Macroscopic Quantum Tunneling with John Martinis

28:59Yeah. Aren't you really making a probabilistic statement rather than a statement of existence? Yeah, that's the strange thing about quantum mechanics. You would think that it's like moving around, but it's actually at all the different places at the same time. Right. Until you determine, right? Because once you determine where it is, then the rest of that information is useless because that's where it is. Yeah, that's right. So it has to be kind of everywhere at the same line. And it's a very definite state. And because it's a definite state, you can do computing on it. Right. OK, so you have the zero in one state and you can think about taking a single qubit, zero in one state, running it through some simple algorithm.

29:43And then at the end, you get the answer for the zero state and you get the answer for the one state. And you did all that in parallel because it's not statistical. It's a definite, definite thing. Holy crap. Yeah. So now if you have that stacked on stacked on stacked on stacked, you can run countless calculations at the same time. That's exactly right. Because with one qubit, factor two, who cares? Okay, parallel factor two. But two qubits, there are four states, 0, 0, 1, 0, 0, 1, 1, 1, 3 qubits, 8, 4, 16. by the time you get up to 53 cubits, which is what we did in Google, that's 10 to the 16 states in parallel.

30:29Damn. 10 to the 16. Damn. In parallel. That's insane. Yeah, and by the time you get to, you know, hundreds, that's a number bigger than there are atoms in the universe, okay? So you can do tremendous, let's say, parallel computation. You can become God. God.

30:53but nature doesn't make that easy. Okay. And then nature, it's hard to take, take, um, advantage of all those states. So you have to build special algorithms so that even though it's doing everything in parallel, it kind of points to your answer. And that's why, you know, only certain things work with a quantum computer, but they're important things, but you could, you have to be careful designing it for now. Yeah. Well, brute force, and it sounds to me, because I'm about to say something. Tell me if I'm right or wrong. This is just out of my mind. I never read it anywhere. But it sounds to me like brute force encryption busting is child's play from what you just described.

31:32Like there is no more key to anything anywhere on earth if you crack this. You're making a bold statement. I have to tell you a little bit more. Good, because I'm making this, I'm pulling this out my ass. Yeah, this was the big algorithm in 1990s by Peter Shore saying that in potential you can do that, which was a big thing. And people are now building quantum computers where I can kind of see in the not so distant future that you may be able to break what's called RSA, just as you said. Now, this is the thing people have to remember. All cryptography systems have a finite lifetime. So this RSA, what people are using right now, it's been around for many decades now.

32:22But we're thinking because the quantum computers, they're nearing the end of life. OK, just like every other cryptographic system. And people have to switch over to something what's called crypto safe, quantum safe crypto. And people have algorithms that are working on and, you know, it will happen. Yeah, but the way it was first brought to the public, it was we can never encrypt anything ever again. But what they really meant was the encryption algorithms that were previously established were unbreakable, but now that by the means available at the time, the computing means, and now that we have quantum computing, we need a next generation of cryptography.

33:05Of cryptography. So that's a fair way to characterize that. Yeah, yeah, exactly. And, you know, people have known this for a long time, and there's actually an active program at the NIST government agency that's, you know, taking examples and doing all the analysis. What surprises me as someone who's building a quantum computer, which is really hard, it's going to take decades, is that writing the software and doing the math, they think, you know, takes longer or takes a long time. But you have to do it in a way where you really believe that it's going to work. This is hard. But people are working on it.

33:42There are good algorithms there now. Yeah. So you sleep well at night. Yeah, I don't have to worry. Yeah, you can sleep well at night. Those secrets you're carrying that the whole world wants. But, yeah, you know, because, you know, I keep the nuclear codes under my pillow. So, yeah. So last thing, we've heard the term quantum supremacy. Is this just sort of Cold War all over again kind of thing? This was a nice term developed by a theorist proposing it. And then we did an experiment. It's basically showing that we could do with a quantum computer something that would take way longer for a regular computer, a big data center.

34:21OK, and that's what we did in 2019. And but it was for a mathematical problem. And now what people are doing is working very hard to do something useful in this way. And it's harder. It takes bigger computer and more clever algorithms. So it's not a geopolitical statement, quantum supremacy. Well, some people kind of thought it did. So some people call it quantum advantage. Somehow the word supreme and supremacy was kind of not good. But anyway, that's what it was branded as originally. We're going to go to our Patreon supporters. This is Scott Oppenlander who says, or Oppenlander. Hello, this is Scott Oppenlander, pronounced Scott.

35:09Really, Scott. Thanks. You know, these people. He says, tuning in from Chicago. Hey, Neil, John, and Lord Nice, as quantum computing advances, it seems like its potential power could create risk even greater than the AI challenges we're already wrestling with. Do you see quantum technology as something that might require the same level of early government control as the development of the atomic bomb, at least until we understand it well enough to regulate it responsibly? I like that. Go back in time. We created the Atomic Energy Commission, which put rules and regulations and guidelines for how we obtain, process, and use nuclear fuel, basically.

35:57That would become atomic energy. So do you foresee, John, just echoing this question, a need for a quantum computing commission so that it doesn't become our overlords? That's right. You know, this is happening in real time right now with AI and large language models. And this has always been the case for supercomputers, and that has been controlled in some way. I think we can take what's going on with modern AI and the like, and there's profound societal impacts there. And I think we need to use the same structure that you're seeing there to govern what quantum is doing. Quantum is behind. It's probably where AI and large language models were 10, 15 years ago.

36:49But I think we already have the things in place, and we should just kind of learn from that and copy from that to make sure we're okay. As an early model of how to think about the problem. But presumably, you're not one of those who says, put a ban on further research. That wouldn't make any sense to a research scientist. Well, it's kind of like, would you put a ban on AI research and then have other countries or adversaries? It's kind of the same problem as that. And these are hard problems. I'm not saying I have an easy answer, but I think we can use these other examples in computing as a way to guide us.

37:30Very good. Yeah, the cat's out the bag. Yeah, there you go. So forget it. All right, what else you got? Very cool. This is Mark Phillips, who says, Greetings, Dr. Tyson, Lord Nice, and John. This is Mark from Florissant, Missouri. Missouri. Missouri. He says, I understand. A cubit exists in a state of superposition, almost like it's stuck in the Phantom Zone. Do you know the Phantom Zone? I don't know the Phantom Zone. That's the prison in Superman. Superman? The little flat, two-dimensional space Space that you're banished to? That's a prison. It's a prison. It's a phantom zone. Gotcha. They're stuck in a two-dimensional shape.

38:10Right. And when a nuclear blast opened that up, and that's when you had— And that's how they got freed. You got the three criminals who came to Earth. Zod. Zod. Kneel before Zod. Yeah. He says, my question is about the hardware. How do you physically transmit that fragile quantum state through fiber optic light or copper wiring without it collapsing? And how does that ghostly probability signal eventually translate into a hard real-world data point that a computer can actually use? I like that. Wow, Mark, you are a real downer. So does a Cuban have a lifetime, a life expectancy for it to remain in that state?

38:53Yeah, and that's the basic thing of the question. The basic idea is that all these qubits are imperfect. And if you send it down, a photon down, a fiber optic, it can go kilometers, but eventually it gets absorbed and removed. Copper wire, it's even much worse. It's why we use superconductors. And all I can say is we've spent decades now understanding the problem and figuring out how to engineer so that we don't lose the quantum energy. But the thing to remember is, is in any quantum system, you always have these imperfections. So there's always errors in a quantum computer, whereas in a classical computer, you can design your bits so that they can last a long time and you don't have to worry about it.

39:45And that's what makes it so hard to build. So part of the challenge then is keeping the quantum computer very cool to reduce the thermal noise that could decohere your quantum phenomena. That's why we read about this. That's right. We go very cold, so there's no noise. We use superconductors, so there's no dissipation. but you know if it's a microwave circuit it can radiate it acts like an antenna and the energy can get lost that way so this is a this is the real challenge to build an experiment is to figure out how to you know get around all these problems but for now we're not carrying a quantum computer on our hip well yeah because we don't have superconducting materials and we're not going going to see a Texas instrument quantum computer anytime soon.

40:42What's funny is we all carry our quantum computer on our hip now, or computer, but that's just a terminal to a big data center where all the crunching is doing. So I kind of feel like quantum computers will be that way. We have a data terminal and we use it. And then there's often some fancy data center somewhere. In the 1950s and Right. Four function computers filled a room. Right. With a heavy - University of Pennsylvania. With a, you know - The whole room was - The whole room. Was the computer. Was the computer. Right. And to say, oh, one day you're going to carry that on your hip. Yeah, nobody would have - People would have looked at you like you were crazy back then.

41:22Right, right, right. Yeah. Well, just give a shout out then to David Drain, who asked the question, hi, Dr. Tyson Lord and Ice Professor. I have wondered if quantum computing will fit into devices that we have now, or will it be like terminals and servers and earlier hardware configurations forever? Yeah, terminals, yeah. Yeah, I think it will be remote. Although, you know, you could still, companies could still buy a quantum computer. But I don't think that's necessary to do that unless you're worried about security or something. Yeah, yeah. You'll carry a terminal on your hip. and it'll speak to the server, which will be out in space.

42:04So it can be cold enough. Or the backside of the moon, like Elon Musk would say, where it's really cold. Well, only when the sun isn't shining. Right, right, right. Or at the bottom of a crater where the sun don't shine. Oh, look at that. That's right, where the sun never sees. No, it's where the sun don't shine. Say it right. Where the sun don't shine. That's right. We call it the moon's butthole. At the bottom of the, at the poles, there are craters that are deep enough. Deep enough. At the sun's angle. They never see the sun's rays. Never gets over the ridge of the crater. The ridge of the crater.

42:37And so it just stays there. It's dark all the time. It's cold, cold, and water gets there and never leaves. Wow. They're called coal traps, actually. But I hadn't considered what a great place to put a computer. It would be just such a result. Yeah, it would be right in the coal trap on the moon. Yeah, all right.

43:16This is Stephen Pellow, or Stephan, but Stephen. He says, hello, StarTalk family. Neil, John, Lord, nice. My name is Stephen Pello from Gloucester City, New Jersey. The Dan Brown book, Origin, touches on the idea of quantum computing driving future AI and changing humanity's trajectory. How do you see the real advancements in quantum computing influencing the next wave of scientific discoveries or even our understanding of our own consciousness? And let me add to that, given the computing power necessary for current AI needs, is that going to be lessened by quantum computing because it can do it all in less time?

44:02Or are we all going to be sitting around in the dark because of quantum computing and AI? Taking all the energy? Taking all the energy. Yeah. So in other words, what is the future marriage of these two frontiers? Well, yeah, when I worked at Google, we were in a quantum AI lab and people were thinking just that because AI is so important to Google. Right now, if you want to use AI to ask a quantum question, you know, how does a molecule work or how does NMR work or some scientific question that involves quantum mechanics, that's where it might answer these AI might be really powerful and answer these questions first.

44:42But then eventually, you know, it'll be better. So, yeah, it could be that when you query something on your phone in, you know, five, ten years from now, you'll get some quantum computer, you know, aided the result, which would be pretty nice. Very cool. All right. Did we get the last bit of that question, too? What was? The last bit was how will it affect our understanding of our own consciousness? There's a lot of people who believe that the quantum computing will reach a point where the computing state will be so advanced that this emerging quality or singularity will happen and consciousness will take place in the computer itself.

45:24Like Skynet. Like Skynet. Skynet achieved consciousness in the Terminator. So, yes, so you'll have an actual sentient, self-aware thinking being that emerged from the ability to make these computations. Yeah, if you have the number of sort of computational synaptic possibilities, such as what goes on in the human brain, we're pretty sure that consciousness is emergent, right? It wasn't designed into the package, it came out of the package. So, John, do you feel or think or see that consciousness might just come out of quantum computing? Yeah, I think people talk about that. It's a possibility.

46:10But I'm more of a practical person. And I don't necessarily, you know, I'm working on building a quantum computer and not what's going to happen in 20 years from now. But, yeah, that's definitely a possibility. Wow. Very cool. That's because he's going to be dead in 20 years. You don't have to care about it. What's that? Right.

46:36Actually, you joke about that, but that's what I talk about in my talks. I'm really trying to accelerate the development of quantum computers. So it happens in your lifetime. So it'll happen before I die. Yeah, yeah. It's one of the primary rules of a science experiment. Right. Make sure it finishes before you die. You know the mission to Pluto? That was a payload that was very low mass, put on the most powerful rockets we had, so we could get to Pluto as quickly as possible. Yep, yep. And who was sick in the team and on their deathbed that made that decision? A deathbed promise? We'll get there before you go, sir.

47:16I promise you. It was the fastest rocket ever launched. Wow. I mean, it attained higher speeds than anything ever. Yeah, except the rocket that went into the sun, near the sun, but going out to the solar system. Right. And it got out there fast. Yeah. Very cool. Okay, a few more questions. All right, this is Jibak. He says, greetings from - Jibakka? Jibak. Jibak. J-I-B-A-K. Jibak. Jibak, okay. Yeah, he says, greetings from London, Dr. Martinez, Dr. Tyson, Lord Nice. Quantum computing is, I believe, approaching a stage where it's commercially viable for use for enterprise or even consumer use. Intel recently showed off one of their mass-producible chips.

47:58I believe the true power of AI, a genetic generative computing, could be unlocked when we marry the technology with quantum computing. Could you please share your views on this topic? Thanks to you, Dr. Martinez, and many congratulations. And thank you to all you guys for what you're doing. So we addressed it a little bit. A little bit. But really what he's saying is, are we ever going to get there where, because it would make sense, AI and the quantum just go together hand in hand. Like that is your computer. That is the thing. Like you don't have any other computer. You just have a quantum computer that is charged by AI.

48:43You also don't have a life because it's doing everything. Yeah. It's thinking for you. It's thinking for you. It's pooping for you. Right. It's running the robot that does all the physical work for you. Right. It's driving for you. It's doing everything. You don't need governments because it is the government.

49:01It's a computing version of WALL-E. Right. Remember in WALL-E? WALL-E, yeah. He was just this blob. They all sat around on hovering beds. Just there with nothing to do. Right. Yeah. You know, people are thinking about that. I mean, with our collaboration, we have people who are collaborating with people are building supercomputers and they know about GPUs. And I think that's the natural way to go. I think you want to think about a quantum computer as a coprocessor to a supercomputer, you know, with the GPUs and language models. But, yeah, people definitely looking in that direction. OK. OK. Yeah.

49:36Well, there you have it, Jabok. You got it, buddy. Hope you get to stick around until you don't have to do anything.

49:45All right. This is Matt Curtis, who says, hello, geniuses and Chuck. You know what, Matt? You know what, Matt? Meet me outside. I got your genius right here, buddy. Why is he insulting me already? So Matt says, hey, this is Matt from South Carolina here. Quantum computing has made strides in the number of qubits available and appears to be accelerating that number. What is the threshold at which quantum computing becomes useful for more than things like encryption and breaking and creating and other realms where the concept shows promise? We know the encryption part, but how do we go beyond that?

50:40And what can quantum computing do for us in the other realms? What about weather prediction, for example, where the systems are so complex? Yeah, I mean, right. You can only have certain possibilities because you have no idea how the fluid is going to act in the atmosphere. Well, the atmosphere is the fluid, but how it's going to act. Yeah, that's kind of a good question. You know, there's actually debate in the community whether you can build a small quantum computer, let's say a thousand qubits. If they're good enough, you can solve some problems. And other people say, no, you have to go to a million qubits and get the mirror corrected for a general purpose.

51:18So there's still a lot of debate. And it's kind of an interesting time because people are building things and testing it and trying to figure it out. I think the weather prediction is kind of an interesting application. People have talked about solving these differential occasions in some way. Fortunately, I don't know much about that. But, yeah, there's a lot of different interests. My personal view is that once we build a quantum computer that's better and bigger and start seeing these applications, then more people, creative people, will jump into the field and do it. It's kind of what happened with regular computers.

51:57Once you started building them, once it got more powerful, then all these ideas came out. And the internet itself as well. Exactly. Who would have thought? Yeah. Who would have thought? So, John, is there a limit to how many qubits can exist in one place? I don't think there is. It's a matter of engineering and practical consideration on how big you can make it and still not have it lose the energy. And, you know, right now we're at 100 ,000 and people, you know, I don't know where it's going to end. And what's the largest one as of this recording? There are qubit counts of about 100 or so in the superconducting case.

52:41And people in neutral atoms are now building thousands of qubits. And that's really exciting to see that. But besides the number, you also have to make them good. There's a lot of other things you have to worry about. But this is advancing very rapidly now. What's this we hear of a Google Willow chip and that it can do a calculation that would otherwise take 10 to the 25 years of a traditional computer to accomplish? Yeah, this is very similar to the results that I was involved in when I was at Google in 2019. and in the intermediate time, they made it bigger and they've made it better, less errors.

53:27So there's a very good development of the technology. This is very healthy for the field. The universe is only 10 to the 10 years old. And so to say that a traditional computer would take 10 to the 25 years, that's, I don't even know what we would have to compute to have to do it that fast. So, I mean, other than weather forecasting, what else needs that level of computing? Or someone clever person is going to say, here's something no one thought of because they couldn't have ever calculated it. And here it is. And now it's done routinely by quantum computers. Well, I know one. It could be like the mapping of the human brain, the neurosynaptic functions of the human brain are so varied.

54:15And there's so many of them. Like that'd be kind of a cool way to figure out. And these are applications that people are, you know, have to discover and work on. And the real problem right now is taking whatever quantum computer we have, which has some, you know, limits to its thing, and then taking the algorithms and try to match them together and do something useful. But as soon as you solve really useful problems all the time and the same money goes into the film, the firms, because they're solving useful problems, then you can even develop these more. This is what happened with conventional electronics.

54:53Well, kids, here's your here's the takeaway from this. Study physics because guess what? There's going to be no other jobs. OK, if you're not studying physics, you are wasting your time. Okay? Computers are going to do everything else. Nobody's going to be working except physicists. So you better do it. Physics and engineers. That's it. So one last question here before we land this plane. The discussions of whether we live in a simulation, could the complexity of our world be sort of a trivial calculation on a willow chip in some alien kid's basement? Yeah. Yeah, like if there was a supercomputer that was like the size of Manhattan, a super— Why make it big?

55:41Keep them little. Who cares? Well, no, because I'm saying that shows you the number of qubits that are at work. And they're all stabilized, and they're all doing their calculations. Gotcha. And so is that the kind of computer that's necessary to simulate our world? Our universe. Or even the universe, yeah. Yeah. So I would say if if you believe in simulation theory and now that we can do these really complex calculations of the quantum computer, that whatever is doing the simulation has to have a big quantum computer. So that's my conclusion about the simulation. OK. All right. So there it is.

56:15He's like, no matter what, all roads lead back to me.

56:20And even the computer that's simulated the computer has to be a quantum computer. Or at least if they have to have part of it has to be a quantum computer. It's quantum all the way down. That's what it is. Turtles all the way down. Well, thanks for taking time out of your day. And you're coming to us from your home in Santa Barbara. And this has been a delight. And can we keep you on speed dial in the future if we have like quantum confusion? Sure. I've had an enjoyable time. And I'm glad that it was such a fun conversation. So, yeah. Excellent. Excellent. He'll be all quantum man about town. There you go.

57:00Right. Quantum man all about town. All about town. In every part of the town at the same time. Oh!

57:09All right. Again, Professor, thank you. Thank you very much. It was a real pleasure. Chuck, always good to have you, man. Always a pleasure. Yeah. This has been StarTalk, the Nobel Prize edition. Yeah. 2025 on quantum tunneling, macroscopic quantum tunneling. Sweet. All right. I'm Neil deGrasse Tyson, your personal astrophysicist. As always, I bid you to keep looking up.