real usefulness is when we get to a point where we can say, “Right over here in this huge case, by using quantum computing, it’s significantly better than classical.” I don’t mean twice as fast because frankly just run it twice as long classically. No big deal. I mean thousands of times faster. I mean, that you can do things that you really can’t do at all classically. This point when we can start seeing this extraordinary differentiation is what we call quantum advantage.
TK: Let’s take a few seconds to talk about this term quantum advantage that Bob brings up. There’s another term, quantum supremacy, and they both speak the same threshold where finally quantum computers have equaled and then exceed the ability and the power of classical computers. Most of us know that Moore’s Law has governed the growth of computing up until this point. Moore’s Law simply says that the density of transistors on a chip doubles approximately every two years. There’s now a counterpart or maybe it will be a displacement of Moore’s Law called Neven’s Law. Hartmut Neven is the director of Google’s Quantum AI lab. Neven’s Law basically tells us that not only is quantum computing increasing in its power exponentially, but it double exponential rates. Now what that means simply put is that there is no analog in nature for the kind of growth that we’re seeing in the power of quantum computing. The reality is according to Neven and his counterpart at Google, we may very well be just years or months away from this point of quantum supremacy. Again, all bits are off when it comes to quantum computing. We just don’t know how much impact it will have. If you want an analogy of some sort, think about the transistor. That radically changed civilization in a way that no other invention had up until that point in time. We may be at the same precipice now with quantum computing but with orders of magnitude greater implication on society. Now, Bob was a bit more conservative when it comes to the timeframe. Here’s what he had to say about how IBM sees the evolution of quantum computers reaching that point of quantum advantage or quantum supremacy.
Bob: When will we start seeing quantum advantage and in what areas? What sorts of use cases? The general what we feel is a somewhat conservative answer just in terms of the timeline is that if we, IBM, because we know we can only speak for ourselves, keep improving the devices and the systems the way we are, we expect to see this quantum advantage within 10 years. We hope in fact because scientists and engineers are amazingly clever and they surprise us all the time. We hope to really see this in three to five years. Now, where will see this?
From my example before with caffeine, one area is chemistry. In the short-term, it’s likely to be things in like materials discovery, developing new alloys, right? Materials tend to have particularly simple structures, repeatable structures, so we think maybe we’ll see something there. Much longer term and I’m very careful. I never say quantum computing will do this. I will say it may do this if everything works out, it may do this. Drug discovery, of course, you know everybody would love to be able to compute the drugs we need versus discover the drugs. Discovery kind of sounds like you’re wandering around looking, right? Well, it’s actually a very targeted type of search but if we can actually simulate these pharmaceutical medicines inside a computer and simulate accurately the way they would interact with you, by doing it inside a computer instead of inside you, it’s a whole lot safer and should be a whole lot faster. That is 10 to 15 or so years out. Now, there are a couple of other areas that are of interest.
One is artificial intelligence, AI is a very broad field. Almost everywhere you look, somebody is using some part of AI, machine learning, deep learning, something. So, it’s not one industry, it’s where are we applying AI and quantum speeds it up. We have some hopeful early results and really, we’re looking at a couple of different possibilities here. One is to say this AI stuff is kind of deep. It’s all really math. Can we use quantum computing to speed up that math?
On the other hand we’re saying, “A traditional way of - traditional algorithm of doing something AI operates this way, is there a completely different algorithm for quantum that solves the same problem but in a way that it takes advantage of these quantum properties?” When I was talking about multiple qubits, there’s this notion of entanglements of how you really tie them together. Is there something about entangling multiple qubits in these very high dimensional spaces that will allow us to find patterns much better? The final general area is in what we call financial services. Let’s say you have a large portfolio, just some sort of financial portfolio. Let’s say you’re a hedge fund. Tom, you had a hedge fund now, okay? You would very much like to accurately assess the risk of your portfolio at any given time because that would guide you and your buying and selling decisions, right? Well, not everybody has a hedge fund, but many people have retirement funds, 401Ks and things like this. So, we may see quantum computing affecting people through better investments for their retirement. They won’t necessarily know they’re using quantum computing but in five, 10 years or so, this technology may be able to be applied in that way.
Wasabi Hot Cloud storage is a proud sponsor of Foresight Radio and their mission to help rethink the future of how we work, live and play. At Wasabi, we’re helping you take control of a future where your ability to affordably store and leverage data will determine success or failure. Our low-cost, high-speed, fully secure cloud storage blows away Amazon, Google, or Microsoft, and there are no hidden fees for egress or API requests. See for yourself with a free trial at wasabi.com.
TK: Now back to my conversation with Bob. If you follow what Bob is saying, it sounds like quantum computing is going to displace traditional digital computers but that’s not at all the case. In fact, what Bob saw was something very different.
Bob: Really the future is going to be a hybrid of classical and quantum. You’re going to see quantum computers working with classical computers. You’re going to see quantum algorithms tied in very closely with classical algorithms. I’ll even go so far as to make a guess. Let’s say, let’s go 20 years in the future and let’s imagine quantum computing is doing really well. Let’s look at a typical business application involving, let’s say, data and transactions and user-interfaces and all these things that you can imagine, but we have the power of quantum computing for the really harry types of computations. Twenty years from now, that application might still be 95% classical with only 5% of it quantum.
TK: If you’re anything like me at this point, you’re just dying to get your hands on the quantum computer. While the most powerful of these are limited to use within laboratory environments, IBM has made quantum computing available to the general public. That is, if you want to take the time to learn how to program one of these. Here’s what Bob had to say about the efforts IBM has made to make quantum computing accessible to pretty anyone who has the ambition and the time to learn how to use one.
Bob: About three years ago, so May 4th 2016 to be precise, we put up on the web a five qubit quantum computer for free, no charge. We basically said, “You want to use a quantum computer, here it is, come and play.” Since that time, we have also put up a 16-qubit machine there. We do have more powerful machines for our commercial ecosystem. So, companies like JPMorgan Chase, ExxonMobil, Dyler and Mercedes Benz, they use the advanced machines. For the general population, there’s this no charge IBM Q experience, and when we first put this up to build what we hoped would be this ecosystem as we would attract people and see who they were, we had no idea. Was this going to be 12 people? Was it going to be 100 people? Would they do anything?
Now, close to three years later, we have over 100,000 users of this, so a lot more people know about quantum than what many people suspect. They’ve run over 9.5 million computations on these quantum computers, these free quantum computers. What’s also interesting is when it comes to research, the people who are going to be advancing this field, we’ve had 175 scientific papers written by using this free quantum computers that weren’t done by us. I mean, we read a lot of papers but these were outside people who were able to use our computers and innovate on top of them, at no cost to themselves. We also have open-source software; we call it Qiskit, Q-I-S K-I-T. It’s available on GitHub. Once again, we’re trying to reduce the friction as much as possible. If you want to learn how to program with quantum, we will give you access to the software, the full stack of software, the computers. There are simulators you can run on your laptop that are consistent with everything else. We have videos, all those sorts of things, user manuals. We have over 50 videos on YouTube at this point to try to pull people in, and the goal here, going back to the question you asked before, when will we get the use cases that do something interesting? It’s going to take a little while for people to learn how to program and think in a quantum way. They have to get started. We have to make it easy for them to do this.
One group I’m really excited about are the students. If you’re in college right now, even if you’re sort of toward the end of high school, three to five years means you will likely be in the job market. So, if we can train people, if we can have these people who are very early in their careers, start to think about problems in the way that you must do in a quantum way of programming for software development. They will come up with extraordinary things and that’s how we’re building the ecosystem where this friction free aspect and access, open access, open source, and we hope good things will happen.
TK: The last things I talked to Bob about was the actual device, the machine. IBM Q is an incredibly elegant computer. If you’ve never seen one, Google it, take a look. All of that device, the majority of it is intended to super cool the qubits which rest at the very bottom of what looks like an upside down chandelier. The beauty of the machine speaks to something far deeper when it comes to quantum computing and it’s something that you hear in the voices, in the attitudes of people that work in the field. It’s this mash-up of computing, science, art and philosophy. Here’s how Bob described it.
Bob: There is this nature, this notion in science of literally elegance, and so on the math side, you want to prove something and you say, “Wow. That’s just exactly right.” In the hardware side too, there’s this combined, and this combination of functionality, ultimately it has to be a quantum computer but elegant design as well, and I think people are preaching more this merger, great functionality, great design and we’re certainly trying to do that with IBM Q.
TK: That was Bob on quantum computing. To find out more about Bob and IBM’s Q quantum computer just check out the links on the Foresight Radio homepage at foresightradio.com. In the next episode of Foresight Radio, we’ll delve deeper into the way quantum computers work. We’ll look at the notion of entanglement, something that even Einstein had difficulty accepting as we talk with scientists who are paving the way for the future of quantum computing. Thanks again to our sponsors for this episode of Foresight Radio, Wasabi. Take a look at how Wasabi is changing the rules of the game for cloud storage at wasabi.com, and thank you for listening. If you enjoyed this podcast, please be sure to subscribe to Foresight Radio and to share it with your friends and colleagues. This is Tom Koulopoulos, I look forward to joining you again soon for another episode of Foresight Radio where we explore the future of how we will live, work, and play in the 21st century.