Pushkin.
In a world where you can put billions of transistors on a single chip, and where anybody can access trillions of transistors in the cloud, it feels like we can use computers for for, you know, anything that can be computed. Kind of amazingly, this is not true. There are lots of things that could theoretically be computed, but that in fact are just too complex, too hard to compute for
even a cloud full of computers. Things like predicting how prospective drugs will work in the body, or modeling really complex financial scenarios, even figuring out the factors of very large numbers. There are all kinds of computations that people just don't do because computers are not nearly powerful enough to do them. But but there is an entirely new kind of computer people are trying to build. It's called a quantum computer. You've probably heard of this. Quantum computers.
If they work like people hope, they will, will be profoundly more powerful than any cloud full of computers that has ever existed. They could allow for new breakthroughs in everything from drug design to energy storage. Also, a big quantum computer could factor very large numbers, which would allow it to crack the encryption codes that secure most of
the data that flows across the Internet. I'm Jacob Goldstein and this is What's Your Problem, the show where I talk to people who are trying to make technological progress. My guest today is Chris Munro. He's the co founder and chief scientist at IONQ, one of several companies that is trying to figure out how to build a big, power,
powerful quantum computer. Chris is also a physics professor at Duke, which was good news for me because a big part of what I wanted to talk to him about was quantum physics, which is of course amazing and weird and essential for understanding how quantum computers work. We started off by discussing the fundamental difference between quantum computers and regular computers. It comes down to the bit, the basic building block
of computing. The essential thing to know about a bit is that it can be in one of two states, on or off zero or one. At least that's the case for traditional computers. Quantum computers use bits called quantum bits or cubits, and they're different than traditional bits in a really strange, really profound way.
A quantum bit can be in both zero and one at the same time. So this is totally new. We never experience, you know, a coffee cup being in two places. I'm looking at a.
Coffee cup here, yes, or more simply a light switch that is both on and off. Right, plainly, the light is either on or.
You don't experience those things in everyday life. So here's my physics, my two minute description of quantum. Okay, there are exactly two rules, no more, no less. There are two rules, and they have almost nothing to do with each other. Here are the two rules. Okay. The first rule is that a quantum system can exist in multiple states two. It could be more, but let's just say two a cubit it's in both states at the same time.
And this is not this rule by itself. If you think about what's going on, it's not such a big deal. And the reason is a quantum system follows a wave equation sort of like water waves. That if I throw a pebble in the pond, it's going to emanate these circular waves that occupy the entire pond.
So you throw a rock or a ball into the pond a minute later wears the waves like the whole pond is way.
Yeah, that's right, that's right, it's not you know, the wave is not localized is kind of technically how we say something, So it's in many places. Now, that's rule number one. So here's rule number two, which is the weird one. Rule number two basically says that that wave, like that wavelike superposition that all quantum systems can can exist in, only works when you're not looking.
So yeah, that's the part that seems like dumb is not quite the right word, but lengthy, implausible, little goofy, like surely the world can't give a shit whether I'm looking at it or not.
Yeah, because what I mean, and what happens when you do look, and quantum says that when you do look the superposition, it pops, It latches onto a definite state, so it localizes, and it does it at random. You can't predict where it's going to localize. All you can say is, well, is supposed to be in these two positions, and I looked at it it was in one position.
More to the point, I mean, are like more to the point, as people are first thinking about this, are they they're thinking of an electron. So the electron is like a wave it kind of exists in all these different places around the nucleus. But then they figure out if you look at an electron what happens.
It pops into one state that almost has no space. I mean, it pops into a point like particle, like a little tiny baseball.
And if you're not looking at it, is it literally not? There? Is that?
Now you're getting philosophical on me.
Here, you did it. You're the one telling me these things.
That's right, Well, it's these two rules. We have to add that rule because.
So so yeah, so state the rules again, just to keep us on the rails here.
Rule number one, any quantum system like a cubit or an electron, can be in a superposition of states. It can be in this fuzzy existence.
It can be in two places at once. It can be on and off.
Rule number two is that that first rule only works when you don't look, and when you do look, it will randomly localize pop out into a singular place.
It will cease being on and off. It will become either on or off. Famously, the cat will stop being both dead and alive, and will either be dead or alive. And just to be clear, observed, doesn't just mean us looking at it, right, that's the kind of caricaturish version. But in fact it means like no particle can touch it or something like that. Right, So observed doesn't just mean seen by a person. It means interacted with by any part of it, by anything, even a single molecule
of air or a photon. One. All of that ruins our beautiful superposition and forces the thing to resolve. So there's two pieces of jargon that I always come across when I'm reading about quantum computers. One, I think you've just described one, this idea that quantum things can be both one and zero if we're not looking at them, is it right, that's called superpositions.
Yes, that's called That's a I would call it a plain old superposition. And you'll see what I.
Meanilla superposition, no sprints, right, Okay, good, So we've talked about superposition. There is this other piece of jargon that people always use when they are talking about quantum computing, and that is entanglement. So just like we'll get to what it means for the computer, but just like, in a basic way, what is entanglement.
What's interesting is that we're bringing Einstein in here. All the time. He had a famous paper in nineteen thirty five that he thought finally put the end to quantum mechanics. He said, and this is ridiculous because these two cubits, and I'm gonna, I'm going to use cubits are so simple. Here you have two cubits prepared in zero zero and one one. Then you separate them. They could be really
far apart. Nevertheless, when one person measures their cubit, they know immediately even before the even before light could traverse the distance between them, they know immediately the state of the other one. And that violates relativity. Another of Einstein said, right.
So Einstein has this paper if I as I understand it, where he basically says, look, if quantum mechanics is true, there will be these particles, whatever very small particles that behave in quantum ways, that become connected to each other in some strange way. They're not physically connected. You can put them as far apart as you want, you can
put them a million miles away. And if you look at one of the particles and it resolves itself into some position whatever, call it zero or one instantly, instantly, at that same instant, the other particle will resolve itself the same way, and surely that can't be true because that doesn't make any sense. The universe doesn't work that way, and therefore quantum physics is wrong. Question spooky action at a distance, he said, mocking list.
Yes, indeed, the title of the paper was a question can quantum mechanics be considered complete? In other words, is it right? There's something else there?
And he's saying, surely, if this could be true, the theory must not make any sense because this obviously can't be true.
Yeah, and he was wrong. He was famously wrong. It spent tested time and time again that this is exact. This is actually how nature works. We have to think of it this way.
And people in China did one recently right where they had one particle on the ground and an entangled particle on a satellite, and they looked at the particle on the ground and it resolved itself in whatever way, and at the same moment the particle on this I mean, I know that's the party trick version of it, but it's a good party trick.
Yeah. So entanglement is I like to think of it. You know, if we move on to quantum computing. I like to think of entanglement as sort of like a wire without any wires. Yeah, and when you're talking about computing, that's in fact what gives quantum computers its power, the ability to wire things together without having wires.
Yeah. So okay, so we've got these two idea, right. Superposition a thing can be in two states at once until you look at it, and entanglement meaning particles, and it's not just two, right, it can be multiple particles can be entangled such that when you look at one and it resolves into some state, you will immediately know what state all the others are resulting into. Right, These are the two intellectual tools we've got. Can we build a computer in our minds from these two tools?
Well, the good news is that there's a cool scaling law because when we go to three cubits, now there's eight states, it's zero one, yes, all three cube, all three bit numbers, there's eight of them. With four, there's sixteen, thirty two, and so on and so forth. So every time you add one cubit, you've just doubled the possibilities.
Yes, exponentially is an overused word, but this is an actual exponential thing.
And this is the structure of quantum computing that gives it its power to calculate things that we could never do using classical computer. It's this fundamental exponential gain.
I get more or less the theory of a quantum computer. It's also clear that quantum computers don't exist in a useful way yet, right So what do you and everybody in the field have to figure out to get from where we are now to having amazing quantum.
Well, in our case using these atoms, it's a simple matter of scaling. We routinely work with twenty to thirty right now. As soon as we get up to about one hundred, we're going to start to challenge will be well beyond what challenges supercomputers, and that's where the opportunities will happen. And the trick is, with one hundred cubits, you need to do a lot of stuff with them. You need to do many more operations. They have to live longer, they have to be even more isolated as
you add more. So it's a tricky scaling problem. Our technology we sort of we sort of know how to do this. Other technologies are still in the lab. I think they don't know the underlying physics of materials to make them clean enough to do. That doesn't mean we're out of the woods. The challenge here is what you said, is trying to isolate it the control. But these challenges have nothing to do with quantum. It has to do with how good of a vacuum, how goods your chip,
how good are your laser beams. These are all things that can be engineered.
So it's it's very you're like a construction guy, like your core problem is just building whatever a box to put an atom in so that the atom will be left alone to exist in its unobserved quantum stamp.
Yep, exactly. Quantum systems only exist that way rule number one, you know, without looking that meaning that they're nearly perfectly isolated. Yeah, so that's the hard part of building a big quantum computer.
So that's the theory. After the break the practice how to build a small.
Quantum computer form real. Now back to the show.
A bunch of companies are building quantum computers, big companies like IBM and Google and Microsoft, as well as a few smaller companies like ion Q. The different companies are trying different approaches. Some require super cold temperatures, in your absolute zero. Some use photons. Ion Q uses ions charged particles.
I asked Chris to walk me through how ion Q builds a quantum computer and so what so lest I mean your company is called ion Q, right, and so let's like tell me about tell me about your qubits? What what kind of atom? What kind of ion? Like what is it that you're using.
The first few generations were uturbium one seventy one. It's a very heavy yes, a heavy atom, the lower right part of the periodic table. And it turns out that it interacts with the lasers in a very simple way, I believe it or not. Okay, so that's what we use. It's a metal. So we have a little wire of uturbuum.
So you have a little piece of metal that's U turbium. I would if I were you, I would sell visors that said uturbuum.
That's not sure. Oh yeah, okay, all kinds of swag.
So okay, so you take this element, this metal U turbium. Go it is.
We have a little vacuum chamber. It's small. They're getting those are getting smaller. It's all at room temperature. By the way, we don't have to cool things as aggressive o. This little wire is inside that vacuum chamber, and there's a few black magic things that happen. It's nothing, nothing super fancy here. We blast a piece of that metal with a laser beam and what happens is we get a puff of metal in gas in vapor form. Its sublimates.
It goes straight from solid to gas without being a liquid.
In a vacuum chamber. So there's nothing else there, and so we get this puff of neutral atoms. Now we have these electric fields from chips that again, nothing really fancy there. The atoms don't see the electric field because the atoms are not charged electrically yet they're neutral. Now, when they float above the region where we want to hold them as ion, we send another laser beam that
removes an electron from each atom. We know exactly how to do that, very efficient and bam, it's now an ion and it says, hey, wait a minute, I'm now stuck. So they just sit there and we do something called laser cooling to bring them to rest.
Sogerat to be clear now that so it's an ion of uterbium and it's held because by electrical charge. That's your Cubit you got a cubit.
Now that's it.
Okay, So that part you've solved, right, Yeah.
And when you put many turbium mindes next to each other, they form a little crystal, an atomically perfect crystal. And you can see that when you shine a different laser on, they glow and you can see like stars in the sky, there's they're not randomly oriented there.
So now you've created a bunch of cubits. You've got them, as you say, like stars in the sky. They're sitting there. They're held in place by electrical charge. What has to happen next.
What people might not know is that an atomic clock is actually based on two levels inside of the caesium atom. It doesn't matter what the atom is.
Oh shit, I was already on uturbium.
And we have very similar levels in utbium that behave as our cubit. It's a pretty good atomic clock, very well defined states. And each each atom has its own cubit. Okay, so we can prepare them all in the state zero.
So what is it? I mean? I know an atomic clock is just a clock that's super accurate, But why does that matter here? I don't even know what it means.
On. Ah, Well, it has to do with the fact that two states in an atom, they have they generally have different energies, sort of like different orbits planets around the sun.
Okay, so it's like what level, what energy level is the electron at it's going between one.
And you can think of that, and those energy levels are incredibly well defined and they're exactly the same for two different atoms.
Okay, and so is this our zero in one? Tell me this is our zero and one.
That's it.
Okay. Uh So now we have our and if we're not looking, they're in superposition. It's always some probability were zero and one.
We start by preparing them all in a very boring state, all in the state zero. That's like initializing the systems, like clearing out your computer. It's putting all the registries into zero. Reboot rebooter, that's rebooting. Okay. After we prepare in zero, now we're going to make superpositions. We're going to different laser beams, going to drive the system halfway to the other level. That's sort of making a fifty
to fifty superposition. You can make as you can also entangle them, and the entanglement is based on the fact that these are ions, and they're like masses connected by springs. They vibrate together. So when you push on one, literally we push them in space. We push them around, and they interact with their neighbors, and they're very well defined ways to do this. Those are called gates.
Like I'm with you. When you've got your ions, I understand you're resetting them and they have their zero in one and then you do something with them, you mess with them in such a way that they become entangled exactly.
Yes, okay, shine lasers that shun these atoms just a little bit.
Love it. I love it that it's another laser. And so now do you have a quantum computer? Have you now? In this?
No? One? Last step? Okay, it's an easy one. It's very similar to remember the initialization step. We prepare everything in the state zero. The final step is to make a measurement. Well, that's the beauty of atoms in a vacuum is that we can send yet another laser beam, very similar to the other ones. And this laser beam, if the atom is in the state one, it will glow and we can collect that. It's like a star. It's really bright. You can see a single atom with
your naked eye in my laboratory. If it's in the state zero, it's dark. So we basically look for bright dark bright dark, and that's what we read. That's it. That's a quantum computer.
And so you have built one of those, built.
Several of them, Duke, We have six of them at I and Q. We've built nine different generations.
And the issue is they're just not big enough. They're not enough cubits to be that.
Yeah, we're at twenty to thirty right now. We need to get to sixty two hundred.
So as you've described it sounds easy. I get it, Like, if you can do twenty, why can't you do sixty?
Okay, I'll make it very short. Remember the vibration I talked about, like when you move one atom, the other one moves. If you put five hundred of these atoms in a chain, that motion becomes very sloppy and noisy. So you want to limit the number you have in a chain. We know we can put twenty or thirty, maybe forty in a chain, but we need to we need to have a modular way to connect to another group of twenty or forty using optical fibers. We know how to do that too.
We have done it, but you haven't quite worked out all the bugs yet. Okay, so it's not there. Doesn't need to be some big breakthrough or some step functions in.
The engineer just have to want it's engine a lot of engineering, and we don't need breakthroughs. We don't need breakthroughs.
We'll be back in a minute with the lighting ground. Back to the show. I know you got to go soon. I want to finish with a lightning round and we can truly make it a lightning round a lot of questions and you can answer them fast so you could go to your next meeting. I understand you're a percussionist in an orchestra and that you play like the weird instruments.
Right.
I watched a YouTube video. I'm curious, what is your favorite weird percussion instrument.
I'm gonna have to say, alephone.
What's the aleophone?
It's a wind machine?
How does it sound? What sound does it make?
Actually, some orchestras serve pieces called for it. It's basically a big piece of burlap on a spindle that has slats and it's like I made one of those for our orchestra those fun.
So how long have you been working on quantum computer?
Thirty years a long time?
Did you think there would be a useful quantum computer sooner than there has been? Like? How is it relative to your initial expectations.
What I've learned is the love that you know. Physics in the laboratory is one thing. Engineering is yet another thing, and then product is a third thing, and they're all different. And when I started in the field, I had no knowledge of product and a little bit of knowledge of engineering. I think given engineering and product nature of this of this evolution, it's actually been faster than I thought it would be.
I'm curious just in terms of your work in quantum physics, right, this is you as physics professor. Now, when you're deep in quantum physics work, does it ever freak you out that quantum physics suggests that the world is so different than the world we experience in our daily.
I should be losing more sleep over it, But I think I've been able to be kind of successful by not thinking, at least during the day of those things.
You're too busy trying to build a computer to think about that stuff.
We know the law I'm more of a mechanic. I've always been very good at applying math, and to me, that's sort of what quantum is. Just it works. We know the laws, don't think too much.
You mentioned when you go home you think about it, Like when you wake up at four in the morning, is there a particular aspect of the quantum universe that you returned to.
Yeah, it's certainly entanglement that it's sort of space seems to wind upon itself. I guess that's one way to think about it.
Entanglement shouldn't that shouldn't happen, right, Like that's basically what Einstein said, and like he seems right, or there's some huge thing about the universe that we obviously don't understand. I guess that's the other thing.
It's so fundamental too, you know that. I mean, what is gravity. There's no microscopic model for gravity. It's just space is distorted so that things sort of fall toward each other. So and entanglement. It's really tantalizing to link those two. Entanglement to gravity. It hasn't happened yet, but that you know, that's an outstanding question. A lot of really smart people are thinking about this. Yeah, that keep I think about that sometimes it's fun.
Chris Munroe is the co founder and chief scientist at ion Q. Today's show was produced by Edith Russello and Gabriel Hunter Chang. It was edited by Sarah Nix, and it was engineered by Amanda ka Wong. You can email us at problem at pushkin dot fm, or you can find me on Twitter at Jacob Boldstein. I'm Jacob Boldstein, and we'll be back next week with another episode of What's Your Problem.