transcript

Speaker 1:
[00:00] One of the most exciting areas of computing research right now is quantum computing. A quantum computer is totally unlike the traditional computer that you're familiar with. It solves problems in a completely different way and has the potential to revolutionize certain fields. However, the promise of quantum computing has led people to make outrageous claims and assumptions that sometimes border on the magical. Learn more about quantum computing, how it works and what it can and cannot do. On this episode of Everything Everywhere Daily. This episode is sponsored by Samsara. If a driver in your fleet got in an accident tomorrow, could you prove what actually happened? Without footage, it's your word against theirs, and that can mean higher insurance costs, drawn out claims and a lot of uncertainty. That's why so many fleets are turning to Samsara. Their AI-powered dash cams capture clear real-time video evidence helping protect drivers from false claims. In fact, Samsara customers have reduced crash rates by nearly 75%. And it isn't just about reacting to accidents. Samsara helps prevent them. With real-time alerts and coaching tools, organizations like the City and County of Denver saw a 50% reduction in false claims against them and a 94% reduction in safety events. Samsara isn't just a dash cam. It's a completely connected operations platform. You get GPS tracking, maintenance insights and compliance tools all in one place. Don't wait for the next accident to take action. Head to samsara.com/everything to request a free demo and see how Samsara brings visibility and safety to your operations. That's samsara.com/everything. Samsara. Operate Smarter. This episode is sponsored by fastgrowingtrees.com. Spring is here, which means it's time to plant and landscape. If you want to make things easier on yourself, check out fastgrowingtrees.com. FastGrowingTrees is America's largest and most trusted online nursery, with thousands of trees and plants and over 2 million happy customers. In addition to everything you might need for your yard, they also have a wide selection of indoor plants and trees as well. I got two indoor ficus trees from fastgrowingtrees.com. It was super easy to take them out of the box, and the process was much cleaner than if I had to go to a local nursery. Right now, they have great deals on spring planting essentials, up to half off on selected plants. And listeners to my show get 20% off their first purchase when using code DAILY at checkout. That's an additional 20% off better plants and better growing at fasttrees.com using code DAILY at checkout. fastgrowingtrees.com code DAILY. Now is the perfect time to plant. Let's grow together. Use DAILY to save today. Offer is valid for a limited time. Terms and conditions may apply. The subject of quantum computers is one that I have been working on for quite a while, and I felt it was finally time to pull the trigger and do the episode. I am also aware that given the current state of research, I will probably have to do another episode in a year or so to cover the latest updates in the technology. There has been a lot of talk about quantum computing, and many people discussing it really don't seem to understand much about it. They often use quantum computing as a synonym for magic and assume that it can do almost anything. To be sure, quantum computers have great potential, but there are limits to what they can do and how they can be used. Before I get into how quantum computers work, let's briefly review how digital computers work. All digital computers and everyone that you have ever used, manipulates and stores data in binary form. Everything is represented as a series of 1s or 0s, known as bits. All of the improvements in computing over the years have been about processing these bits faster and storing more of them. Quantum computers do not work this way. Instead of bits, they take advantage of properties that manifest at the quantum level and manipulate what are called qubits or quantum bits. A qubit can be a 1, a 0, or can be both at the same time. This property is known as quantum superposition. It can only exist at the quantum level and is one of the foundations of quantum computing. The idea of something being two things at the same time can be difficult to get your head around, but the best analogy I've heard is that it's like spinning a coin. A coin can be heads or tails, but while it's spinning before it stops, it has properties of both heads and tails. Another quantum phenomenon core to the workings of a quantum computer is quantum entanglement. Quantum entanglement is a phenomenon where two or more particles become linked so that measuring one instantly determines the correlated state of the other, even when they're separated by great distances. To use our coin analogy, imagine two coins spun together, so that when one of them lands heads, the other will instantly land tails, no matter where it is. Quantum superposition and quantum entanglement are very difficult concepts to grasp because we don't experience these effects in our day-to-day lives. The great physicist Richard Feynman once said, I think I can safely say no one understands quantum mechanics. And Richard Feynman won a Nobel Prize for his work in quantum mechanics. If you remember back to my episode on the ultraviolet catastrophe, our understanding of quantum effects doesn't come from intuition and direct observation as in Newtonian physics, but rather from the confirmation of mathematical models. The idea of using quantum effects for computing came from Richard Feynman himself. In 1981, Feynman argued that a machine built according to quantum rules could model nature far better than ordinary computers. And in the 1980s, the British physicist David Deutsch expanded on this, developing the concept of a general-purpose quantum computer. The reason why a quantum computer can, in theory, do things that a normal digital computer can't, is that a quantum computer doesn't try one solution at a time. It can explore many possibilities simultaneously. A digital computer can only process things sequentially. To use another analogy, let's assume that you have hundreds of closed boxes, and one of them has a gold bar. A digital computer would solve the problem of finding the bar by opening one box after another until it found it. Even if a digital computer uses parallel processing, it's just several processors working sequentially. A quantum computer would be able to check all of the boxes at the same time. So, again using our analogy, instead of opening every box, imagine shaking the tables that the boxes are sitting on. The box with the gold brick being heavier would behave differently from the other boxes, allowing you to identify it. Again using the analogy, shaking the table can help find the gold brick, but it can't tell you what is inside every box. It can just help us solve this one problem quickly. Theoretical quantum computing gained momentum in the 1990s when the theoretical computer scientist Peter Shor of MIT developed Shor's algorithm, showing that a quantum computer could factor large numbers far faster than known classical methods. Soon after, Stanford's Law of Grover introduced Grover's algorithm, demonstrating faster database search. These breakthroughs proved that quantum computers could offer genuine computational advantages. Theorizing a quantum computer was much easier than actually building one. Digital computers went from very large vacuum tubes to slightly smaller transistors and finally to integrated circuits. Because quantum effects only manifest at the atomic level or smaller, engineering challenges for building a quantum computer were and are enormous. Building a quantum computing device is one of the hardest engineering projects ever attempted because it requires controlling matter at the smallest scales, all the while preventing outside disturbances. In effect, engineers must build a machine that is both extremely powerful and extremely fragile. The first challenge is creating reliable qubit hardware. Unlike an ordinary transistor, a qubit must preserve delicate quantum states such as superposition and entanglement. The next problem is environmental noise. Heat, stray electromagnetic fields, vibration, radiation, and tiny material defects can disrupt a qubit state through quantum decoherence. That means engineers must isolate the processor from nearly everything around it. Many machines operate inside dilution refrigerators colder than outer space with extensive shielding and vibration control. Precision control is another major hurdle. Quantum operations require exquisitely accurate pulses of microwaves, lasers, or voltages delivered at exactly the right time with the correct strength. Small errors accumulate quickly. A useful machine may need millions or even billions of perfectly timed control operations, all coordinated with almost no error. And scaling is perhaps the hardest challenge. Demonstrating a few qubits in a lab is very different from building thousands or millions that can work together. As qubit counts grow, wiring, cooling, signal routing, calibration, and crosstalk between neighbor and qubits become serious obstacles. In some systems, the control hardware can become larger and more complex than the quantum chip itself. Error correction adds another layer of difficulty. Individual qubits are too error-prone for large computations. That means for each logical qubit that's used for a computation, it might require dozens to hundreds of physical qubits simply to provide error correction. As quantum computers develop and require longer processing time, they may require as many as thousands of physical qubits per every logical qubit. Qubits can be made from superconducting circuits, trapped ions, neutral atoms, photons and other systems. But each approach has trade-offs in speed and stability. Simply producing qubits that behave consistently across different devices is extremely difficult. From the 2000s onward, laboratories and companies began building real hardware using these qubit technologies. In 2019, Google announced that it had achieved quantum supremacy. Quantum supremacy sounds really impressive, but it has a very specific meaning. It means a quantum computer has completed at least one specific task that would be impractical for a classical computer to reproduce in a reasonable time. Soon after Google made the announcement, IBM argued that improved classical computing methods could do the same benchmark in days, not thousands of years, like Google claimed, which would put the entire quantum supremacy claim in question. The state of quantum computing as of the recording of this episode is still very, very early despite press releases to the contrary. If you wanted to compare this to the evolution of digital computing, we would probably be in the 1940s when the very first computers were created. Several different types of quantum computing systems are being developed today because no one yet knows which technology will scale the best. The most mature approach today are the previously mentioned superconducting quantum computers used by companies such as IBM and Google which use ultra-cold electrical circuits as qubits and benefit from chip manufacturing techniques. Another leading design is trapped ion quantum computers pursued by companies like IonQ and Quantinuum where charged atoms are suspended and controlled with lasers, offering excellent accuracy but slower speeds. Neutral atom systems developed by firms such as QERA use arrays of laser-trapped atoms and are considered promising for scaling to large numbers of qubits. Other approaches focus on different strengths. Photonic quantum computers pursued by Xanadu and others use photons as qubits and may be useful for networking and room temperature operation. Silicon spin qubits use the quantum spin of electrons inside semiconductor chips and could potentially leverage existing chip factories. So how much better would quantum computers have to get to do something practical like break contemporary cryptography? To break most modern public key cryptography you would need a large-scale, fault-tolerant quantum computer. Not a small, noisy prototype, nor today's machines that make headlines with their raw qubit counts. It would need stable, logical qubits protected by quantum error correction capable of running long, highly accurate computations. It would require several orders of magnitude, more physical qubits than are currently capable. A cryptography-breaking system would probably need thousands of logical qubits, and potentially millions of physical qubits. There is also the previously mentioned issue of quantum decoherence. Quantum coherence is the ability of a quantum system to maintain its delicate superposition and phase relationships long enough to perform meaningful quantum operations before environmental noise destroys them. Superconducting qubits used by IBM and Google typically maintain coherence for tens to hundreds of microseconds, with some devices ranging into the millisecond range. To break modern cryptography, the relevant question is not just how long a single bit remains coherent, but whether the entire quantum computation can preserve usable quantum information long enough to complete billions or trillions of operations. In a fully fault-tolerant machine, individual qubits would still decohere quickly, but continuous quantum error correction would refresh and protect the logical state, allowing the computation to run for hours, days, or potentially longer. One assumption many people have made about the future of quantum computers is that it will advance in a way similar to how digital computers have advanced. However, that isn't guaranteed. Quantum computers are not digital computers. It's an entirely different technology that just happens to use the word computer. Even if the technology continues to advance rapidly, the first quantum computer capable of breaking modern cryptography would probably cost billions of dollars. However, even that future is not guaranteed. There are many serious physicists who think that the problems with scaling quantum computers can't be overcome, or at least can't be overcome affordably. Each logical qubit can experience errors when enough underlying physical qubits suffer noise, decoherence or faulty operation. The more physical qubits in a system, the larger these problems and engineering challenges become. To keep heat and noise down, quantum computers require extreme cooling to operate at temperatures only several thousandths of a degree above absolute zero. The other big problem is that there aren't many use cases for quantum computers yet. Outside of certain areas such as factoring large numbers, chemical simulations and some optimization problems, skeptics question whether there are enough commercially viable workloads. But assuming that quantum computers can in fact scale and tackle large problems, they will then impact the entire economy. Many people worry about its impact on things like bitcoin, which will be covered in a future episode. But if a quantum computer can break cryptography, every bank, financial institution, online retailer, major technology company, and government is going to be at risk. The takeaway from this episode is that quantum computers are real and can, in theory, solve problems that would be extremely difficult or impossible for normal digital computers. However, the current state of quantum computing is still very primitive despite press releases to the contrary. It will be necessary to increase the number of physical qubits by two or three orders of magnitude in order to get to a point where real problems can be solved. To get to that point, many important engineering obstacles will have to be overcome, some of which might involve the invention of entirely new technologies or techniques. Quantum computing is not the next phase of general computing. You will not have a quantum computer on your desk, nor will Microsoft make Windows Quantum. However, if the engineering obstacles can be overcome, years from now, quantum computing might play an important role in the world. The executive producer of Everything Everywhere Daily is Charles Daniel. The associate producers are Austin Oetken and Cameron Kieffer. My big thanks go to everyone who supports the show over on Patreon. Your support helps make this podcast possible. And I also want to remind everyone about the community groups on Facebook and Discord. That's where everything happens that's outside the podcast. And links to those are available in the show notes. As always, if you leave a review on any major podcast app or in the above community groups, you too can have it read in the show.