177 | Monika Schleier-Smith on Cold Atoms and Emergent Spacetime

When it comes to thinking about quantum mechanics, there are levels. One level is shut-up-and-calculate: find a wave function, square it to get a probability. One level is foundational: dig deeply into the underlying ontology. But there’s a level in between, long neglected but recently coming to life. In this level you think about — or do experiments with — entangled quantum systems in the real world, putting entanglement to use. Monika Schleier-Smith is an experimental physicist specializing in cold atoms, which can be both entangled and manipulated. We discuss how to use such systems to study everything from metrology to quantum gravity.

Support Mindscape on Patreon.

Monika Schleier-Smith received her Ph.D. in physics from the Massachusetts Institute of Technology. She is currently an Associate Professor of Physics at Stanford University. Among her awards are a MacArthur Fellowship, a Sloan Fellowship, and the I. I. Rabi Prize in Atomic, Molecular, and Optical Physics from the American Physical Society.

• Web site
• Stanford web page
• Google Scholar publications
• Wikipedia

[accordion clicktoclose=”true”][accordion-item tag=”p” state=closed title=”Click to Show Episode Transcript”]Click above to close. 0:00:00.0 Sean Carroll: Hello, everybody. Welcome to the Mindscape Podcast. I’m your host, Sean Carroll. And I do realize that by reading my books, listening to my lectures, even listening to the Mindscape Podcast, you can get an impression of quantum mechanics and in modern physics more generally, that is a little way out, right. We’re talking about parallel universes or quantum gravity, emergent spacetime, evaporating black holes. It may seem a little bit removed from the nitty-gritty of not just experimental physics, but also of your everyday experience.

0:00:35.2 SC: But of course, quantum mechanics itself was not invented by theoretical physicists just trying to think of cool things, they were forced to come up with these crazy ideas by trying to explain the data, by trying to explain experiments. And these experiments have not stopped, they’ve not gone away, and I’m not talking about experiments that use quantum mechanics, those experiments have also been going on a long time, every particle physics experiment like the Large Hadron Collider uses quantum mechanics to make the predictions. I’m talking about a new generation of quantum mechanics experiments that really digs into entanglement. Quantum entanglement is really one of the things that is special about quantum mechanics, makes it very different from the classical world, the idea that different physical systems can be related to each other in some deep quantum way.

0:01:23.9 SC: What can we do with that? We can build quantum computers. We can do other things as well. So today’s guest, Monika Schleier-Smith, is an experimental physicist who works on cold atoms, and the reason why cold atoms are really interesting is because when they’re cold you can entangle them and you can control the amount of entanglement, you can shoot little photons at them, and you can sort of manipulate them in a very, very delicate way. So ordinary atoms that are hot are just bumping into each other randomly, they will become entangled, but then unentangled and you don’t know, and you can’t really control it. When you really cool them down, you have this pinpoint precision control, so you can do a lot of different things, you can do sort of down-to-earth things, which I think was the original motivation, metrology, measuring things to exquisite precision.

0:02:12.2 SC: But then guess what, the crazy theorists have come in and ruined everything by pointing out you can also build models of quantum gravity. And this is one of the things that Monika does. She, in her lab at Stanford, builds groups of atoms that are entangled with each other in the right way that… Well, it’s just beginning, I wouldn’t say that we’re there yet, but it’s resembling the features of a model of holographic duality, the famous AdS/CFT conjecture that Juan Maldacena put forward almost 25 years ago now, where you could have a theory without gravity and a theory with gravity and secretly, they’re the same. Likewise, Monika can build collections of entangled atoms that if you look at them in exactly the right way resemble a system with gravity, and then you can use what you know about gravity to make predictions about what those atoms are going to do.

0:03:04.2 SC: It’s still beginning stuff, very, very cutting edge, but the hope is we’ll both just learn more about quantum mechanics and how to manipulate entanglement and collections of atoms, but maybe also learn a little bit about how gravity and spacetime do emerge from these collections of atoms. So the real lesson here is that there’s no real clear, sharp distinction between the down-to-earth useful applications of quantum mechanics and the pie-in-the-sky theorizing that me and my friends like to do so much. So let’s go.

[music]

0:03:39.9 SC: Monika Schleier-Smith, welcome to the Mindscape Podcast.

0:03:56.2 Monika Schleier-Smith: Thanks for having me.

0:03:57.5 SC: So I want to start with a slightly left field question, because you’re an experimental physicist, I’m a theoretical physicist, in certain circles, there’s a feeling that theorists grab all the glory, they hog all the credit, like experimentalists are in there doing the hard work. Most physicists are experimenters, many more than theorists, but it’s the theorists who end up writing the books and being on TV and things like that. Is this a feeling that you either have or get in your everyday work or is in your world it’s happy collaborations all around?

0:04:33.4 MS: Yeah, that is actually not a feeling I’ve gotten. I do remember one of my college professors saying everybody comes into college thinking they want to be a theoretical physicist, and nobody thinks to do experiments, and there’s something to the idea of doing experiments, but no, often… I often tell my students, one of the things I love about being an experimentalist is if I can have an idea, I don’t need to convince, and it’s an idea that merits experimentation, I don’t need to convince somebody else to do the experiment, it’s on me…

0:05:04.3 SC: You could just do it.

0:05:04.5 MS: To do it in my own lab. And so I think that’s… And technology is always advancing, so that’s something that we can take advantage of, so, yeah, I think it’s wonderful being an experimentalist and getting to actually kind of do things in the lab. The data sometimes just speak for themselves in a way that you don’t have to argue whose idea it was. If you did the experiment, you did the experiment.

0:05:27.2 SC: And did you start off wanting to be an experimentalist, or did that come to you at some point?

0:05:33.2 MS: Yeah, that’s good question. So I think if I sort of think back to being a freshman in college, I was interested in… I actually loved abstract math courses where I stayed up all night trying to prove something, that kind of thing, so I certainly have had that theoretical bent and also enjoyed physics and chemistry. And I think that for me, I wanted to go into a field of physics, once I realized it was physics I wanted to do, where I could maybe think about some interesting ideas, but not have my whole job be to think. And so what I love about my job is there’s sort of the hands-on part, these days it’s more the graduate students and postdocs doing the hands-on part…

0:06:17.8 SC: Alas, yes.

0:06:19.3 MS: But there’s a varied aspect to it. We get to do a little bit of theory on the side, but the whole job isn’t to think.

0:06:23.5 SC: And you’re doing experiments in the realm of quantum mechanics, which is great, because I’ve talked about quantum mechanics on the podcast before. Most listeners know that this is one of my things, but I would love to hear how you explain quantum mechanics to the person on the street. I want to say as an experimentalist, but just explain it however you would explain it. I’m curious as to whether or not we think about it differently.

0:06:45.2 MS: Yeah, so to me, the most kind of remarkable and revolutionary aspect of quantum mechanics is the concept that information is not something that has to exist locally. So classically, when you think about information in your computer, it’s stored in individual bits that are ones or zeros, and in quantum mechanics, first of all, we can have the concept of superposition, where a quantum bit is not just a zero or one, but it can be in a state where until you measure it, it’s somehow undecided what state it’s in, and there’s some randomness potentially in the measurement outcome, so that that randomness is part of what’s special about quantum mechanics.

0:07:26.7 MS: I sometimes like to use the analogy of a coin toss, so when we do that measurement, it’s like tossing a coin that could land heads or tails. But the cool thing about quantum mechanics is also that there can be correlations in the randomness, so I can be tossing a coin here, and you can toss a coin there and we could arrange a situation where every time I get heads, you also get heads, and every time I get tails, you get tails, and that would never happen with an actual coin toss, but that could happen with measurements on a pair of qubits or quantum bits, and that tells us that there’s some information that isn’t your information, where are you? Pasadena?

0:08:08.9 SC: Right down in Boston, actually.

0:08:10.4 MS: You’re in Boston, so it’s not in Boston. It’s not in Palo Alto, right? Because everything I’m measuring looks completely random, but when we actually compare notes, we realize there’s some information there, so there’s this sort of information hidden in correlations, and that phenomenon known as entanglement is what’s really special about quantum mechanics. And I find that kind of amazing that information can exist in this delocalized fashion.

0:08:34.2 SC: Yeah, I think that this is exactly right. So let me run by you something that has been on my mind, because I actually am beginning to write a textbook, undergraduate level textbook, on quantum mechanics, reading other people’s textbooks, and I’m struck by how little they talk about entanglement, they just solve the Schrödinger equation over and over again, and the fact that there is entanglement is mentioned, but then it’s breezed right on by. Am I right that the modern frontier of experimental quantum mechanics as well as theoretical is very entanglement-centric?

0:09:06.6 MS: Absolutely, and you mentioned writing a textbook. I haven’t written books, but I teach a course for freshmen on quantum mechanics, on quantum information. And that’s actually part of what I have found amazing in teaching this course is that you can sort of start talking about entanglement from day one, and certainly really start to get an understanding of its implications in a couple of weeks. And actually at week seven, we get to the Schrödinger equation, but we sort of flip it around and it lets you get to some of the really cutting-edge aspects of what’s special about quantum mechanics.

0:09:40.4 SC: Yeah. It makes perfect sense to me. So yeah, I’m going to try to revolutionize the teaching of undergrad quantum mechanics by putting entanglement front and center a lot quicker. But I need to ask now, do you have strong feelings about the interpretation or foundations of quantum mechanics?

0:09:58.2 MS: I think that that… It’s like, it’s a fascinating topic, it’s great for after-dinner conversations, and if you get a bunch of physicists around the table, everyone will have heated arguments. For me, it’s sort of… At the end of the day, so far, there is a theory that predicts very well everything we do in the lab. Sometimes what we do in the lab doesn’t match the theory, but then it’s usually we didn’t do what we thought we did, and we need to track down in some source of experimental error.

0:10:29.5 MS: And so certain aspects, things like the Heisenberg uncertainty principle, this idea that you can’t know where something is and how fast it’s moving at the same time, when you do experiments that deal directly with quantum uncertainty, you really develop an intuition for where it comes from and what are the mechanisms that enforce that principle. And so I think that there are lots of things that are sort of mysterious about quantum mechanics and one should stop and be bothered about it and think about interpretations, but it’s not something I have strong feelings about. There might be some interpretations that I would take issue with, but… Yeah.

0:11:11.0 SC: Well, next time we’re at the same conference, we’ll have dinner together, we can talk about the interpretation of quantum mechanics, I can give you a pitch for that. But then, you mentioned the idea of measurement. I guess the three big ideas that you mentioned, all of which I agree with, that come in when you start talking about quantum mechanics, superposition, measurement and entanglement, right. So is there a simple explanation for why we don’t observe all this weird quantumness in our everyday life? You did the coin flip example, but then you were quick to say, not with real coins. Of course, with qubits. So why not? If the real world is quantum mechanical, why don’t we observe all these things?

0:11:49.7 MS: Yeah, I think a good example is perhaps the Heisenberg uncertainty principle that I mentioned. So there’s a fundamental limit on how well you can know where a thing is and how fast it’s moving, its position and its momentum, but for sort of macroscopic objects that we encounter in our everyday lives, those uncertainties are tiny compared to what we can see with our eye, compared to motion that has to do with that object, is at some finite temperature. So usually in sort of our day-to-day lives, we’re not observing things at the scale where you would see these quantum phenomena. And again, when you do experiments, you often really deal first-hand, or even try to plan experiments, you deal first-hand with sort of why it is hard to scale up these phenomena like quantum superposition…

0:12:44.5 SC: I guess. Yeah, that’s really what I was going to try to get at because you gave a very sensible answer about using the uncertainty principle, to why quantum uncertainties are there for individual particles, etcetera, but for an experimentalist like yourself, isn’t much of your job trying to make quantum systems bigger and still be quantum?

0:13:06.4 MS: Yeah, yeah. And so one thing that I have, for example, thought about a fair amount, and I haven’t done an experiment like this, but there’s various groups doing experiments along these lines, is how would you make what you might call a Schrödinger cat state, so a quantum superposition of a cat being alive and dead. [chuckle] And okay, so we like to simplify that down a little bit as physicists, so rather than a superposition of a cat being alive or dead, you could ask the question, can I make a superposition of a group of atoms, either all being in their ground states, or all being in their excited states? And fundamentally, it’s unknown which of those scenarios is the one we’re in until we do a measurement, right?

0:13:48.5 SC: Right.

0:13:49.9 MS: That actually turns out to be a state that is potentially very useful, if you can make it. It’s a resource for improving precision measurements of time. So why is that hard to do? You need to somehow, first of all, manipulate this collection of many atoms. Actually, let me give an example. You could imagine maybe you can do this by having… I sometimes like to think of the ground state and the excited state as a spin that points down, or a spin that points up.

0:14:22.8 SC: Good.

0:14:23.3 MS: Or equivalently, maybe a spin that points left, and a spin that points right, but two opposite orientations.

0:14:27.5 SC: Probably for a lot of people, the spin pointing up or down is… It makes more sense intuitively to them than an excited state and non-excited state.

0:14:35.0 MS: Yeah. Yeah, exactly. And so you could imagine maybe trying to do an experiment where I send some light that interacts with a group of atoms, maybe I send a photon. If the photon is in one polarization state it will rotate the spins one way, if it’s in a different polarization state it’ll rotate the spins another way, for example. And then if I… But in that type of a system I could create something that’s the superposition of the spins oriented in one direction or the other, and maybe I can just scale this up and do this with more and more spins, or more and more atoms and make more and more macroscopic superposition states. And then you run into the problem that the larger I try to make my system, the higher the probability that one atom does something it shouldn’t do.

0:15:26.0 SC: Or you don’t want it to do.

0:15:28.8 MS: Like it decays from an excited state to the ground state and there’s a photon that leaves and hits, and in principle, could be observed by somebody, and that would give that observer information about the state of that one atom, and that would, so to speak, collapse the entire superposition state. So basically, the bigger you make the system, the harder it is to control what you don’t know about it.

0:15:49.1 SC: Right.

0:15:50.4 MS: And this idea of superposition, it’s all about having this kind of quantum system where there’s something unknown about it until you perform the measurement, so there’s that risk that you accidentally let some information leak to the environment that destroys that delicate superposition state. And so that’s something that when you plan an experiment you can immediately see why it gets hard.

0:16:11.8 SC: But you just said that you have contemplated this, but you haven’t actually done this particular experiment, trying to put as many entangled particles together?

0:16:19.4 MS: No, no. In part, because it’s hard. [laughter]

0:16:21.3 SC: It’s hard. [laughter] Yeah, that’s a good reason.

0:16:22.8 MS: But there are the state-of-the-art… There are a couple of physical systems where people have done things like this for let’s say around 20 spins, right? So that’s kind of the state of the art.

0:16:31.4 SC: Oh, okay, that was… I was just going to ask…

0:16:32.3 MS: So you can go beyond just a single spin or qubit in a superposition of zero and one, you could scale it up to about 20, and I’m sure this will keep getting pushed. Yeah, but that…

0:16:43.4 SC: Well, this is what we need to push for quantum computers, right?

0:16:46.3 MS: Exactly.

0:16:46.3 SC: When you say it’s 20 qubits, that’s a 20-qubit quantum computer, is we haven’t had anything more than that, I guess, is what you’re saying?

0:16:56.3 MS: Yeah. I mean, I guess, you know… If you read the news, you know there are 50-ish qubit quantum computers, depending what you define as a quantum computer, but yeah. Yeah, so.

0:17:07.8 SC: And what exactly is the kind of system that you’re working with? You have some kind of conceptual things, but what happens in your lab when you’re pushing these quantum systems around?

0:17:16.8 MS: Good. Yeah, great. So what do we do in my lab? First of all, the general system I work with are systems of laser-cooled atoms. And so perhaps it’s a little counterintuitive, but we can use lasers to bring atoms to temperatures that are, in our lab, millions of a degree above absolute zero.

0:17:35.1 SC: Maybe it’s worth explaining a little bit about that. When I think of shooting a laser at something, it sounds like I’m going to heat it up, but you’re somehow cooling it down?

0:17:41.3 MS: Right. Yeah, yeah, exactly. There’s a trick we use, which basically relies on the Doppler effect. I wish I could wave my arms and the listeners of the podcast could see it.

[laughter]

0:17:54.3 SC: They can imagine.

0:17:54.8 MS: But the basic picture you can have in mind is we have some atoms flying around, they’re initially at room temperature, and we have laser beams, basically, that are counter-propagating. So if we want to slow the atom down in a particular direction, we would have two laser beams that are pointing in opposite directions, and instead of tuning those laser beams so that they can resonantly be absorbed by the atoms, we tune them so that they’re like a little bit too low in energy, or too low in frequency to be on resonance with the atoms.

0:18:30.4 MS: But there’s this phenomenon of the Doppler effect. It’s the same thing where if I’m driving along the highway and a truck comes the other way, the faster it’s moving, the more its frequency will be shifted upward, sound higher pitched. In the same way to the atom, depending which way it’s moving, the laser beam will seem like it’s higher pitched, or at a higher frequency, and this results in a phenomenon where the atom is more likely to absorb a laser beam, that’s propagating in a way that will kick it to slow it down, than a laser beam that’s propagating in the other direction.

0:19:04.1 MS: So that basically means you get this preferential absorption of photons that slow down the atom, give them the momentum kick that slows down the atom, and then the photon, it’s slowed down, and now it emits a photon, and actually, that emitted photon then has a higher energy than the one that was absorbed, because the atom is now… If it’s sitting still… Yeah, if it’s gone from moving to sitting still the energy it emits will be different. So that gives you a way to extract energy from the atoms and sort of put it into the light field, and also the atom emits some photon in a random direction, so actually the entropy of the light field goes up, there’s some sort of disorder that’s getting pumped into the light field and the atoms can be more ordered then and ultimately sitting still.

0:19:54.1 SC: So just to I make sure I get this, I’m a perfect podcast host for this, ’cause I actually don’t know anything you’re talking about. I mean, I’ve heard colloquy before, but I’m not an expert, so I’m learning here in real time. So the great thing about the procedure you just outlined is that it’s not like you’re trying to be Maxwell’s demon and observing the velocity of all the atoms, you’re flooding them with light with the property that they will, if they’re moving too fast in some direction, they will absorb a bit of light and slow down, and then they’ll kick out another photon and you’re left with the very bunch of cold atoms.

0:20:28.6 MS: That’s right, yeah.

0:20:30.0 SC: Okay, and what kind of atoms are we talking about? What element?

0:20:33.7 MS: So in my labs, we work with either rubidium or cesium, those are both in the first column of the periodic table, so they have one valence electron, which gives you sort of a relatively simple atomic structure to… Electronic structure to work with, they have sort of convenient laser frequencies that we can use to address them, yeah, so.

0:20:55.0 SC: And how many atoms are in a little group that you’re working with at any one time?

0:21:00.2 MS: Yeah, so typically, this starting point of laser cooling gives some cloud that could easily have 10 million atoms in it, let’s say, but we don’t… And then that’s sort of the first stage of our experiments. And then I have a couple of different labs that do different things, but in… Actually, one of the things we’ve been doing is doing experiments where we have little clouds of atoms that are individually trapped and asking how much control can we get over the ways that these atoms can actually interact with each other. And sort of the general motivation is we talked earlier about entanglement, and in order to generate entanglement, we need some way for the atoms to interact.

0:21:45.7 MS: Now, ultimately, the sort of conceptually simplest thing one might want to do, and there are some labs that do this, is have sort of an individual atom trapped at the focus of a laser beam and have perhaps some array of those individual atoms so that you can kind of control, think of each one as a qubit and form some kind of a quantum register.

0:22:05.9 SC: So they have every single atom pinned at one location?

0:22:09.2 MS: Yeah, exactly. In things that I’m currently doing in my lab, we work with little clouds of atoms, each one pinned in a focused laser beam, and that for what we’re doing right now happens to have some advantages, but that’s kind of what you can picture is sort of a bunch of little clouds of atoms sitting in kind of a line, trapped one next to the other.

0:22:30.5 SC: And then how many atoms are we left with or how many effective qubits?

0:22:33.7 MS: So each little cloud has a few thousand atoms, and we can have, let’s say, 20 of these little clouds, yeah.

0:22:41.9 SC: Okay. Alright, so the individual clouds, does it even make sense to ask this question, are they solid or gas?

0:22:52.7 MS: Yeah, yeah, yeah, I think… They’re gas… So the atoms within the individual clouds, they can move around freely there to lowest order, not really interacting with each other, they’re far apart compared to the size of the atom, so they’re sort of at micron scale distances apart. Whereas the size of the atom is on the angstrom scale.

0:23:15.9 SC: And are they like sitting in a bowl or something, what prevents them from just falling to the floor?

0:23:20.6 MS: Yeah, and so that’s where lasers come in again, so so much of what we do involves using lasers, so the first step I said was cooling them, but also then we hold them in place, we essentially are levitating them. Yeah, essentially trapping them by using the fact that the atom can basically experience an attractive force when it’s… Or a force that attracts to the intensity maximum of the laser beam. You can kind of think of it as you have this oscillating electric field of the light, which can polarize the atom. And so that allows the atom to… So the light sort of induces some polarization in the atom that’s along the field direction, so it can lower its energy by being in the presence of that oscillating field, which is the light.

0:24:13.3 SC: Maybe one somewhat down-to-earth question here is, what does the lab look like? I think the most people have an image of the Large Hadron Collider as a physics lab, and I think it’s a different scale that you’re working on here.

0:24:25.3 MS: Yeah, that’s a great question. So the typical scale, first of all, in terms of sort of the number of people who are operating such a machine where we laser cool and trap the atoms and do our physics, it’s typically three or a team of three or four people working together…

0:24:41.6 SC: Very different.

0:24:42.8 MS: And they’re in a room that’s something like 500 square feet, to give us kind of a sense of scale, and so they will have… You can picture basically having several optical tables, so imagine a table that’s something like 5 feet by 9 feet or so, that has on it… Well, so in a typical lab, we might have one table that has a vacuum chamber where our experiments happen, so we need the atoms, these atoms, these particular atoms we care about, let’s say the rubidium atoms, to be isolated in ultra high vacuum without anything else…

0:25:22.8 SC: You don’t want them to bump into air molecules or whatever, right.

0:25:28.1 MS: And so there’s that vacuum chamber, all the pumps needed to create the vacuum, and then various… Lots and lots of kind of optical components, so laser beams that are used to do the cooling and the trapping and manipulating the atoms, which we usually prepare this laser light on another table, send it into optical fibers that carry it to where the experiments happen, so you kind of… You can adjust something at the laser end without needing to readjust everything at the vacuum chamber side. And so there’s… Yeah. So there’s sort of a couple of these tables with lasers on one, the vacuum system on the other, lots of electronic components, often kind of home-built to do what we need to do.

0:26:10.3 MS: So every laser needs to be at exactly the right frequency, and we’re constantly measuring and feeding back to make sure it stays at the right frequency. So lots of feedback loops and things like that.

0:26:22.2 SC: It sounds like a lot of the day-to-day work of one of your graduate students is tending the lasers. [chuckle]

0:26:29.0 MS: Well, that actually depends a lot on kind of the stage we’re at in our research. So I would say in sort of the early stages, there’s kind of, really, a custom-built apparatus that the graduate students, they need to design and build and set up the lasers and get them, all the light into the optical fibers and at the right frequencies, and then eventually, the graduate students aren’t pressing buttons to turn on and off the light. That is all automated. So they’re writing computer code that tells all of the lasers, for example, what they need to do in the magnetic fields and so forth. And so then the day-to-day life becomes more about writing that script that’s telling the apparatus what to do, analyzing the data, which is some images, let’s say, from a camera that tell us what the atoms are doing. Yeah. And so that… There’s sort of a shift then to kind of sitting at a computer and making sure and telling the experiment what to do, analyzing data, and every so often, perhaps, going and fixing something on the experiment table.

0:27:28.8 SC: And the ultimate goal is we want to entangle these things. So what is it that is entangled? Is it individual atoms within a cloud that you’re entangling with each other, or are you entangling different clouds?

0:27:40.7 MS: Yeah. So one of the things we’re currently very interested in is how one can have some kind of programmability of what the sort of graph is of entanglement that we can create. And I will say that in my lab, we’re actually currently working on, can we actually quantify and prove that there’s entanglement in our system, and that takes on a fairly sophisticated set of measurements that are in progress. But what we’ve done so far is show that we have a high degree of control over, basically, the structure of interactions, and those interactions are the mechanism for generating entanglement.

0:28:17.2 MS: And so what we are most interested in… Well, yeah. And so both of these things are interesting: Generating entanglement within a cloud, generating entanglement between the clouds. What you would care about depends a little bit on what application you have in mind. And for us, there are a few different directions that we’re intrigued by, which range from sort of preparing states that could have applications in precision measurement or in computation or simulating phenomena from other areas of physics. And one that I’m intrigued by is connections to gravity that you actually probably know much more about than I do.

0:28:53.2 SC: We’re going to get there, believe me. I’m just…

0:28:55.1 MS: But yeah, so… And so, again, so depending which of those things you want to do, you might want to entangle things in different ways.

0:29:00.2 SC: Yeah. No, I’m just luxuriating in all the vicarious pleasures of being in the lab without actually doing any of the work. So how does the entanglement come about? Do the atoms bump into each other or interact with each other, or do you… Does some separate manipulation get them into an entangled state?

0:29:16.5 MS: Yeah. So one way, generically, one way to entangle atoms is actually to let them bump into each other. And that is something that is not the focus in our lab, and the reason that we’re kind of interested in actually going beyond that and being able to let atoms interact that aren’t directly bumping into each other is related to this idea I mentioned right at the outset, that the special thing about quantum mechanics is that information doesn’t have to be local, that it can be stored in these kind of non-local correlations. And if you want to kind of efficiently build up non-local correlations, then it would be great to be able to have kind of interactions that don’t rely on things being right next to each other.

0:29:56.3 MS: And so one of the things that we do in our lab is actually use, again, use light, so use photons to carry information between atoms that are far apart. So the atoms, again, are kind of angstrom-scale objects, but we can have a photon convey information from one atom to another atom that’s a millimeter away or from one little cloud to another cloud that’s a millimeter away. And so for us, that’s a very long distance.

0:30:22.0 SC: Exactly, yeah.

0:30:23.0 MS: On the scale of our experiments. And so that’s… Yeah. That’s one of the key approaches that we use in our lab.

0:30:30.0 SC: And you already alluded to this, but could I get more details on this question of how do you know that things are entangled? Is it just you trust the rules of quantum mechanics? ‘Cause my impression is that entanglement itself is not measurable, at least in a single measurement. Maybe you could sort of do it over and over again.

0:30:47.3 MS: Yeah. And so in the experiment, you always need to basically prepare the same state many times and do a set of measurements that… So that allows you to do kind of complementary measurements that give you kind of more… Yeah, more than one piece of information about the same state, even though the measurement does fundamentally change the state. So that’s why we need to redo the same experiment and measure a different quantity.

0:31:12.8 MS: One of the ways that we are actually currently working on is using some insight from kind of the field of precision measurement. So there are known fundamental limits having to do with the Heisenberg uncertainty principle that I mentioned before, where if I have a collection of atoms that are unentangled, what is the best measurement that I can do? In our case, we’re not looking at position or momentum, but you could ask, for example, what is the best measurement I can do of the strength of a magnetic field that makes these spins that I… I tend to think of these atoms as spins, that makes these spins rotate.

0:31:52.6 MS: And again, this phenomenon of quantum superposition means that when we measure the state of an atom, it’s kind of like a coin toss. There’s some randomness. And we know if we were to flip 100 coins, on average, 50 would land heads and 50 would land tails, but there’d be some fluctuations around that. And that just comes from the statistics of a binomial distribution, so. And so those fluctuations scale roughly with the square root of the number of coins you tossed or in our case, the square root of the number of atoms that we did our measurement on.

0:32:24.1 MS: So that sets a limit to how precisely you could measure a magnetic field using the atoms, and if we can do better than that limit, then actually that is one way in a very sort of operatively useful way saying there’s entanglement in the system, so that’s the type of measurement that we’re currently working on, it requires making sure all the technical noise is not a limitation and that you’re really just seeing the quantum noise. Yeah, and so far, we’ve seen kind of evidence that we have the right structure of interactions to give rise to entanglement and we’re working on showing is it really entanglement? And like you said, that requires performing measurements of the spin in different directions, so if there’s a full set of measurements, what needs to do.

0:33:08.0 SC: So just to be clear to the people who are not quantum mechanics experts here. If I have an atom, a single one, that may or may not be entangled with other atoms, there is literally no measurement I could make on just that one atom that would tell me whether it was entangled. Is that right?

0:33:23.5 MS: Can you, sorry, can you repeat the question one more time?

0:33:26.2 SC: If I focus in on one atom that may or may not be entangled with others, and I think about measuring just that one atom, there’s nothing I can measure that would tell me whether there was entanglement with it elsewhere.

0:33:37.3 MS: That’s exactly right, yes. Yeah.

0:33:39.2 SC: Okay. Yeah, so you got be clever, this is why you get paid the big bucks.

0:33:43.2 MS: If we had that state we described before, the so-called Bell state, where it’s a superposition of both coins being heads and both coins being tails, if we just look at one of them, everything looks random, we can’t tell… And that’s generally true, that I need to somehow look at both parts of the system.

0:34:00.6 SC: And I got the impression from looking at your website, etcetera, that one of the things that you are specializing in is the fact that it’s relatively easy to get two atoms to be entangled if they’re right next to each other, but you’re trying to be separating the sort of physical proximity of different atoms from their entanglement, so you can have different kinds of entanglement structures. I guess… Maybe explain what I mean when I just said that, what does it mean to have different kinds of entanglement structures? Can an atom be entangled with many other things, or is it just one at a time?

0:34:32.0 MS: Oh, yeah, that’s a great question. First of all, there is sort of a rule called monogamy of entanglement which says, you know, if I can have two things that are kind of maximally entangled or I can have many things that are all sort of weakly entangled. So there are some trade-offs there. One thing that… And yeah, as you said, we’d like kind of control over the structures of entanglement, one thing I’ve worked on in the past is having sort of the opposite limit of two things which are strongly entangled, which is having many that are sort of collectively entangled.

0:35:05.3 MS: So if I have a cloud of atoms and every atom can talk to every other atom, that gives a way of making a certain type of kind of collective entangled state that does have applications actually in enhanced precision measurements. And that is actually something I worked on in my PhD thesis, was using that type of collective entanglement to make states that are useful for enhancing the precision of atomic clocks. So that’s kind of one simple limit, is like everybody is talking to each other and there’s a collective form of entanglement all to all. Another limit is sort of maybe pairwise entanglement that you could get by having some nearest neighbor interactions. And what we would love to do is kind of be able to explore between those two and really control the structure of interactions and explore a wider range of quantum states.

0:35:55.2 SC: And did you ever expect when you started doing this kind of thing that you would be modeling quantum gravity?

0:36:00.5 MS: No.

0:36:00.6 SC: But it happened, apparently that is what is going on. I’m not sure what the best way to approach this is. Do we have to talk about AdS/CFT?

0:36:15.0 MS: Perhaps. Yeah, I could say a little bit about how I got interested in this topic.

0:36:20.6 SC: Yeah, why don’t we do that.

0:36:21.0 MS: I think maybe… Again, so my background had been in this area of quantum metrology, I was building a lab where I have ways of letting atoms talk to each other with light that’s a little bit different from letting atoms collide with each other, and somewhere along… Some time when we were working on getting the lab set up, I had a conversation with a theorist named Bryan Swingle, who at the time was a post-doc at Stanford, and he has a background in, I guess, everything from condensed matter physics to quantum gravity, and so very complementary to mine. But was interested in whether one can do experiments that would probe a phenomenon known as information scrambling, which is sort of trying to get at what happens to information that falls into a black hole.

0:37:16.0 SC: So sorry, just to again, be clear, is the phenomena of information scrambling unique to black holes or is it a property that systems have including black holes and maybe… Maybe explain what it is.

0:37:29.3 MS: Right, yeah, yeah. So, to the best of my understanding, this sort of term was maybe first used in the context of thinking about black holes and asking about, you know, is information lost that falls into a black hole. And our understanding is, no, it’s not lost, but it is very quickly scrambled, which is to say it kind of gets hidden in complex quantum correlations and entanglement, and then it’s something… Information that was initially locally encoded in one quantum bit would become quickly very delocalized. And there is a conjecture known as the fast scrambling conjecture, that there’s sort of a fundamental limit to how fast this can happen, and that that would happen in essentially black holes, or perhaps we should say systems that are… And now you mentioned this concept, AdS/CFT, in systems that are dual to black holes under the framework of what’s known as holographic duality, so I almost feel like I should let you explain this. You’re more of an expert.

0:38:32.3 SC: Let me say a couple of words about it and then you can fill in for what is relevant to what you think about it, so… We did have Netta Englehardt on the podcast a few months ago, and she’s an expert. So the idea that they’re… Juan Maldacena back in the ’90s explained that there is a certain set of theories, quantum field theories, without gravity, so things that we think we understand pretty well, in n dimensions where n is some number like 4.

0:39:01.1 SC: And then there’s also theories with gravity, superstring theories in particular, but maybe it’s broader than that, that have a certain background geometry, anti-de Sitter space, so like a cosmology with a negative vacuum energy, and there’s a relationship between these non-gravitational theories in n dimensions and these gravitational theories in n+1 dimensions. And the relationship is supposed to be, they’re the same theory. And it’s not completely clear to me that it’s true that they’re the same theory, but they are certainly very, very similar in relevant operational ways. So the idea is that we can learn about a real theory of quantum gravity in the context of this dual theory without gravity, where presumably we understand things better. How’d I do?

0:39:46.4 MS: Right, yeah, and so to me, what was intriguing about this idea, or one, there are actually a number of things that are intriguing about it, but one was this idea that in certain cases, you could have on one side of the duality a strongly interacting quantum system whose properties might seem like they should be hard to calculate, and on the other side, there’s a way to actually visualize aspects of this highly entangled quantum system in terms of curved space and gravity, that it might give us some kind of way, new ways of being able to think about strongly interacting quantum systems, and have some ways of visualizing what generically you might worry is something that requires an exponentially large description. So that was one thing that to me was kind of intriguing about hearing about this.

0:40:36.0 SC: So the relevance to what you do is that you can imagine setting something up that resembles in some way the non-gravitational side of the duality, but then if dualities like this are real in some other way of looking at it, you’re doing an experiment that evolves gravity.

0:40:57.4 MS: Exactly, yeah.

0:41:00.1 SC: Or that mimics gravity, simulates gravity. I mean, you’re not actually putting something heavy in your lab and feeling its gravitational force.

0:41:05.4 MS: Right, and that’s also a fascinating direction of research, can one do precision measurements in a regime where quantum mechanics and gravity both matter. But that’s not what we’re doing, that’s right. So we are kind of asking, can we build quantum systems in the lab where there is this idea of some kind of emergent extra dimension, that has curvature that might be… That maybe we can think about as a gravitational system, is that something that might apply to systems we can actually build, and then… I think that connects to broader questions, there are big questions about is gravity in our universe, is that actually something that emerges fundamentally from quantum mechanics? I don’t feel equipped to answer that question, but we can explore the concept, and maybe it will help us actually think about have new ways of understanding quantum systems and entanglement.

0:41:57.6 SC: You know that when Penzias and Wilson discovered the cosmic microwave background, their famous paper just said we measured some excess antenna temperature, I’m not going to say what it is, right?

0:42:07.5 MS: Right, yeah, we’ll let other people worry about that.

0:42:10.9 SC: That’s okay. They won the Nobel Prize anyway, it still counts. So in other words, let’s see… The hope is that there are certain kinds of theories, there’s kinds of physical situations you can set up where they have the properties that if all this fancy theorizing is correct, there’s a dual description that seems gravitational.

0:42:32.8 MS: Right.

0:42:33.5 SC: And presumably these set-ups of entangled sets of qubits or atoms or whatever are not just lying around, you have to work hard to create them.

0:42:43.7 MS: Right, yes.

0:42:44.9 SC: And that is what you’re trying to do. Is that fair?

0:42:47.6 MS: Yeah, so I think we are… So let’s see, so we’re trying to create… As a starting point, one thing we’ve done is kind of build what I would say is kind of a toy model for this idea of an emergent geometry that describes something about the structure of the correlations in the quantum system. And for me, sort of that, starting from… And this is something where based on things we wrote on paper, we had an idea of what we should expect to see in the experiment, but nevertheless, seeing that feels like a starting point for thinking about how you can connect to these theoretical ideas and then the hope, I think, ultimately, is that one might build a system where one doesn’t know, sort of, that there is some picture in terms of an emergent geometry, but one discovers that perhaps by measurements then it tells you something about the system, right. So that might be the longer term goal.

0:43:45.4 SC: So, let me actually just dig into that a little bit. With what you’ve done so far, I guess the question is, are you learning things from the experiments right now, or are you just checking that you’re getting the answers you expected by doing the experiments so far?

0:44:04.2 MS: I would say that we are, first of all, kind of developing a level of control in the experiments to be able to… So just as an example, I mentioned first hearing about some of these ideas of holographic duality in the context of the phenomenon of fast scrambling. And one of the things that struck me there was that the models that people write down on paper that are supposed to behave in this way of exhibiting fast scrambling have rather exotic-looking interactions that are not local, right? And that was one of the things that first made me think, well, maybe actually, we do know how to realize, not precisely those models, but systems with non-local interactions, is that something where, we might be able to explore that phenomenon or build other toy models that would be sort of hard to realize in other systems where you have, you know, the nearest neighbor interactions on the lattice?

0:44:56.6 SC: Good, I mean, let’s dig even closer there, because I think that probably a lot of people think of entanglement as an interaction, but it’s not, right? I mean, you have something different in mind when you say… I mean, how do you make a long distance interaction in your world?

0:45:11.7 MS: Right, and so in our experiments, what we… So essentially, interactions actually are always local, but we can make things that look effectively like non-local interactions by letting photons carry information very quickly, at the speed of light, from one atom to another. And so effectively at the end of the day what we have is what looks like the atoms interacting, we actually can kind of ignore the photons at the end of the day and say it looks like these atoms interacted even though they’re a millimeter apart. That’s our mechanism for generating what I would call sort of effectively non-local interactions.

0:45:48.2 SC: And again, just for the people out there who listen to this and similar podcasts, this has nothing to do with the spooky action at a distance that Einstein worried about, right? Because that’s when you measure a spin and now you know if it’s entangled, you know the state of some other spin on Alpha Centauri or whatever, but you’re actually just creating entanglement or manipulating entanglement between spins, you’re not collapsing the wave function by doing some measurement on them.

0:46:13.9 MS: Exactly, so the first step is to generate these interactions which can create entanglement, and then ultimately to see whether you have entanglement, you do need to perform that measurement, and you would perform a set of measurements that do look like they’re showing what Einstein called spooky action at a distance, we understand how it comes about. And the one key thing is that information did need to sort of travel from point A to point B to generate that entanglement.

0:46:40.3 SC: And so maybe you said this, but it got lost, I mean, have you made quantum systems that exhibit this fast scrambling that they hypothesized for black holes?

0:46:50.6 MS: No, not yet. So, one of the things we have started to think about is sort of with the toolbox that we’re building, what are some models that might exhibit this fast scrambling? And we had, I had a toy model that I’d been thinking about for a while, where instead of having, let’s say, nearest neighbor interactions on a lattice, you have interactions at a distance of one site, two sites, four sites, eight sites, any power of 2. And that was actually somewhat different from, you know, models that are known to be holographically dual to black holes, but it has this feature that actually in, a sort of, information can spread exponentially fast, kind of from one point to any other point in the system.

0:47:32.4 MS: So the characteristic timescale from if I had a local interactions for information to spread from one site across the entire system, would scale linearly with the system size, here it can scale with the logarithm of the system size. So that, you know, I had kind of this toy model and was, had an idea of how we could actually do that in the lab, and then somewhere actually in the context of kind of thinking about this toy model, my, actually my graduate student at the time, Greg Benson, started kind of asking theorists whether there’s like a more rigorous way to think about this model than my kind of hand-waving. I can count… Yeah, my hand-waving sort of arguments.

0:48:13.0 MS: And through a rather serendipitous chain of events, Greg came into contact with Steve Gubser at Princeton, who rather remarkably, it turned out that this toy model we’ve been thinking about, if you tweak it a little bit, connects to a version of this AdS/CFT correspondence of holographic duality that Steve formulated, that is, actually really kind of builds on some very deep ideas in number theory, and I can go deeper into that if it’s of interest, but that actually then led us to this idea of, by tweaking that toy model, we actually could build something in the lab that has some sense of an emergent geometry that looks a little bit like anti-de Sitter space.

0:49:06.2 SC: Okay, so let’s just catch our breath here. So, if we didn’t know any better, if we were just dumb and naive had a bunch of atoms on a lattice, and we poked at one of those atoms with a bit of information or something like that, we’re being very hand-wavy here, what you’d expect is the information would then spread out to the nearest neighbors of that atom and then their nearest neighbors, so it would spread out linearly over time. But what you want is, or what the conjecture is, happens in gravity and black holes is faster than that, so you poke one atom and it spreads out all over the place, and you’re able to build toy models of that.

0:49:45.2 MS: That’s right, yeah. And so we haven’t really directly probed the scrambling, I would say, what we’ve done so far is show that we can build this graph of interactions.

0:49:54.0 SC: Right, okay. And then presumably where Gubser etcetera come in, is that it’s one thing to make these non-local interactions, it’s another thing to make them of precisely the right form to look like gravity in some dual theory, right?

0:50:10.4 MS: Yeah, that’s certainly true. That’s another thing to make them of the precise form to look like gravity in some dual theory, and I mean, so far, kind of the toy model we’ve realized in the lab, has, it has some parameter we can tune and there’s a particular point where it does have this feature of, to the best of our understanding, actually, you can’t quite really do the theory for a scalable number of particles, but to the best of our understanding it could exhibit fast scrambling at some point, but I’m not claiming that it’s the holographic dual of a black hole.

0:50:44.5 MS: And then there’s a different place you can tune it to where it has some features that bear a resemblance to a system where there is kind of an emergent geometry, that looks like a curved space, but it’s not actually a fast scrambler in that regime, so.

0:51:00.6 SC: Okay, well, that’s fine. So what is the… Where do we make our money doing this? Are we saying that… I mean, at the end of the day, you have a bunch of atoms, and they don’t really have gravity in some sense, are we trying to learn things about systems that have some dual gravitational description or is there a question that we don’t know about those systems that you’re going to answer experimentally?

0:51:27.2 MS: I think, what I would like to understand better… And again, to be honest, I would say lately we’ve been focused on building up a toolbox, and for a while, I actually sort of stopped chatting with theorists because I thought…

0:51:42.8 SC: Too distracting, yes.

0:51:43.6 MS: You can always have billions of ideas, but if you can’t do the experiment then where does that get you. No, but now… I think to me, one question is, let’s say we have this particular toy model where it does seem that there is a sense of what we might call a holographic bulk geometry, and is there some predictive power to that. And I don’t have a sharp answer to this, but is there a sense where having this geometrical picture helps us predict something about the dynamics of the quantum system.

0:52:13.3 SC: I see. Okay.

0:52:14.1 MS: Helps us understand what’s the most efficient way to transfer information from point A to point B in this collection of qubits, for example. And I think that there are things there, where actually the gravity… It’s not precisely gravity, but this picture of this bulk geometry actually should give a useful way of thinking about the behavior of the quantum system. And I feel that having a toy model in the lab to play with that, and hopefully soon more than one model, is kind of a starting point, so I guess I feel that there’s always… There’s value in doing experiments, it sort of… [chuckle]

0:52:47.9 SC: Well, yeah.

0:52:48.8 MS: For me at least, clarifies things to really think, how do you do this in the lab. Just to give maybe also one example, so what I think is fascinating theoretically, is the idea that there is a direct connection between… We talked about entanglement earlier, between the property called entanglement entropy, which is a measure of entanglement in the quantum system, and the geometry in this extra dimension. So there’s this idea that the amount of entanglement entropy in some region on the quantum system that lives on the boundary of this higher dimensional space is connected to the area of a particular surface in the higher dimensional space in the bulk.

0:53:33.9 MS: So that is something where… That would be kind of cool to see in a system where you can directly measure that in the lab, and there are some experiments where one can measure entanglement entropy. But also you can ask, do you need to measure that? Or at least as an experimentalist, you naturally say, well, are there simpler experimental observables that will kind of get at some of the same physics? And in our experiment so far, entanglement entropy isn’t something we can measure, and we were actually kind of surprised… Even though when we’ve been doing some theory before, we had made some plots of what the entanglement entropy would do in a system like this, and we were sort of surprised that actually by being forced to work with what we could, there were simpler observables that kind of actually showed some of the same physics.

0:54:18.2 MS: And so that type of thing, you don’t really have to think about till you’re confronted with it in the lab, but perhaps you can learn something from that.

0:54:24.2 SC: Yeah, no, I’m entirely on board that experiments are just simply worth doing, right? You can’t just have theorists talking to each other all the time. But just so, I think I finally get some logic from your explanation that I hadn’t gotten before, so let me play it back to you, and you can tell me if I’m on the right track. So you have this system of atoms with entanglement, and you’re poking at them with photons and whatever. And in principle, I could imagine taking a giant computer and solving for what the system is going to do, using the Schrödinger equation, but in practice, that’s just impractical, there’s too many degrees of freedom, too many things going around.

0:54:58.2 SC: But if the specific arrangements of entanglement and interactions have this holographic dual, that is to say there’s a way of thinking about them that looks like gravity in an extra dimension, it’s not gravity in an extra dimension, we know how many dimensions there are, etcetera, but you can use that knowledge to make a prediction for what’s actually going to happen in your experiment, and then you can test to see whether that comes true.

0:55:22.1 MS: Yes, yeah, yeah, and I think that’s sort of… The sort of fundamental challenge that we never really explicitly stated in our discussion about entanglement earlier, is the fact that having a full description of a quantum system that I could, let’s say, put on my classical computer, that description grows exponentially with the number of particles, and even for 50-some particles that just won’t fit on the world’s largest super computer. Right.

0:55:47.8 SC: That’s right.

0:55:49.9 MS: And so that’s where the question is, under what circumstances is there a simpler description, and in some cases… And there’s lots of work on that by people who are pushing the state of the art in numerical simulations of quantum systems. And in some cases, where it’s a one-dimensional system, and you have nearest neighbor interactions, it’s kind of known how to more efficiently represent the classes of quantum states that you would naturally get from that in certain cases. But in more generic cases, it’s not obvious necessarily. And the hope is that this holographic duality could perhaps give an approach that’ll lend itself to some quantum systems where currently we don’t know how to efficiently describe them.

0:56:36.7 SC: And is it too much to ask that we might learn something about quantum gravity by doing this? Are you going to address questions about black hole information loss?

0:56:47.3 MS: So I don’t think that I personally can solve those problems.

[laughter]

0:56:52.2 MS: Because a lot of other people have thought about them much more deeply than I have, so what can we contribute? I think we can talk to the people who’ve thought more deeply, right. And we can, again, build these model systems and see how they behave in regimes where you might not be able to calculate it, right? And that’s… And I would say actually right now, we’re still working in the regime where we can calculate what happens well, in our system, that has something to do with the fact that we work with little clouds of atoms that you can think of a bit more semi-classically.

0:57:26.5 MS: And so there’s still… I mentioned a lot of this is about just developing the experimental tools, so there’s a path to getting to systems that are more in a regime where it’s, let’s say, one qubit per site, and it has a simple theoretical description on paper, but nevertheless, it’s hard to calculate what will happen. And so, in that regime, the hope is that with the right sort of dialogue between us and other experimentalists who can build systems in the lab and theorists who might know exactly what is the right question to ask to learn something about information and black holes, I think that that dialogue is essential.

0:58:05.0 MS: And part of it… So for example, but you mentioned before, if one can build a system that’s a fast scrambler, that’s not a sufficient condition for having the holographic dual of a black hole, but it is a necessary condition, and then you can start, at least to the best of my understanding, and so then you can start to ask, you can explore, can we build things that maybe it’s not actually easy to calculate what will happen, but you can measure it in the lab, and people are thinking about more and more refined tests for whether the system is or might not have a gravity dual beyond just asking how fast is information scrambled. And so I think there’s a lot of parallel work in terms of making experiments more capable and thinking about what are the right measurements to perform.

0:58:54.8 SC: I don’t know if this is a fair question, but in the original work by Maldacena, the non-gravitational side of the duality was this very specific theory, right, super Yang-Mills theory with a lot of supersymmetry, etcetera, etcetera. So should we be… How confident are we that just from a bunch of rubidium atoms, you can mimic a system like that, are we confident that we have the right properties?

0:59:22.6 MS: No, certainly not, but to me, one of the questions is like how generic are these ideas about holographic duality. And I believe that theorists are often restricted to thinking about models that they have the right tools to analyze. It would be great if the concepts generalize beyond those models.

0:59:49.2 SC: Yes. Okay, good. That’s a very good answer.

0:59:51.2 MS: And so what we can do on the experimental side is build systems that we might have some reason to believe are interested in, like, oh, I think this seems like it should be a fast scrambler, maybe it’s interesting. And then learn something in regimes where you can’t easily analyze it on paper.

1:00:08.9 SC: Yeah, I’m a big believer in doing the experiments and being surprised and then realize, oh, I should have thought of that all along, right?

1:00:14.2 MS: Yeah.

1:00:15.4 SC: And I guess it’s worth… Before we end, I do want to come back to the fact that there are other reasons to do these experiments, other than pretending to make black holes or whatever, so you know what those reasons are better than I do, but one that you mentioned is precision timing measurement. So how does that work?

1:00:36.1 MS: Yeah, so generally speaking, one of the reasons that we love atoms is that atoms are our best measurements of time, and the second is defined in terms of an oscillation frequency in the cesium atom. And one place… And actually, the best clocks made of these laser-cooled atoms really are at the point where one of the limitations on their performance is quantum uncertainty, that sort of coin-toss in a ways that comes from the randomness of these acts of projecting an atom into one of two states.

1:01:12.5 MS: So this is one place where entanglement can help you introduce correlations that if one coin lands heads, another is more likely to land tails, so to speak, and so the fluctuations in that total count are reduced…

1:01:26.1 SC: I see.

1:01:27.1 MS: And so that’s a direction where entanglement certainly can help, and there’s lots of work going on in terms of just taking some things that have already been kind of demonstrated and applying them towards the world’s best clocks. One of the questions I’m fundamentally interested in is the sort of simplest states one could make that have been made are so-called squeezed states that have that sort of collective entanglement, every atom is entangled with every other in sort of an equal way, and those are certainly… It’s well understood why those are useful for those applications.

1:02:04.9 MS: But one of the questions I’m kind of interested in is if you can have richer structures of entanglement, does that also have a benefit for sensing and in what cases does it have a benefit, and that might be that you’re trying to get more information than just measuring one single quantity, it might be that you’re trying to actually image a magnetic field or something, you’re trying to… Or get information, how is some signal varying as a function of time, what are the spatial correlations in a signal, so kind of… There’s not that much… I would say we’re at an early stage of understanding kind of how richer structures of entanglement can offer benefits in both sensing and time-keeping tasks.

1:02:48.2 MS: And so that’s something where the more we kind of build up this toolbox, the more we can start to even explore at a fundamental level how to harness entanglement, how to fully harness it as a resource for precision measurement.

1:03:01.6 SC: I’m sure there’s very good answers to this question, but why do we want to have even more precise measurements of time than we already have? We’re pretty good at it now.

1:03:09.9 MS: Yeah, well, one of the… We talked earlier about sort of different ways of probing gravity actually are harnessing holographic duality, but perhaps actually, can you really do measurements in a regime where quantum mechanics and gravity are both playing a role. And the best atomic clocks are just extraordinarily precise, to the point where if you have just a very small change in the height of the clock, I think by now probably in the millimeter scale, one can actually… That changes the rate at which the clock ticks because of gravitational redshift, and so you can actually really sort of resolve that and see these effects of general relativity in the atomic clock, right.

1:04:03.5 SC: And in this case, you’re talking about real honest to goodness gravity, making apples fall from trees, not dual gravity in your theory.

1:04:10.1 MS: Exactly. Yeah, yeah. And so I would actually say one of the motivations for better clocks is that actually you can do better and better precision tests of fundamental physics. There are ideas for using clocks to detect gravitational waves, and regimes where sort of different parameter regimes from LIGO and things like that, so that’s kind of one direction.

1:04:30.2 SC: And it seems… Maybe this is too naive on my part, but it seems, from all the words you use, that maybe quantum computing is an application for this kind of thing, or do you learn anything that has been relevant to quantum computing?

1:04:41.0 MS: Yeah, so one of the things that we are interested in… So, certainly generally entanglement is kind of the fundamental resource for quantum computing, so in a very general sense, anything that advances what types of entangled states you can prepare might have some computing applications, but at a more direct level, there’s sort of the goal of building a universal error-corrected quantum computer, and then there’s a kind of goal of asking, are there certain classes of computational problems that might naturally map to existing or near-term hardware in the lab? Can we ask whether the quantum systems that are natural for us to build can solve certain classes of problems?

1:05:26.3 MS: And so in that vein, there’s a whole class of optimization problems that can be mapped to essentially minimizing the energy of an interacting spin system, and generically those problems… So these are things in the vein of travelling salesman problem, they can be certain scheduling problems and things like that. Generically, those require some spins with non-local interactions, and it turns out that… So if you can… And have a particular problem that you’re solving, you need to be able to basically program the graph of those interactions, and you’d like to also control the sine of the interaction, do the spins want to align or anti-align, and how strongly do they want to do that?

1:06:13.2 SC: So sorry, when you say the graph of the interactions, we have some atoms that we know their locations, and then the graph of the interactions is given atom A and some other atom B, how strong is this interaction?

1:06:26.0 MS: Exactly, yeah. And also, what is the sine of it, do they want to align or do they want to anti-align…

1:06:29.2 SC: For every single pair of atoms or whatever.

1:06:32.5 MS: Yeah, yeah. And so having these programmable non-local interactions is a great way to start exploring, does this quantum system give you a way of efficiently solving these certain classes of problems that are known to be hard classically, and it’s not… There are lots of… There’s lots of theoretical work on, can quantum systems help with this or not. And there are certain cases where it’s known that they won’t help, or certain approaches won’t help, and certain cases where it looks like the quantum systems can help, and then a huge wide open space where I really think you need to kind of play with the systems and learn from experiments.

1:07:20.1 MS: And so we naturally now have a way of making these non-local interaction graphs, we see some evidence that we have a way of naturally generating what looked like low-temperature states of this interacting spin model, and we would love to go deeper into exploring for the cases, the types of graphs that are hard classically, can our system find the ground state and so forth?

1:07:41.5 SC: And it seems to me as a quasi-outsider that the field is advancing pretty darn rapidly. How do you see what this field is like 20 years from now in terms of how many spins are being entangled and what kind of systems you’re looking at?

1:07:57.9 MS: Yeah, 20 years. I think it is really hard to predict because like you said, even just in the past few years, there have been quite rapid advances. I would say one area where there’s been, in this field of cold atoms, substantial progress in the past few years is the ability to really scalably create arrays of individually tracked atoms, where one can have an ordered array of hundreds of atoms that you can look at one by one, and pretty soon, hopefully manipulate any one of them individually.

1:08:33.5 MS: There are ways of having nearest neighbor interactions that are explored in a number of labs, and it’s actually also something we work on in a different set up than the one I was describing, having… So you can either have local interactions. Or one thing that so far from that toolbox is kind of missing is having this longer range connectivity that could be really valuable in this context of efficiently implementing certain quantum algorithms or generating certain structures of entanglement. And so to me, this is maybe not an answer for the field as the whole, but one… As a whole, but one thing I’m kind of excited about this, can we merge these different technologies of the ability to have more control over the graph of interactions with this really scalable single particle control.

1:09:22.1 SC: And this is the kind of thing where you could do it much, much better if you had a billion dollars, or is it just you need time in the individual 500 square foot labs, and eventually we’ll get to the point where you need a lot more money.

1:09:33.0 MS: I would say… I would say for things I’m working on, I would say we need time in individual 500 square foot labs and the ingenuity of grad students and post-docs coming up with clever solutions, and that’s partly… I think my taste is doing some things that are a little bit… You don’t already know how to do it, if I knew how to do it, and that just money would solve it… [chuckle]

1:09:57.9 SC: You’d be something else.

1:10:00.1 MS: I’d actually be less excited, [chuckle] than I am about things where I have a sense that there is something interesting to explore, but I don’t really quite know what the answer is yet, yeah.

1:10:08.4 SC: I think that’s the perfect place to end. And you’ve convinced us all that we need to understand entanglement better, theoretically and experimentally. So Monika Schleier-Smith, thanks very much for being on the Mindscape Podcast.

1:10:18.8 MS: Thanks so much, again, for the invitation.

[music][/accordion-item][/accordion]