Quantum computers have been on my mind a lot lately. A friend who likes investing in tech, and who knows about my attempt to learn quantum mechanics, has been sending me articles on how quantum computers might help solve “some of the biggest and most complex challenges we face as humans,” as a Forbes commentator declared recently. My friend asks, “What do you think, Mr. Science Writer? Are quantum computers really the next big thing?”
I’ve also had exchanges with two quantum-computing experts with distinct perspectives on the technology’s prospects. One is computer scientist Scott Aaronson, who has, as I once put it, “one of the highest intelligence/pretension ratios I’ve ever encountered.” Not to embarrass him further, but I see Aaronson as the conscience of quantum computing, someone who helps keep the field honest.
The other expert is physicist Terry Rudolph. He is a co-author, the “R,” of the PBR theorem, which, along with its better-known predecessor, Bell’s theorem, lays bare the peculiarities of quantum behavior. In 2011 Nature described the PBR Theorem as “the most important general theorem relating to the foundations of quantum mechanics” since Bell’s theorem was published in 1964. Rudolph is also the author of Q Is for Quantum and co-founder of the quantum-computing startup PsiQuantum. Aaronson and Rudolph are on friendly terms; they co-authored a paper in 2007, and Rudolph wrote about Q Is for Quantum on Aaronson’s blog. In this column, I’ll summarize their views and try to reach a coherent conclusion.
First, a little background. Quantum computers exploit superposition (a particle inhabits two or more mutually exclusive states at the same time) and entanglement (a special form of superposition, in which two or more particles influence each other in spooky ways) to do things that ordinary computers can’t. A bit, the basic unit of information of a conventional computer, can be in one of two states, representing a one or zero. Quantum computers, in contrast, traffic in qubits, which are constructed out of superposed particles that embody numerous states simultaneously.
For decades, quantum computing has been little more than a hypothesis, or laboratory curiosity, as researchers wrestled with the technical complexities of maintaining superposition and entanglement for long enough to perform useful calculations. (Remember that as soon as you look at an electron or cat, its superposition vanishes.) Now, tech giants like IBM, Amazon, Microsoft and Google have invested in quantum computing, as have many smaller companies, 193 by one count. In March, the startup IonQ announced a $2 billion deal that would make it the first publicly traded firm dedicated to quantum computers.
The Wall Street Journal reports that IonQ plans to produce a device roughly the size of an Xbox videogame console by 2023. Quantum computing, the Journal states, could “speed up calculations related to finance, drug and materials discovery, artificial intelligence and others, and crack many of the defenses used to secure the internet.” According to Business Insider, quantum machines could help us “cure cancer, and even take steps to reverse climate change.”
This is the sort of hype that bugs Scott Aaronson. He became a computer scientist because he believes in the potential of quantum computing and wants to help develop it. He’d love to see someone build a machine that proves the naysayers wrong. But he worries that researchers are making promises they can’t keep. Last month, Aaronson fretted on his blog Shtetl-Optimized that the hype, which he has been countering for years, has gotten especially egregious lately.
“What’s new,” Aaronson wrote, “is that millions of dollars are now potentially available to quantum computing researchers, along with equity, stock options, and whatever else causes ‘ka-ching’ sound effects and bulging eyes with dollar signs. And in many cases, to have a shot at such riches, all an expert needs to do is profess optimism that quantum computing will have revolutionary, world-changing applications and have them soon. Or at least, not object too strongly when others say that.” Aaronson elaborated on his concerns in a two-hour discussion on the media platform Clubhouse. Below I summarize a few of his points.
Quantum-computing enthusiasts have declared that the technology will supercharge machine learning. It will revolutionize the simulation of complex phenomena in chemistry, neuroscience, medicine, economics and other fields. It will solve the traveling-salesman problem and other conundrums that resist solution by conventional computers. It’s still not clear whether quantum computing will achieve these goals, Aaronson says, adding that optimists might be “in for a rude awakening.”
Popular accounts often imply that quantum computers, because superposition and entanglement allow them to carry out multiple computations at the same time, are simply faster versions of conventional computers. Those accounts are misleading, Aaronson says. Compared to conventional computers, quantum computers are “unnatural” devices that might be best suited to a relatively narrow range of applications, notably simulating systems dominated by quantum effects.
The ability of a quantum computer to surpass the fastest conventional machine is known as “quantum supremacy,” a phrase coined by physicist John Preskill in 2012. Demonstrating quantum supremacy is extremely difficult. Even in conventional computing, proving that your algorithm beats mine isn’t straightforward. You must pick a task that represents a fair test and choose valid methods of measuring speed and accuracy. The outcomes of tests are also prone to misinterpretation and confirmation bias. Testing “creates an enormous space for mischief,” Aaronson says.
Moreover, the hardware and software of conventional computers keeps improving. By the time quantum computers are ready for the marketplace, they might lose potential customers—if, for example, classical computers become powerful enough to simulate the quantum systems that chemists and materials scientists “actually care about in real life,” Aaronson says. Although quantum computers would “retain their theoretical advantage, their practical impact would be less.”
As quantum computing attracts more attention and funding, Aaronson says, researchers may mislead investors, government agencies, journalists, the public and, worst of all, themselves about their work’s potential. If researchers can’t keep their promises, excitement might give way to doubt, disappointment and anger, Aaronson warns. The field might lose funding and talent and lapse into a quantum-computer “winter” like those that have plagued artificial intelligence.
Lots of other technologies—genetic engineering, high-temperature superconductors, nanotechnology and fusion energy come to mind—have gone through phases of irrational exuberance. But something about quantum computing makes it especially prone to hype, Aaronson suggests, perhaps because “‘quantum’ stands for something cool you shouldn’t be able to understand.”
And that brings me back to Terry Rudolph. In January, after reading about my struggle to understand the Schrödinger equation, Rudolph emailed me to suggest that I read Q Is for Quantum. The 153-page book explains quantum mechanics with a little arithmetic and algebra and lots of diagrams of black-and-white balls going in and out of boxes. Q Is for Quantum has given me more insight into quantum mechanics, and quantum computing, than anything I’ve ever read.
Rudolph begins by outlining simple rules underlying conventional computing, which allow for the manipulation of bits. He then shifts to the odd rules of quantum computing, which stem from superposition and entanglement. He details how quantum computing can solve a specific problem—one involving thieves stealing code-protected gold bars from a vault–much more readily than conventional computing. But he emphasizes, like Aaronson, that the technology has limits; it “cannot compute the uncomputable.”
After I read Q Is for Quantum, Rudolph patiently answered my questions about it. You can find our exchange (which assumes familiarity with the book) here. He also answered my questions about PsiQuantum, the firm he co-founded in 2016, which until recently has avoided publicity. Although he is wittily modest about his talents as a physicist (which adds to the charm of Q Is for Quantum), Rudolph is boosterish about PsiQuantum. He shares Aaronson’s concerns about hype, and the difficulties of establishing quantum supremacy, but he says those concerns do not apply to PsiQuantum.
The company, he says, is closer than any other firm “by a very large margin” to building a “useful” quantum computer, one that “solves an impactful problem that we would not have been able to solve otherwise (e.g., something from quantum chemistry which has real-world uses).” He adds, “Obviously, I have biases, and people will naturally discount my opinions. But I have spent a lot of time quantitatively comparing what we are doing to others.”
Rudolph and other experts contend that a “useful” quantum computer with robust error-correction will require millions of qubits. PsiQuantum, which constructs qubits out of light, expects by the middle of the decade to be building fault-tolerant quantum computers with fully manufactured components capable of scaling to a million or more qubits, Rudolph says. PsiQuantum has partnered with the semiconductor manufacturer GlobalFoundries to achieve its goal. The machines will be room-sized, comparable to supercomputers or data centers. Most users will access the computers remotely.
Could PsiQuantum really be leading all the competition “by a wide margin,” as Rudolph claims? Can it really produce a commercially viable machine by 2025? I don’t know. Quantum mechanics and quantum computing still baffle me. I’m certainly not going to advise my friend or anyone else to invest in quantum computers. But I trust Rudolph, just as I trust Aaronson.
Way back in 1994, I wrote a brief report for Scientific American on quantum computers, noting that they could, in principle, “perform tasks beyond the range of any classical device.” I’ve been intrigued by quantum computing ever since. If this technology gives scientists more powerful tools for simulating complex phenomena, and especially the quantum weirdness at the heart of things, maybe it will give science the jump start it badly needs. Who knows? I hope PsiQuantum helps quantum computing live up to the hype.
This is an opinion and analysis article.
Will Artificial Intelligence Ever Live Up to Its Hype?
Is the Schrödinger Equation True?
Quantum Mechanics, the Chinese Room Experiment and the Limits of Understanding
Quantum Mechanics, the Mind-Body Problem and Negative Theology