Online science dialogue in Spanish "La computación cuántica y su impacto".

Summary:

On 30 March 2022, the Rafael del Pino Foundation organised the dialogue "Quantum computing and its impact on our society", with the participation of Ignacio Cirac, director of the theoretical division of the Max Planck Institute for Quantum Optics in Munich; Marcos Allende López, specialist in quantum technologies, blockchain and digital assets in Washington DC, and Antonio D. Córcoles, expert in experimental quantum computing at IBM, Yorktown Heights, New York.

Ignacio Cirac: We are living in a very special moment. Those ideas that many people had in the 1990s and which seemed unrealisable, can now be realised. They are being done in laboratories, in companies, in industry, and we are already able to build some small prototypes of quantum computers and do some computations with them.

Quantum computing is part of quantum information theory, which brings together two scientific and technological revolutions that took place in the last century. On the one hand, the discovery of quantum physics, which describes the microscopic world of atoms, molecules, photons and electrons. On the other hand, information theory, which tells us that information is something that can be measured, compressed, corrected, and which has given rise to many applications. Quantum physics is the basis of electronics and electronics is the basis of much of the equipment we enjoy. Information theory allows us to use today's means of communication. If we put these two revolutions together, it gives rise to quantum information theory, which has applications in computing, communication or metrology and allows us to use quantum physics to send and process information.

Quantum physics describes the world of the very small. A quantum computer is a computer that works according to the laws of quantum physics. Quantum physics is an old theory and we have already exploited it in electronic systems, but it has aspects that we have not exploited because we did not have access to them until recently. These are the ones that are related to the most mysterious facts of quantum physics that are represented, for example, by Schrödinger's cat. If we could extrapolate microscopic principles to the macroscopic world, quantum physics would tell us that this cat is in a superposition, that it could be with a physical property that is not well defined. The physical property could be to be alive or to be dead, and it could be alive or dead. Schrödinger's paradox also tells us that if we extrapolate quantum physics to the macroscopic world and try to understand it, it would tell us that we could have a box in which we would have this superposition, a cat that has not yet defined how it is, and only when we open the box is the state of that object defined, in this case whether the cat is alive or dead and that occurs with a certain probability. Quantum physics tells us, in short, that the objects in this microscopic world do not have defined properties and only become defined when we observe them.

This is not only true for cats, but also for other physical properties, such as magnetisation. If we have a magnet, we can have it with the north pole pointing upwards or pointing downwards. According to quantum physics, the magnet could have both positions at the same time. In the case of a circuit we could have the currents going clockwise or counterclockwise. According to quantum physics we could have a circuit in which we have a superposition having the two superpositions at the same time and only when we observe the position of the north pole or the direction of the current in the circuit is defined.

In the microscopic world we can use this kind of phenomena, in atoms, which are like little magnets, which can have the north pole pointing up or down, with the spin up or down. In the case of circuits, we can also miniaturise it and have a very small microscopic circuit, formed by superconducting currents that move either clockwise or counterclockwise. In these microscopic systems we can observe this superposition, we can have both properties at the same time. This is what in quantum physics we describe with formulas, assigning for example the value zero when the spin of the atom is up and the value 1 to the spin down. This means that we have a state in which we have a superposition, which is characterised by two numbers, alpha and beta, which tell us whether when we observe the object, we will find it more or less likely to point downwards. In short, in the microscopic world we have access to these laws of quantum physics, which tell us that we can use superpositions, and it is only when we observe that the properties of these objects, in this case atoms or superconductors, are defined.

What is more interesting is that this is true even if we have many objects. We can have many atoms, or many superconducting circuits. We can have all atoms with spin down, in the state we call zero, zero, zero, zero, or in the state with all spins down and one up. There is an exponential number of exponential configurations of the total number of atoms. If we have a microscopic system made up of these atoms, or these semiconductors, we can have superposition of an exponential number of configurations. We can have all the atoms at zero, or some at zero and some at one, and because there is an exponential number, there is an exponential number of configurations. This is what gives the computing power in quantum computers. Each of these states works in parallel when we do some operation. It's as if we have a very large number of computers doing operations at the same time, and this is what gives the computing power.

When we do the computation with a quantum computer, we have to measure, and if we measure, we destroy the superposition according to the laws of quantum physics. So just having these superpositions is not enough. In order to have a quantum computer, you have to have these superpositions somehow interfere with each other in such a way that at the end you don't have a superposition but the result of the computation that you want to do.

You could say that what a quantum computer does is that it processes information, zeros and ones, in terms of superpositions, that is, using the laws of quantum physics, and, with that, it has other options and can solve problems in a more efficient way, much faster, than the biggest supercomputers we can have, which do not work with these superposition rules.

What are the applications and current status of quantum computers? The shift from today's electronic computers to the quantum computer is an essential change. It's not a change that something goes faster. It's that we do it in a different way and, therefore, it has an enormous potential. It's a bit like the change from the abacus to electronic computers, in which we went from using the laws of mechanics to the laws of electronics, and that gave us a lot more power. Now, with the laws of quantum physics, we have much more power.

Quantum computers could have applications in various segments of industry. So we think that quantum computers will have a big impact on industry and society. The question is when this is going to happen, because we already have the first prototypes of quantum computers, so, in principle, one would think that we can already apply them to all these interesting problems. However, this is not the case at the moment. The reality is that these quantum computers operate in very extreme conditions, either at very low temperatures or in a vacuum. We have to make them interact with virtually nothing, and that makes it difficult to operate them and get the full benefits out of them.

That puts us in the same situation with quantum computers as we were with computers eighty or ninety years ago, those early computers that were not capable of many calculations and took up entire buildings. The situation, now, is a bit similar. We have the first prototypes, which are already able to show that they work well, that they have some capability beyond the classical computers, but they are still not as good as we would like them to be. In fact, the fundamental problem with quantum computers to be able to use them the way we want to use them is errors. When we have these quantum bits in these superpositions, we cannot observe them because the superpositions disappear, but this happens not only if we observe them but also if they interact with any other object. So, if we don't completely isolate our system, these quantum bits are somehow erased and you get errors in the computation. So we have two options. Either we learn to live with these errors, that is, even though they have errors, we may have advantages. That's what we call noisy computers, or NISQ. Or we correct the errors. That is, it is possible to use more quantum bits and every time errors occur we correct them. This is what we call scalable computers.

Both options are being pursued. It is just that one of them may take a long time. The first prototypes of quantum computers date back to 1996. The industry entered quantum computing around 2012. In 2019 there was an experiment that showed that a quantum computer, albeit with errors, is capable of faster computation than a supercomputer. We have entered the NISQ era, where we have these prototypes, which we are trying to make better, but they are still buggy and not quite powerful. Later on, and we don't know when, the scalable ones will come and this will be the time when we will be able to enjoy all the advantages of quantum computers. In a way we have just entered the NISQ era of these buggy computers, and waiting for scalable computers will probably take much more time, much more effort and much more advancement of the technology.

In terms of applications, there are many that quantum computers can give us. They are problems that they can solve much more efficiently than a classical computer or a supercomputer. There are some specific problems, for example, related to the encryption of secret messages. There are other problems that are optimisation problems, such as the problem of the traveller who has to go through several cities and have the shortest route, which is the shortest way. There are many optimisation problems that quantum computers can improve. There are problems related to data analysis, to artificial intelligence, to machine learning. And there are other problems that are related to simulation, that is, solving chemical problems, or material problems, or even physical problems, that classical computers cannot solve. These problems cannot be solved by classical computers because these systems that we want to simulate, these chemical compounds or these materials, themselves comply with the laws of quantum physics and the physical properties are encapsulated in the superpositions. And since there is an exponential number of configurations, in order to characterise the physical properties we have to compute an exponential number of coefficients of these superpositions. This makes it almost impossible to solve with classical computers. It is well known that problems that arise in chemistry, such as chemical reactions or composition, or the spatial structure of chemicals, or in materials, or in super-energy physics such as the solution of the standard model equations, are all very difficult for supercomputers. However, this is where quantum computers can help, and it seems that these problems can be solved more efficiently, even in the NISQ era. What remains to be seen is whether the other problems can also be solved more efficiently in the NISQ era or whether we will have to wait for scalable computers.

Quantum computing is a form of computing that uses the laws of quantum physics and we are currently experiencing an extraordinary moment. Because of the technological progress that has taken place over the last twenty-five years we are able to build these quantum computers and do demonstrations that go beyond pure trial and error experimentation. They are already able to challenge supercomputers in some kind of computation.

In the long term, quantum computers will have a huge impact because there are many problems we know that quantum computers can solve faster, more efficiently, than classical computers, than computers that do not use these superpositions. In the meantime, until scalable computers exist, research is being done to find specific applications, especially in optimisation problems, data processing problems, or simulation problems. But the most important applications are those that we have not yet discovered. It has happened in any technological or scientific revolution that we have witnessed, that the objectives that were set at the beginning changed over time and resulted in a series of applications that were practically impossible to predict. In the field of quantum computing something similar is going to happen.

Marcos Allende López: Today, it is very difficult to make predictions about how quantum computers will be used. In the end, we are still developing the technology. Once it is implemented, any technology has a much greater impact than expected. In the 1960s, an IBM executive said that four or five computers would cover all of Earth's demand for classical computing, and look where we are today, with almost everyone having one or more in their homes. When quantum computers reach their expected potential, we will be able to perform very complex simulations and calculations in many cases. These include medicine, materials physics, finance, social sciences and many other areas that will allow us to test extremely interesting and useful models. For example, in medicine, biology, genetics or pharmacology there are many applications that are already foreseen, such as the ability to simulate drugs with a computer, which is something we cannot do today. Or the possibility of designing a drug and interacting with a pathogen to see if this really works. The ability to work with computers capable of doing these kinds of calculations is going to have a tremendous impact on the development of all these areas. In biology or genetics, we could also simulate very interesting models that could provide very useful information about our evolutionary history as a species, or that of other species, or even give us the possibility of simulating how to modify our DNA and become a better species, and I put "better" because here we get into a series of ethical issues. In the social sciences, simulations of human behaviour at the social level could also be carried out. Even in finance, applications could be made, for example, to simulate models of portfolio evolution or the evolution of financial environments. In materials physics, new substances could be developed for different industries, such as energy. We are still waiting for a tipping point where quantum computers are proposed that are scalable and allow us to go to that quantum computer in a linear way. The moment those tipping points are reached by the major entities that are developing this hardware, we will be able to make a more certain prediction. We'll see if it can be ten years or less.

Antonio D. Córcoles: I agree with everything that Marcos mentioned, particularly the simulation of materials in many different fields, such as medicine, agriculture, energy. All of that has a lot of potential and very rapid progress is being made, both experimentally and theoretically, for the algorithms that are needed to transform those calculations. There is going to be a series of almost continuous progress, so the question of whether we are ten years away is never going to be answered because we are always going to be dragging that progress along. The progress we have made in theory and in experiments in the last ten years is quite spectacular, but there is a long way to go. From that point of view, the consideration of noisy computers can teach us a lot about how to attack these problems until we get to scalable ones.

Ignacio Cirac: With these NISQs there are already some things we can do in the next two years that we can learn from. Maybe it is not as spectacular as those industrial problems, like medicine or the environment, but it has a lot of value, at least in the long term. In the simulation of materials, of scientific systems, not only from the scientific point of view, but also from the point of view of some industrial applications, it turns out that when we have a lot of objects and these objects obey the laws of quantum physics, we cannot solve them, we cannot make predictions with them on supercomputers. However, quantum computers have a very special advantage there. If they already have advantages for the problems we have mentioned, for these they have many, many more advantages that are double exponential. So, even if there are errors there, these very big advantages can compensate for the fact that there are errors. I hope that, in the next two or three years, problems are going to be solved, not just academic problems, but problems that will teach us something about these many-particle systems and allow us to develop new theoretical technologies. As a physicist, when I have one of these problems, when I want to describe superconductivity, I can't do it, we don't have the techniques. These first-generation quantum computers are going to teach us not only how to solve them, but to develop the techniques to be able to describe them. Once we have those techniques, even though they may be classical, that will have applications. Today, one of the most talked-about near-term applications is the generation of certified random numbers. These random numbers are not only important in the lottery, but also in many aspects of physics and even industry. It is very difficult, if not impossible, to certify that a number is random. When someone draws a number from the lottery, how do we know that it was not pre-destined? It is impossible to know. However, with a quantum computer, you can create random numbers and certify, know that they have not existed before and that they are truly random. This is something that can work already with these first generations. Of course, they don't solve a problem in medicine, or something like that, but it's a problem that we can't do classically. From there, people can already think about what we can use random numbers for that are certified. There are people who think that, if we had certified numbers, in blockchain we wouldn't need so much data mining, we could improve energy consumption. In other words, there are many things that are connected to each other and a discovery in quantum computing can be a surprise and can go in any direction.

We are at a very special moment. There are few times in the history of technology when we find something like this. We already have a computer that is capable of doing something that we can't do with classical computers. That is, if there is a problem, we already know that we can do something with those computers, even with their bugs, which are more powerful. And we have an idea of a couple of things that they can be more powerful for. That's an undeniable fact. The question is whether there are other problems that they can be used for. That's where we have to test because it's very likely that, by testing and testing, things will appear that we can't imagine. On the other hand, there is great enthusiasm, especially in the media, about quantum computers, saying that the computers we are building today are going to solve humanity's problems, but this is not true and, moreover, there is little evidence. From this point of view, one is no longer so optimistic. We know some things where there might be possibilities, but we also know many other things where, just by doing some calculations, you can see that they are not going to work.

Antonio D. Córcoles: He visualised the development of quantum computers as a kind of dial in which, as research progresses, the noise can be reduced. As we learn how to design the architecture for error correction to get to scalable computers, that evolution is going to be quite big. The interesting thing about the noise systems we are working on now, which are small in size but quite respectable, is learning about noise. That's what's going to give us the next evolution. Learning about noise may seem to apply only to noisy computers, but there are techniques that we can extract to apply that kind of learning to what are going to be scalable computers. We have quantum computers that you can access in the cloud and you can study the noise of these systems as a project. Once you have the hardware, people are going to play with it and define new problems and find applications. It all goes together in progress. We have an evolution that doesn't just happen in the lab, it goes globally.

Marcos Allende López: One example I read about last year is that Roche and Cambridge Quantum Computing were with an NISQ trying to develop a potential cure for Alzheimer's disease. The strategy that should be followed depends very much on the resources available. Trying to develop a scalable quantum computer is not affordable for everyone because it requires an investment of hundreds of millions. Companies like IBM and others have already invested so much money that it is difficult to recommend someone to start from scratch trying to get into such a race. But there are many other applications where there are opportunities for everybody. For example, there are people who are focusing on software, to develop programs that work on these computers. This also has its limitations, because developing programs for computers that do not yet exist is complex. Then there are some first applications that are working because they are a necessity. In the end, one of the things that quantum computing brings is a cybersecurity threat, against everything that is encryption of information at a distance, which we have today in a globalised world that uses algorithms that are going to be hackable with quantum computers. Techniques are beginning to be developed that are based on the generation of random numbers, which allow both the construction of new asymmetric algorithms and symmetric algorithms, in which the generation of keys is greater. This is very important in blockchain.

Different Spanish entrepreneurs are doing a great job developing different proposals. The ecosystem is awakening in Spain. There is more and more interest, dialogue, dissemination and funding. In the academic world, there was a great deal of rejection in those more theoretical degrees that business could be introduced due to a fear that business could contaminate the teaching plan. By preventing companies from entering, it was possible to have a more purist vision of being able to carry out a more fundamentalist study of theoretical physics. If more alliances could be established between universities and companies, even in theoretical physics, it would be very useful for everyone.

Antonio D. Córcoles: Developing quantum computers is something that costs a lot. It is very difficult to compete starting from scratch. Academia should try to see how it and business can add a bit of synergy to what they do together. Part of the effort at IBM is to try to collaborate with academia, in Spain and outside Spain, to help the research itself and to provide a basis for students to choose a future career related to their studies in a field other than academia, but without trying to define their line of work. In Spain there are many opportunities in software because hardware development requires a very big effort and a lot of financial muscle. There are many small start-ups that are working not only to develop software for algorithms, but also software to control cubits, to control doors better and better, and apply it to hardware that already exists in other companies. That is something that is accessible to small companies.

Ignacio Cirac: When a new theory emerged, people always said that Spain should do the theory, that it should make software, but you want Spain to have technology. Things seem to be changing, at a slow pace compared to other countries, but they are changing. Not in quantum computing, but Spain is a power in quantum communication. There are European networks led by Spanish groups, we have Spanish industry working, developing, implementing quantum communication systems and there they are leading from the European point of view. If you think that the German government has given two billion for quantum computing and the Spanish government has given forty, fifty, sixty, eighty million, there is a difference, but at least they have given something to start with. Quantum computing is something very long-term and if we don't start now and it takes twenty years, we are going to stay there. So it is very positive that it has gone in the direction of promoting more experimental research in quantum communication. I hope that companies will also come in, not only in communication but also in computing. It is worrying that there is not a stronger connection between academia and business in a field like theoretical physics, but it is even more worrying that it does not exist in engineering. Finally, quantum computing is something that breaks some paradigms. Many people in academia are involved in start-ups. In many countries around the world, master's degrees for quantum scientists are being created. In Germany it has been done for a couple of years now, but also in Spain. There is the intention and there are people who support it.

The Rafael del Pino Foundation is not responsible for the comments, opinions or statements made by the people who participate in its activities and which are expressed as a result of their inalienable right to freedom of expression and under their sole responsibility. The contents included in the summary of this conference, written for the Rafael del Pino Foundation by Professor Emilio González, are the result of the debates held at the meeting held for this purpose at the Foundation and are the responsibility of the authors.

The Rafael del Pino Foundation is not responsible for any comments, opinions or statements made by third parties. In this respect, the FRP is not obliged to monitor the views expressed by such third parties who participate in its activities and which are expressed as a result of their inalienable right to freedom of expression and under their own responsibility. The contents included in the summary of this conference, written for the Rafael del Pino Foundation by Professor Emilio J. González, are the result of the discussions that took place during the conference organised for this purpose at the Foundation and are the sole responsibility of its authors.

• Facebook
• X
• LinkedIn