I went through a theoretical physics phase in high school. Our school only offered the standard high school courses, but I started collecting, and became enamored, with books written by famous contemporary physicists: Carl Sagan, Stephen Hawking, Michio Kaku, Brian Greene, etc. In particular, I admired how they were all about, seemingly effortlessly, communicating complex ideas and abstract mathematical concepts in layman’s terms—to get impressionable youth like me excited about their field. I briefly considered making a career out of math and physics, but ultimately took the computer science route instead1.
Quantum Supremacy is the latest book by Michio Kaku, a theoretical physics professor at CUNY by title, but also a prolific author and media personality. It’s a book that introduces quantum computing as an area of active research and technical development. Before Generative AI sucked all the oxygen out of the school, this was one of the most promising technological advancements post smartphones; the hype was and continues to be about what quantum computers can accomplish, well beyond the bounds of modern PCs and even supercomputers.
Initially, I wasn’t sure why a physicist would so extensively opine on computing. Granted, the cliché is that physics ultimately underlies all of the hard sciences2, but the science of computing focuses on problems that seem far afield from what astrological and interstellar phenomenon that serves as the physicists’ playground. If anything, computer science overlaps with mathematics more than physics.
Well, the book explains that quantum computing is about leveraging atoms as units of computation. Not only are atoms one of the smallest divisions of matter we can reasonably control—ignoring subatomic quarks for a second as we can barely observe them—but within the realm of quantum mechanics, atoms exhibit a property called quantum entanglement, where multiple atoms can impact each others’ state as a part of a whole system. These properties then allow for massive increases in computational power, compared to the digital computers we’ve built for the past 80 years.
As far as I understand it, the main difference lies in the qubit, the quantum equivalent of the bit as the base unit of computation. Both hold the binary 0 and 1 state, and you can string together multiple bits or qubits to represent exponentially more states, e.g., 8 bits lets you represent numbers from 0 to 255 (2^^8). The major advantage of qubits, though, is that they can represent all these states at the same time, compared to bits which can only store one state at a time. A rough analogy would be a password-cracking algorithm: traditional computing would have the algorithm go through all possible passwords one at a time and faster CPUs let the program run through each attempt more quickly; quantum computing would simply try all the passwords at the same time all at once.
It’s been tough to wrap my head around; the entire discipline of computer science, not to mention all modern technology and computing, is built on top of the abstraction of bits and deterministic execution of linear programs. By contrast, not only is the quantum mechanics notoriously unintuitive, but the implications of using these rules for computing results in probabilistic output based on massively parallel algorithms. Beyond the physical difficulties of creating and stabilizing these qubits, the foundational theories of computing have to be updated to account for this wholly new way of processing information.
But back to the book—once Kaku gets satisfied with explaining the theoretical physics behind quantum computing, he spends a lot of the rest of his writing in pure speculation mode. Most of the chapters describe major problems in today’s world: genetic modifications, nuclear fusion, weather predictions, etc., and then follow through with how a massive increase in compute, courtesy of quantum computers, could render solutions. It’s not specified exactly how this would happen; there’s a tinge of the same assertions that Artificial General Intelligence (AGI), if ever manifested, would quickly solve all of humanity’s problems3. Ironically, a chapter goes deeper into the history of AI and the Turing test. Even as it postulates the advent of quantum computing to aid AIs of the future to pass this test, ChatGPT launched before this book was published, and indeed leveraged large amounts of data and compute to simulate human responses—on simple ol’ GPUs.
Quantum Supremacy is a fun read. The accessible prose and metaphors echo those introductory physics books I read in high school, and it serves as a friendly gateway for those who want to dive deeper into the topic. It got a bit too fantastical for me by the end, but the breadth of examples is a reminder that we still don’t understand much of how atoms interact: how diseases fundamentally manifest; how we can harness all the potential energy sources around us; how air particles interact to form weather systems. Quantum computing is certainly applicable to all these real-world challenges, but there’s potential for it to also shed light on the physics behind it all, and perhaps that’s why a physicist is as excited about its development as any technologist.
I entertained the idea of tacking on an astrophysics minor to my CS major, but was talked out of it by my undergraduate counselor—a decision I still regret.↩
Well, the natural rebuttal is that math underlies physics.↩
Or, of course, doom us all.↩
