Quantum computing may have some tech analysts fretting over the threat they pose to digital security, but properly scaled and managed, it could also pave the way to better healthcare and transport, as well as greener energy sources. Cybernews sat down with one expert to get a clearer picture of quantum computing’s past and future.
Quantum computing is something you might have increasingly heard about if you follow the tech world. Though they are not yet fully scaled and efficient, or “fault tolerant,” rapid progress is being made, with the National Institute of Science and Technology (NIST) declaring that the post-quantum cryptography standard designed to protect the new world of computing will be finalized in two years.
That said, NIST has already run into trouble, with one of its supposedly quantum-proof protocols being hacked in less than an hour using a conventional computer and another being breached over a weekend using a single laptop. Though some cybersecurity experts think this may not be such a bad thing in the long run, others believe that the vastly improved quantum computers could pose an existential threat to the global digital system that the human race has come to depend upon.
Jack Hidary, the CEO of Sandbox AQ, a spin-off company of Google owner Alphabet, has been working with quantum computer scientists for years and is not oblivious to these pitfalls. But he is also optimistic that if properly managed, our quantum future is bright. He should know – his company was among just a dozen in the world appointed by NIST to its National Cybersecurity Center of Excellence to develop post-quantum computer migration strategies.
Not only is Hidary confident that the fault-tolerant computers of the near future will advance scientific research into transport, healthcare, and energy, he also believes the new industries generated by this development will create an abundance of career opportunities – and not just for the tech-savvy.
The Cybernews YouTube channel sat down with Hidary to pick his brains on this complex but fascinatingly relevant subject and hear the case for quantum computing.
"There are certain problems that quantum computers will be able to solve in very short amounts of time that today would take long, long cycles."Jack Hidary, CEO of Sandbox AQ
When they are designed to be fault-tolerant, will quantum computers be able to solve problems instantly?
I wouldn't use the word instantly, but in very short amounts of time, and if we limit it to certain problems, then yes: there are certain problems that quantum computers, when they are scaled and fault-tolerant, or error-corrected as we say, will be able to solve in very short amounts of time that today would take long, long cycles.
We keep hearing that quantum computing will be so much more powerful than classical computing. Are we talking about the sun versus a single nuclear power station in terms of a scale comparison of output?
The difficulty is that it's not a speed-up of all things. Quantum computers can do very few things better – but the things they do better, once we get to scaled computers of the future, they will do exponentially better. If you had something as tall as a streetlight and then something that was in space, that's the differential. But only for a handful of things: you would not get any more speed-up of your Excel spreadsheet, it would not run your database faster.
One thing to remember about quantum computers is that there's no hard drive and no memory, so we have to hybridize them with classical computers. The exciting thing about the birth of quantum computers is that they're being born on the cloud. Over the last forty years, if we look at the history of IBM, Digital, Sun, Dell, Lenovo, all these different revolutions in speed and power, what we see in the majority is when new hardware came out, you had to buy that computer. Then you ran it until it became obsolete. That is a costly methodology, and it means a slow rate of replacement.
So you are saying that cloud computing is a way around that?
Exactly. We already have central processing units (CPUs) and graphical processing units (GPUs) in the cloud, and we’re seeing the rise of quantum processing units (QPUs). CPUs are your workhorse general chip, good at tasks: an Excel spreadsheet, Microsoft Word, Google Docs. But when it comes to more specialized tasks such as rendering video games or lifelike cartoons, or if you have certain calculations you want to do in neural networks with thousands of hidden layers between input and output, then you want to use a specialized chip called the GPU, a graphical processing unit. The future of computing is not quantum or classical, it is quantum and classical – a hybridization of these different chips and what they do well.
Tell us more about how quantum computers can help with greener energy supplies.
Right now, when a utility [company] looks at solar and wind, they do not look at it the same way as other sources because they cannot necessarily depend on the sun being up or the wind blowing. So they call those intermittent power sources. If you then couple a solar panel or wind turbine with a battery system, so that whenever the wind is blowing, it's pumping electrons into the battery system, you have a baseload: a consistent supply of electrons.
Australia has installed a major grid-based battery system. There will be others around the world, but right now, there is very little storage of electrons in the US, fewer than 1%. With the rise of better battery systems, we could store them, but part of what's holding that back is the cost and weight of the material used to make lithium-ion batteries. We need to move beyond lithium-ion to think about new chemistries that do not use cobalt to make these batteries less costly and weighty. Quantum computers will be helpful in advancing those kinds of chemistries.
"As these computers get more fault-tolerant, we'll start to see an impact on healthcare, life sciences, clean technologies in storage and batteries, strong and lighter materials for cars and planes. We're in for a wonderful decade of innovation."Jack Hidary
So as these computers start to get more fault-tolerant, bigger, and more robust, we'll start to see an impact on healthcare, life sciences, clean technologies in storage and batteries, stronger and lighter materials that might be better for cars and planes. I think we're in for a wonderful decade of innovation.
Quantum computers still need to get to base-level maturity, but starting in about five years, they will start to really rock and roll. They will complement the tools we have today. We're not waiting for quantum computers – right now, we are doing life-sciences, drug development, and battery chemistry work on GPUs. That's a very exciting development. We'll continue to do that, and then we'll add in the QPU when it becomes available: so we're looking at a synergistic effect of the GPUs that are getting faster and more capable every six months, as well as the QPUs as that begins to kick in.
And how far away do you envisage this being, can you put a timeframe on it?
It's hard to put an exact timeframe, but I can tell you that work is progressing really well. What I'm encouraged by – and my team and I have no bias in the system in that we work with many kinds of quantum computers – is the incredible pace of discovery and progress.
So I'm cautiously optimistic. The industry has taken a new rate of acceleration, with billions of dollars of capital in the last two years, even during COVID. It's a global effort: this is by no means limited to one or two countries. While that is happening, people are refining the error-correction codes, and I think all this effort will merge.
You'll have more robust hardware. When it comes to qubits, it’s not just the number we're interested in, it's also the fidelity: if I have a qubit and set it in a certain state, what is the ability of that qubit to maintain that state? Does it get all wobbly on us? What we're seeing is higher fidelities coming out of the industry. That will give us the kind of robustness that will allow us to get these calculations done.
So in layman's terms, fidelity equals reliability?
Yes, it's the reliability of that qubit to be put in that state and remain there. Quantum devices such as qubits are very fragile, in the sense that if I have a magnetic field nearby or other perturbations near that system, it could shake it out of that state very easily. In classical computers, there is also error-correcting, your phone has error correction. Not as many as in a quantum system, but there can be errors that emerge.
A common error-correction scheme in the classical world would take three bits to represent every one logical bit. So there's kind of a ‘majority rules’ error-correction scheme. We would take three bits and encode all of them with [the value of] one. If, over time, there was some perturbation of the system that caused one of them to flip to zero by mistake, you'd look at that and say: “Ah, I still have two out of the three [valued at] one, therefore the triplet meant to be one.” And so we'll still count that as a logical ‘one’ in the classical system.
We have to be even more robust when it comes to quantum computers with error correction because they are so much more sensitive to the outside world. That's why for those systems that are cryogenic, it really helps to stop the errors: when you can cool it down and nothing is moving around.
I was doing some research trying to find out when the journey to quantum started. The earliest thing I found was the Shrödinger's Cat experiment in the 1930s, which was intended to debunk the theory that quantum particles only change when a human being observes them. Does the journey to quantum begin there, or did it start before that?
There are several journeys that came together to make quantum computing happen. In 1900 Max Planck, a physics professor at the University of Berlin, got more perturbed by the data that was coming out of experiments of the time, analysis of the makeup of an atom with electrons and protons. The data did not match the physics theories of the late 1800s. One of two things was possible: either the data was wrong, or the physics theories were inadequate. And it turns out that the data was correct. They repeated this experiment again and again, and they got the same solutions.
Planck went to his cottage about an hour outside of Berlin, shut himself off from the world, and put down some calculations that ended up being a paper that was published in 1901, where he coined the word “quantum” in the sense of the Latin for “how much.” Einstein published a paper in 1905, inspired by Planck, to explain with this “quantized” idea how the photoelectric effect works. It was well known for 20 years already that if you had photons hitting certain metal plates, it can dislodge an electron and cause it to start moving. And electrons on the move is what we call electricity: that's the principle behind solar panels today.
But it was a big conundrum in the early 1900s, how that possibly worked – because it did not conform with the electromagnetic answers that we got from the traditional Maxwellian understanding. Einstein gave birth to a deeper understanding of the photoelectric effects using the quantum paradigm, and that led to Shrödinger's equation. He developed the cat experiment – ironically, not to elucidate an aspect of quantum but to attack it.
"Einstein felt intuitively that the world is deterministic, the way that Newton had the idea that if I know the state of affairs at time zero and wind the clock ahead, I can know exactly where everything is. And quantum says we don't have that knowledge."Jack Hidary
Because Einstein and Shrödinger were in cahoots to say that while quantum was an interesting development, it could not be the ultimate understanding of the world. They were not happy that quantum was probabilistic as a theory: Schrödinger and Einstein felt intuitively that the world is deterministic, the way that Newton had the idea that if I know the state of affairs at time zero and wind the clock ahead, I can know exactly where everything is. And quantum says we don't have that knowledge. Einstein wrote a paper in 1935 with Podolsky and Rosen at Princeton, and Shrödinger came up with his thought experiment. Both attempted to bring down the walls of quantum mechanics, although both ended up being incorporated as part of the canon!
It was not until 1979 that Paul Benioff published a paper that laid out the idea that we can use quantum mechanics to do computing. In 1980 Richard Feynman made a famous speech about quantum computing. Benioff had the idea first – but Feynman was far more famous because he had won a Nobel prize, and so when he amplified the idea, it became more widespread. And also, we want to give some credit to Yuri Manin, a mathematician from Russia [who helped to pioneer quantum computing theory around the same time].
NIST says in two years, this process will be complete, that the post-quantum standard will be finalized…
Within two years. It's already kind of baked, in the sense that six years ago, NIST started a process with multiple stakeholders in academia, industry, and governments around the world. People from 25 countries participated, 82 initial submissions were given, that was culled to 15. This has been a series of validation processes with peer review, thousands of people involved around the world. It had to be: for cybersecurity protocols to be trusted, they have to come from an open, inclusive process.
NIST was able to announce that the protocols would move forwards and be the standard. We're all familiar with these protocols, we've had years to engage with them. One of the winners that comes out of this is ladders-based cryptography – it will now replace Rivest-Shamir-Adleman (RSA), the mainframe encryption that we use for internet communication and other applications that was standardized in the 1980s. It's served the world very well – it's fair to say that the global economy is built on public-key encryption. Without it, we couldn't transmit credit cards, or have e-commerce, banking transactions, or confidential information storage of patient data. When we get scaled quantum computers, they will be able to crack RSA encryption.
The governments of the world, led by NIST, got together and said: “We need a new standard, because RSA and elliptic-curve cryptography are cracked by quantum.” You might ask: “Why switch now if the quantum computers that could do this are not yet here?” The answer is in four letters – SNDL, “store now, decrypt later.” There are adversaries out there who know that they'll have quantum computers in a number of years that they will use to crack RSA. They siphon off information over the open internet as you're communicating and make copies of it. They cannot read it now, so they store it and decrypt it in the future.
"Don't be intimidated. If you want to change careers and get involved in quantum, there are a lot of resources - this is the golden age of online education."Jack Hidary
Are they playing the long game?
Definitely. But if you're going after [big] pharma, they've got a pharmacopia of trade secrets: the formulae behind all their drugs, the knowhow, and thousands of compounds they're developing into the medicines of the future. And even if it takes years to decrypt, that's still very valuable information. So they're playing a global long game, and for that reason, the whole world has to upgrade the encryption standards to the new quantum-safe protocols. That will lead to better cyber protection.
But do you think that this very process, being open and inclusive, as you said, could itself be hijacked by state actors?
It's up to each government to put directives out in terms of migrating the 20 billion devices [worldwide that will need updating]. You can keep the phone, but you have to upgrade the software. Every messaging app that you use, Whatsapp or Signal or Telegram or any other, will have to be upgraded over the next year or two to the new quantum-safe standards.
This is going to be a big industry, upgrading all our devices…
Yes, there's a lot of work to be done. It will be by many people: IT organizations, consultancies, systems integrators, and original equipment manufacturers. The routers and phones of the future will come standard with this, they'll bake it into the software. This is going to be a multi-party integration and upgrade; it's going to take four or five years to move our planet to the new standards.
What would you say to people wanting to learn more about quantum computers?
Don't be intimidated. If you want to change careers and get involved in AI or quantum, there are a lot of resources – this is the golden age of online education. There are online courses, books, and videos on YouTube. Start to learn: whether you want to be a sales executive or an engineer or scientist, there are ample opportunities.
So you don't need a background in mathematics to get involved?
No, you develop all that now with online tools. This is the moment: we need more people in the industry, there's a shortage already even though it's young. So I would strongly encourage those viewing and listening today. Accessible resources are launched on the internet every day. We welcome you to the field.
More from Cybernews about quantum computing:
Subscribe to our newsletter