Quantum information science will contribute to one of the next revolutions in computing, but realizing that promise will stretch our scientific understanding and technological skills to their limits. We sat down with NIST Physical Measurement Laboratory Acting Director Carl Williams to get his take on the potential and challenges of this disruptive new discipline.
EDITOR'S NOTE: Since the publication of this Q&A in 2018, Carl Williams has retired from the federal government.
CJW: Quantum information science (QIS) is really the merger of two of the great scientific quests of the 20th century: quantum mechanics — the foundation of everything that works on the subatomic level — and information and computation theory. In the latter part of the 20th century, we realized that these two were intimately connected. QIS uses the stranger properties of quantum mechanics — its probabilistic nature, superposition, entanglement — to transmit, compute and move information. We have already been using quantum mechanics, albeit at a semi-classical level, to build some of the defining technologies of the 20th century, like the transistor and the laser. QIS is really just the next step along that road.
CJW: QIS is a revolutionary technology. It allows you to do things in a totally different way. It really requires intimate control of the quantum world to build the technology. As such, it is likely to be the foundation of a lot of future technologies, and therefore, extraordinarily important for our economic and national security.
I think that it will be on par with what we’re beginning to see today with precision medicine — the fact that we can take a person’s genome and develop treatments that are tailored to that individual. It’s going to cause a major change in how we do things and give rise to a lot of new possibilities.
And I think this is, in fact, why a lot of businesses are now investing in QIS. They’re not investing because they believe there’s going to be a product three years down the line. They are investing because they need to understand how this is going to affect their bottom line in a decade.
CJW: NIST is a measurement institute, it’s what we do. Measurement is the key to doing anything, and knowing how, at the highest level, you do it and do it well, that is, accurately, precisely, and reproducibly. The quantum world and QIS represent the ultimate in precision control. They are an incredible measurement challenge.
In many ways, NIST is ideally positioned to contribute to QIS, in part because our atomic clocks essentially have the computational power of one quantum bit, or qubit. And voltage standards are based on Josephson junctions, which are one of the technologies that are being used and explored for QIS applications. Though QIS has only been a formal established program at NIST since 2000, we have been manipulating and using it for voltage metrology for more than 25 years. So, we’ve been thinking about how this can impact the future for a long time, and we’ve developed a lot of expertise — everything from developing very sensitive single-photon detectors to prototype quantum computing devices to figuring out how you benchmark those devices to numerous other applications.
CJW: While other agencies have their application spaces in defense, intelligence, national security, computing, and other areas of basic research, NIST’s role is to provide the ground truth and the foundation upon which all characterization measurements can be built. This includes defining the base capabilities of a quantum computer, and tests such as a suite of algorithms upon which, if somebody’s machine can solve these problems efficiently, demonstrates a capability beyond what we currently believe to be classically possible.
With regard to industry, NIST is often seen as a convener. We are hoping to establish a consortium with industry in quantum information science, called the Quantum Economic Development Consortium. It’s important early on that we create a broad quantum ecosystem within the U.S., and that means understanding where industry thinks this technology might go, even though they’re still trying to figure that out. Feedback is essential to making wise decisions. This consortium that we are in the process of setting up will allow public-private partnership so that we can collectively determine where our gaps are, what investments we should be making, what opportunities we might be missing, what things we need to do collectively to support the future quantum industry.
In addition to that, we will be there to support industry’s measurement needs and to transfer our knowledge to them as they work to develop their technologies and bring them to market. We’re not here to pick winners and losers. We’re here to help create a level playing field upon which companies can compete. And we look forward to working with all of them and may the best ones win.
CJW: There are roadblocks to the enabling technology. For instance, we need everything from better RF electronics and stabilized lasers to dilution refrigerators and cryogenic systems. Mastering those key things will enable us to handle more complicated systems. So, the first thing we need to do is build a better technology foundation for manipulating multiple quantum systems. Over the past 20 years, we’ve progressed from controlling two or three ions in an ion trap to routinely controlling tens of these things simultaneously.
To make these systems more robust, we need to be able to shield them from outside interference. That either means very low-temperature cryogenic devices or vacuum systems that basically isolate atoms or ions. We need to be able to detect individual spins, detect individual electrons, and detect individual photons.
So, the range of technologies and components we need to do all this is really quite remarkable. And, in fact, what we have been doing in the past 10 or 15 years has really developed those core pieces. But it’s one thing to have the core tech housed in a big complex instrument and it’s another thing to have it integrated into something that looks like a smartphone. We’re some ways away from having a system that you don’t need a bunch of physicists to run, but we’re getting there.
CJW: A quantum computer is never going to replace a classical computer. There is no such thing as doing word processing on a quantum computer; you just get gibberish. It’s not going to replace it.
Richard Feynman — probably one of the first people to really envision quantum information science — basically realized that classical computers do not efficiently solve quantum problems. You can do drug design, even simulating quantum physics all on a classical computer, but if you want to just add one more molecule, atom, nuclei or electron, you need exponentially more powerful computers. A quantum computer can solve these and other specialized problems far more efficiently than a classical computer. In fact, a quantum computer can solve some problems that, in principle, the classical computer could never solve, or at least never solve in a reasonable amount of time. So, quantum computers are going to complement classical computers. In fact, the control system for my quantum computer is going to be a very complicated classical computer, so they’ll always be around.
CJW: If you have an https website, you are protected by public-key infrastructure. One of the things that a quantum computer would take away are the algorithms that we currently use to protect that infrastructure. If a quantum computer were to be built, then banking, secure transactions, literally, all of e-commerce would be at risk of being hacked. NIST and others are already involved in a massive effort called post-quantum cryptography or quantum-resistant cryptography to come up with new algorithms that will be resistant to quantum attacks.
But let me de-hype it just a little. The first quantum computer is not the risk because many people will have access to it and one can monitor that it’s not being used for this purpose. It’s when a quantum computer can be had relatively easily that this becomes a risk. By then, we have to have replaced all the existing public-key infrastructure. But replacing that infrastructure will take a decade.
CJW: Until the past few years, the U.S. did reasonably well in producing a quantum workforce. There were a few problems between the different subject areas that contribute to QIS, but we produced plenty of quantum physicists who did quantum information science. We also produced a good number of quantum computer scientists, but most of them did not end up with jobs in academia and elsewhere. That is beginning to be addressed. There is a need for computer scientists and quantum engineers, because physicists don’t build smartphones.
But as industry moves into the field, they’re putting a strain on our existing workforce. They’re hiring people faster than we can train them. How many do we need to produce? How do we give them quantum intuition? We need people who have the intuition. We all know that our children are much better at using smartphones than we are because they’re growing up with and playing with the technology. Well, we need that kind of quantum intuition, that same level of familiarity, to build our QIS workforce. We need to have more people at the undergraduate level who have an understanding of the difference between quantum and classical intuition. And that means changing our education system, ensuring that it’s multidisciplinary and creating a workforce that can contribute to this the same way that we had to create a workforce that could deal with integrated circuits and eventually with smartphones and the internet. And that workforce is going to take some time to develop. It’s a revolution in its own way.
CJW: In order to support science, we increasingly have to sell science. That leads to a certain amount of over-hyping, and almost everybody does this. Having said that, when I look at where this field is going, I think that, given what we understand about physics and the laws of physics and the universe, QIS probably will be one of the most remarkable technologies, with some of the most far-reaching applications, that has ever been developed. Do I know what the most important application is? No, don’t have a clue. This would be like going back to 1950 and predicting how many computers we would need in 1990. Or what we would be using computers for. It would be like predicting that we would touch the internet almost continuously every day when it didn’t exist, or back when it was only an academic toy in the mid-1980s. So, if you pick up a paper and it tells you something about quantum technology or QIS in the future, and it’s making a promise, it’s probably a little over-hyped because I don’t think any of us are visionary enough to truly understand the implications of what’s coming.
A very interesting article. The insights are really helpful and informative. Thanks for posting.
Nice article post
Thanks for sharing admin