Generative Quantum Eigensolver with Alán Aspuru-Guzik
Download MP3The New Quantum Era, a podcast by Sebastian Hassinger.
Kevin Rowney:And Kevin Rowney
Sebastian Hassinger:I'm gonna begin this episode with a quick apology for leaving you all without an episode for over a month now. As you know, we try to get an episode out every couple of weeks. End of year is always challenging for scheduling, because a lot of our guests are in the academic world, exams, finishing up grading things. Those types of events tend to get in the way, and then I tend to get behind on scheduling. But, we're picking back up now.
Sebastian Hassinger:We've got a bunch of incredibly fun and interesting things planned for the year, which we'll be sharing more about as the year progresses. And today's episode is one that I've been working on for quite a while now. I've been wanting to interview our guest today for quite some time. He's a pioneer in the field in the true sense of the word. He's been working for quite some time in quantum computing.
Sebastian Hassinger:His name is Aspuru-Guzik. He is, was a an author of the foundational paper, the Variational Quantum Eigensolver Technique. If you haven't heard of it, the, Variational Quantum Eigensolver or VQE is a a technique for combining classical computing with quantum computing to try to come up with approximations, particularly of the ground state of chemical systems. It's really interesting because I I thought it was very innovative. The paper came out in 2013.
Sebastian Hassinger:It was the first, along with, QAOA, the first two hybrid approaches to try to get some practical results, something useful out of the NISQ devices that have been sort of the typical, output of the hardware vendors for the last 10 years, noisy intermediate scale quantum devices. Variational techniques are actually, they're mathematical techniques that are quite familiar both in mathematics themselves that they were initially for solving partial differential equations and boundary value problems. They've been adopted in machine learning, and they're used in physics before quantum computing as well just to, to in quantum mechanics to to get those approximate ground state energies and wave functions. So, you know, it was very much an off the shelf kind of technique, but adapted very cleverly for quantum computing. The reason that it, you know, that I find it clever is that it takes advantage of the inherent strengths of qubits and, of course, the the linear algebra that that, is native to quantum computers, but manages to or or attempts to sidestep the limitations of the devices we have today by only running very short circuits, on the device before returning the result back to the classical code that then, can take another pass at another, iteration of the the approximation to get closer to your your eventual end answer.
Sebastian Hassinger:The as I mentioned, the the mathematical approach is also familiar to machine learning, and therefore, this type of, of hybrid, variational approach is also core to a lot of what is considered quantum machine learning, early research into quantum machine learning. And so the topic today is work that Alain has been carrying out over the last couple years to update the variational quantum eigensolver technique with, sort of more cutting edge machine learning techniques. Really exciting stuff, and I think you'll enjoy the interview quite a bit. Welcome back to the podcast. I'm here today with yet another very special guest.
Sebastian Hassinger:Super excited to be speaking with him, Alán Aspuru-Guzik from the University of Toronto. He's a professor of chemistry and computer science and really a leading pioneer in the field of quantum computing from way back when. I actually first saw you speak, Alán, at, IBM's Think Summit in 2017, when they were announcing the 53 or 54 qubit system was coming out the next year. And you presented on stage your variational quantum eigensolver, eigenvalue solver, which was a really interesting, I mean, since it was my first exposure to quantum, it completely, steamrollared my brain. I don't think it even blew my brain.
Sebastian Hassinger:It just completely flattened me. But it was really exciting because it it was something potentially practical we could do with near term noisy quantum devices. And that was 2013, I think, that paper came out. So 4 years earlier and more recently, you sort of revisited the variational space, with an updated kind of approach. So first of all, thank you for joining us.
Sebastian Hassinger:And, can you can you give a little bit of background on the work that led to that that first VQE paper?
Alán Aspuru-Guzik:Absolutely. So I will start even earlier. So the VQE work is around 2013, 2014. But let me start in 2004. So many of the listeners on this on this podcast, we probably were not born then.
Alán Aspuru-Guzik:But, just to give you, like, a 2 1 minute and a half introduction, when I was finishing my PhD, I joined the group of Martin Head-Gordon as a postdoc. There, D Wave Systems, at the time, was thinking about gate model quantum computers. And cofounder, Geordie Rose, handed me over 5 papers that he found were interesting and funded my advisor to think about quantum computing for chemistry. Me as a postdoc said, oh, this is really cool. And collaborating with Peter Love and Tony Dutoi and, of course, Martin, we looked at those papers, and I found out kind of an easy mapping to do chemistry on quantum computers.
Alán Aspuru-Guzik:And, basically, that paper simulated quantum computation of molecular energies in 2005. Got got me you know, published in science, got me like a like a door at Harvard where I started my research group. But the VQE comes from that pressure because at Harvard, they said I said, look. I'm gonna do all the things that are not quantum computing because quantum computing for chemistry is not a new field. And I said, what are you talking about?
Alán Aspuru-Guzik:We hired you to create a new field. Go and just create a field. Create the field of quantum computing for chemistry.
Sebastian Hassinger:No pressure.
Alán Aspuru-Guzik:Not not that always we're not working at that, but I said, okay. I mean, I was pretty much one of the first, so we found the first. So I said Yeah. Bring it on. And I needed to show something by inter my intermediate promotion.
Alán Aspuru-Guzik:It was around 2010. I need to show something. So, actually so I said, you know what? I need to get some results. So I met this fellow, Andrew White in Australia, fantastic quantum optician.
Alán Aspuru-Guzik:And we met and we said, like, why don't we try to build a minimalistic, phase estimation algorithm for molecules? And we published a paper in Nature Chemistry in 2010. That is the first time, actually, due to a quantum chemistry calculation on a quantum computer with quantum optics. There was a group in China that got it, like, a couple weeks later after I was in the archive or, you know, like, around the same time. I forget if it's before or after, but I'm traumatized by that scoop because we put it together around the same time.
Alán Aspuru-Guzik:And one in NMR, one in one in quantum optics, and then we just suddenly, did that. And you know, it was really cool because we even had error correction. With 2 qubits, you can actually do majority voting error correction. So we got the potential in your surface of the hydrogen molecule. And that's a very interesting paper.
Alán Aspuru-Guzik:It's, I think, one of the first applied papers in quantum computing ever. Right. And what we did is that paper that was really cool, obviously, besides doing showing error correction and showing that a molecule can work. It's showing the most explicit diagonalization of a 2 by 2 matrix ever done in history because, really, that's what he was doing. Right?
Alán Aspuru-Guzik:I mean, like, a quantum optics table about these size just diabolizing a matrix. Okay. So it is in that context now I need to get tenure that VQA is born. Okay? So don't underestimate the precious academia.
Sebastian Hassinger:I was gonna say. So Everybody complains about it, but it sounds like it was a great motivator for your productivity.
Alán Aspuru-Guzik:I mean, my group was blessed that, I had, like, this, like it's almost like you know, my group has Toronto also. I have, like, the rosters of the rosters of the field come as sponsors. They wanna work with me, some bricks. Happy to have some of the best people. And and then in this particular case, you know, my team and I were focused on a lot of NMR and actually, we did we did, even NV Center experiments and, you know, we collaborate with a bunch of people doing small experiments of quantum simulation.
Alán Aspuru-Guzik:And there's a moment where, you know, like, we were thinking about why is phase estimation really measuring energy in such a complicated way? Why don't we just measure the energy directly? So, I went to to Jarrod McLean and to Man-Hong Yung, and I told him, like, why don't we measure the energy directly for Hamiltonian? And then the next day, Jarrod came to my office with Man-Hong and said, we figured out how to do it. This is how we do it.
Alán Aspuru-Guzik:You just write the Hamiltonian, map it into spins, and, you know, you measure it. And and I said, let's write a paper. But you know what? At the time, it was very hard to get a very high profile paper if it's only theory. Right.
Alán Aspuru-Guzik:They said, let's work with, let's work with with Jeremy O'Brien, to do a because he was a great guy. I met, you know, very energetic professor of my same age from Bristol at the time. And he's a guy now. He's the CEO of PsiQuantum.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:They're in a black hole somewhere in the Bay Area building a quantum computer. Yeah. I haven't talked to him since then. Maybe one couple Zooms because they are so secretive. Although, they have hired some of my former group members there.
Sebastian Hassinger:I'm sure they have.
Alán Aspuru-Guzik:But, anyway, the whole story is that, no, there's PsiQuantum doing its thing. But back then, the beginning was PsiQuantum also, right, was this thing. You know, like, it was the beginning of my former company, Zapata Computing, the beginning of psych quantum. So many things was the idea of, okay. Let's just measure directly and then variationally optimize.
Alán Aspuru-Guzik:Let's actually do what people do in quantum chemistry. Right. And and that's how VQE was born. It was together with QAOA, which was presented at the same conference in Aspen. Eddie Farhi stands up and persist QAOA.
Alán Aspuru-Guzik:Then I stand up and present VQE. We were like, dude, this the same thing almost. That was the beginning of the that was the beginning of the the the variational algorithms finally. Well, and I
Sebastian Hassinger:was gonna mention QAOA because at least my impression back in 2017 was that VQE was, you know, sort of the the most popular kind of path to take for trying to make progress in in, you know, chemistry applications in quantum computing. And QAOA sort of led directly into the broader topic of quantum machine learning. Is that would you say that's sort of accurate?
Alán Aspuru-Guzik:Yes. Quantum optimization general, of course, quantum machine learning. At the same time, my former student also, Patrick Rebentrost, was already working as a puzzle with Seth Lloyd. And together, they were rocking in a bunch of ideas about quantum machine learning. So I think quantum machine learning, many other people, like, worked on it besides Seth Lloyd.
Alán Aspuru-Guzik:But, in in my around my closer circle, I think the person that I recall around the same time, really thinking heavily about machine learning was Seth and Patrick. And they were they were rocking at that as well. So Peter Wittek, the the late Peter Wittek, also wrote a textbook on quantum machine learning.
Sebastian Hassinger:That's right.
Alán Aspuru-Guzik:So there were other there were other types of algorithms. But, obviously, later, people did, variation of quantum classifier. We were very, very early on did the first autoencoder, the variation of quantum autoencoder. So the quantum autoencoder was a really cool thing because I was doing the first autoencoder from
Alán Aspuru-Guzik:molecular design. It was my most cited papers.
Alán Aspuru-Guzik:And we're like, oh my god. Maybe we can do a quantum out of color. And then with, Johnny Olson and many other people in my group, Jonathan Romero, we published the quantum out of color, which is also one of the seminal papers in quantum machine learning. Right. That's our contribute our contributions, we also worked a lot in the quantum.
Alán Aspuru-Guzik:So that's that's kind of the era. Right? So Yeah. Yeah. That's that's kind of what I can tell you about the VQE.
Alán Aspuru-Guzik:I guess you wanna talk more about the GQE, but now people saw history too.
Sebastian Hassinger:Yes.
Alán Aspuru-Guzik:Yeah. People saw history between 2004. Now we jumped into 2012 13, 14, the thinking then around then. I guess the the quantum outer color paper came out around 2016, if I remember. So those were the kinds of things I was thinking around that time, and then, I guess, we can talk about what's what we're thinking now.
Sebastian Hassinger:Well, that's, I mean, that's where I wanted to to get to after this sort of you know, like you said, the historical context is super important. But I think, January of 2024, you published a new paper, which was, the generative quantum eigensolver and its applications, I think, is is the title. Yes. I don't I think it's been somewhat flying under the radar Yeah. To my my way of thinking because I think it's a really powerful sort of revisiting of those, those foundational ideas behind the the VQE paper, but updating it for, you know, based on everything we've learned in the last, 10 years or so and also for the, the increasing power and, powerful techniques that we've been developing around machine learning and accelerated computing with GPUs.
Sebastian Hassinger:So, and obviously, generative AI has become this incredibly huge topic, over the last couple years. Speak to me, like, how did you apply, sort of generative techniques to the to the variational foundation of of of VQE? Or is that is that an accurate depiction for one thing?
Alán Aspuru-Guzik:Yes. It's it's it's a reasonable depiction, but it didn't come that way.
Sebastian Hassinger:Okay.
Alán Aspuru-Guzik:So the story is, a little bit after the VQA paper came out, of course, we wrote a paper called the theory of the variation of quantum magnet solver. Still, with Jarrod McLean and many others in my group, that is also very heavily cited. It's kind of cited together with the variation magnet solver, then a paper where we have the theory and a bunch of ideas that we have of extensions. Very late a little bit later, there was a very important paper published by Gerald as well before he joined Google, where Jarrod independently, with his collaborators, found this idea of a barren plateaus. The idea that, in the in the in the long end limit, like, the the expectation value of the gradient of, of a variational algorithm will actually be 0, will be flat.
Alán Aspuru-Guzik:Right? This that just has to do with just the statistics of how you do quantum measurements. And that means that without, like, previous knowledge of the of the problem initialization of what your tricks, you might get lost in the optimization. And, that seemed like the tool most of ourational algorithms. Yeah.
Alán Aspuru-Guzik:And is that
Sebastian Hassinger:is that related in some way to sort of the local minima problem with annealing? Is it similar in the sense that you're you're sort of without, preknowledge of the context of the problem? You may end up thinking that you've arrived at a solution, but it's just a local locally minimal, energy level.
Alán Aspuru-Guzik:I think it's there might be some connection between them. I we worked in my group on both, but I think I see them differently in different kind of there might be manifestations of the of the rule of no free launch in some sense. But Yeah. What what I mean is that yeah. So, like, you can't rely on the gradient.
Alán Aspuru-Guzik:You can always do gradient less optimization, but if you have too many, many parameters, right, maybe even there, you you have issues with the noise of your measurements. And so then a mantra in my group became I mean and and people in my group know how it works. Like, in my group, basically, I kinda like to set challenges and then keep barking at my students with a challenge and then, see who drives us up to the challenge. And a lot of interesting algorithms have come out of that. So I told them once, like, hey, guys.
Alán Aspuru-Guzik:We don't want radians. So there was an algorithm that I proposed to my group. We call it quantum Angry Birds. I'm not gonna tell you what it is, but if you remember the video game Angry Birds. Yeah.
Alán Aspuru-Guzik:Right? It had to do with initial conditions and quantum dynamics, and we we call it quantum Angry Birds. Can you aim at something by, like, initial conditions? And we start to thinking about algorithms. We never publish quantum Angry Birds, but we started thinking about this type of algorithm.
Sebastian Hassinger:That seems like a missed opportunity. That's very good marketing, quantum angry birds.
Alán Aspuru-Guzik:Well, we just couldn't make something work with quantum angry birds.
Sebastian Hassinger:For sure.
Alán Aspuru-Guzik:Well, I'm guessing the idea. If somebody somebody has an idea about how to make quantum angry birds work, just write to me, and we can polish the paper together. But those are the kinds of things. Like, I sometimes come down in, like, group meeting. I'm like, okay.
Alán Aspuru-Guzik:This is what I'm thinking. I have to say, be clear. I mean, the the stuff that happens in my group, many times, it's just the puzzles and the students, they come up with the idea. I help a little bit. You know?
Alán Aspuru-Guzik:It's like, we're an anarchist collective, as Peter Love likes to call, so it's call us. Right? So so then I I kept telling them, like, I want an algorithm without gradients. Like, we need to think about a way of, like, having a variational algorithm where you don't need the gradients in the quantum computer. I want to calculate the gradients outside of the quantum computer.
Alán Aspuru-Guzik:And then this brilliant scientist, which now works at NVIDIA, spent a couple years in my group, Kohei Nakaji, has all the credit of taking that idea and coming back to me one day and says and he he's very humble and very smart. He just comes and, you know, tells me a lot. Like, I found a way of doing this. And he says, like, I'm gonna use a generative model to generate the variation of trials. Mhmm.
Alán Aspuru-Guzik:And then I'm gonna optimize the gradients in that generative model. Right? Like, so the gradient is pushed out to the quantum computer into this thing. Mhmm. This thing prepares the quantum states in the quantum computer.
Alán Aspuru-Guzik:And that thing, I would tell you what I call it, that thing. It can be any machine learning model. We use a transformer, but it can be anything that generates generatively. Mhmm. You like GaN?
Alán Aspuru-Guzik:GaN. It can be an autoencoder. You can do an autoencoder.
Sebastian Hassinger:So Interesting.
Alán Aspuru-Guzik:It's it's cool because it could be even a quantum autoencoder. It could be kind of cool to have, like, quantum those quantum. So there's a bunch of things you could do, but at the moment, like, we use I have transformers were amazing in GP 2D. Yeah. Yeah.
Alán Aspuru-Guzik:There's there's a few reasons we use transformers, but that's why the paper is called generative quantum magnetism. Or it doesn't matter what you do.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:Later in the paper, the application is called GPT QE because we use a GPT, and I'll get into that in the remainder of the podcast. But, at the moment, just think about it as a generative model generates a circuit.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:So the first problem is how do you optimize that? So we have to have the loss function for the for the for the, for the model, and that's what, that's what, Kohei found out, how to to have some sort of Boltzmannian, e to the minus of beta energy kind of loss. A loss that as you keep optimizing the parameters in a circuit, the ensemble of the ensemble of the states that you prepare keeps being lower than lower than energy. So you generate a bunch of states, and then the quantum computer only measures the energy of the states. Now it doesn't need to measure the gradient.
Alán Aspuru-Guzik:Mhmm. But then the quantum circuit, based on that ensemble of energies, feeds the fits its parameters to generate tokens, which in turn would generate lower energies.
Sebastian Hassinger:So so the the model itself is trained on on prior or on test shots or on prior work from that device? Or how do how do you train the model on, so that it will, generate the circuits?
Alán Aspuru-Guzik:Let's get on to that because that's gonna be a big part of what my group is doing. But at the moment, at the moment, what we do is, what we do is, basically, just learn on previous evaluations. The the first thing is, like, if you're evaluating the water molecule at a particular geometry, whatever system you're looking at, you could start with random parameters and slowly lower the energy. It goes converges really quickly, almost like like BQE converges. It starts learning what parameters lower the energy.
Alán Aspuru-Guzik:So what
Sebastian Hassinger:we Okay.
Alán Aspuru-Guzik:It can it start can start from a cold start. Okay? And what the paper has and the reason this paper is is still appropriate is because we got some reviews that were focused on high performance, large scale. People really wanted to see very large scale results. Right.
Alán Aspuru-Guzik:And and we told the reviewers, like, you know, come on, guys. Like, this is a new algorithm. Like, we don't care too much about running it into a supercomputer. I'm just showing you that it can run larger. So we started, like, 3 or more VQE papers that are so VQE papers that are being written.
Alán Aspuru-Guzik:And then, well, Kohei and I just corresponded. We'll revise the prepping very soon and resume to another journal. So this paper is it's been there in the archive because we started so many collaborations. I think, okay, Kohei has about 10 or so going on. Right.
Alán Aspuru-Guzik:On the on the GQE, but the paper is kind of like there. We're gonna revise it with all our new knowledge. Right. So but, anyway, the point is that the the point is that, that that preprint, right, shows that we can do that. We can generate with the you can even start random, for example, and then slowly, the algorithm learns, which is important to know.
Alán Aspuru-Guzik:Now this uses GPT, pretrained. That's when things get really interesting. Okay? Because we could generate large quantum mechanical models. Right?
Alán Aspuru-Guzik:Models that could be could be pre trained. And what we know how to do now nowadays, at least what we are publishing so far, You could remind our collaborations to stand this in a lot of ways. But what we're doing is okay. So if I have the same number of electrons at this point and I have, for example, the same molecule, which is different geometries. If we have data on a few geometries, we can always learn predict the other geometries way faster.
Alán Aspuru-Guzik:So it learns the parametrization of the Hamiltonian, and we have done a lot of work on that. Even our first paper in the outer corner showed that if you learn across a potential energy surface, you can't predict the rest. Our quantum autoencoder paper originally did that. We did it also, with a paper with Alvar Severa Alerta that combines quantum machine learning with out encoders with variation sorry. Variational ligandsolver.
Alán Aspuru-Guzik:That shows also that we can do that. Like, it's called the meta variation ligandsolver. That's an interesting paper we published along the way called metaVQE. So so learning across a potential in the surface has been done before. We did it in metaVQE.
Alán Aspuru-Guzik:We did it in the autoencoder. Now we're doing it here again. It's kind of like, we know it's gonna be done, but what is nice about it is that then you can imagine, okay, this is VPT. GPT can read the entire Internet. So in principle, can I generate a big data set of molecules?
Alán Aspuru-Guzik:And then the question is how to learn across molecules, and that's where the work is happening.
Sebastian Hassinger:Is there any, I mean, is there any, sort of relation to how, Google has approached AlphaFold, for example. I mean, they're training that model that, the DeepMind model on on known molecular geometries as a way to sort of, short you know, shortcut to guesses about other protein folding geometries. Right? Is there is there any relation to that type of work or or similarity?
Alán Aspuru-Guzik:Yeah. Yeah. It's related to that. We wanna use a bunch of known examples, and then what GQE will do that not what all the algorithm can do is it would be very, very good at generating quantum state right off the bat with machine learning. Right.
Alán Aspuru-Guzik:That can definitely then run be running a quantum computer. And then, if people are complaining about the the accuracy of of of the GQE, The GQE well, by the way, you can just keep running longer and lowers the energy.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:But you could always switch to VQE mode and now get the gradients. But now
Sebastian Hassinger:I see.
Alán Aspuru-Guzik:But now, you know, you're in a very good situation because you're so close to a minimum.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:Then you're not
Sebastian Hassinger:gonna starting from a much better ansatz, essentially.
Alán Aspuru-Guzik:You're starting basically at the right answer.
Sebastian Hassinger:Yeah.
Alán Aspuru-Guzik:The ansatz will be kind of like the circuit type, but within the ansatz, you will be very close to the right answer. Like, almost there.
Sebastian Hassinger:Interesting.
Alán Aspuru-Guzik:So you could just I mean, my my my dream, I mean, just talk about the future. Let's talk about right now, I think John Preskill, I think, is called MISC. I forget. I didn't see his paper very carefully, but he talks about the new era. I love John Preskill.
Alán Aspuru-Guzik:It's kinda like our guru of new era's life. He just starts talking in very elegant way as always, and John starts saying Yeah. Now we enter the source war era.
Sebastian Hassinger:We bring up Preskill once per q to b, and he tells us whether we're in a new era or not.
Alán Aspuru-Guzik:Yes. Exactly. Preskill. That's Preskill. Nice.
Alán Aspuru-Guzik:I always I always admire Preskill's elegance. I aspire I aspire to be, you know, as elegant as symbols, you can tell, like, where are my Mexican, whatever else? I have a different type of elegance. I am more like comedian. Okay?
Alán Aspuru-Guzik:And he's more of a formal anyway, I I like him a lot. He also makes making jokes. I like him. Anyway, the part is that he announced anywhere. And the question is, okay.
Alán Aspuru-Guzik:We have this error, and then we have the error corrected error. So if we extrapolate to the error corrected error of course, eventually, of course, all these algorithms that we also have worked on, but now here I'm gonna say, like, Dominic Barry and my former student, Jarrod McLean and also Ryan Babbush Actually, more Ryan Babbush than Jarrod, I think, in that case, have been working very, very hard on on this, you know, long term algorithms. My group also works in that. Recently, we have new work, by most and by anybody that I I wanna I wanna promote on on new ways of doing quantum simulation as well.
Alán Aspuru-Guzik:So we also work on that my colleague Nathan Wiebe. So we're also working by loop a little bit less focus, but also on the long term vision. Right. But I think, eventually, we wanna know my prediction in what would you wanna run a quantum computer in the future? You might run.
Alán Aspuru-Guzik:I mean, this is speculation. But you might run a GQE to get started.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:Then a GQE. And then the GQE is the guest for a for a full Hamiltonian simulation.
Sebastian Hassinger:Right. You seem
Alán Aspuru-Guzik:to have a good overlap with the ground state. So Yeah. I think all the algorithms that I've worked in my career and the others have worked in their careers, we've come together in a workflow that Sebastian, you you work with our colleagues at Amazon, then might run-in AWS. It might run-in IBM quantum computer. Well, and that Whatever he runs.
Sebastian Hassinger:Uh-huh. Incredibly good point, Alan, because, like, you know, you know, even predictions of about, like, what's the timeline till large scale fault tolerant quantum computers that's you know, depending on who which CEO you ask, you'll get different answers. But what's certain is that there are small numbers of logical cubits on the immediate horizon. Right? And we've heard announcements from Quantinuum, from Cuera, from IBM, from others who are all on, you know, the on the verge of starting to ship logical cubits.
Sebastian Hassinger:It'll be, you know, tens of logical cubits to begin with. But I I wonder, you know, if that speaks to exactly what you're saying. It's to to get the value out of, you know, a dozen, 2 dozen, maybe 50 logical qubits, you'll need these really sophisticated sort of hybrid architectures, solution architectures that that leverage the power, the best powers of classical computing and the best that we can deliver in quantum computing to to get to something that's together, the sum of the parts is more than, you know, than individually. Right? Like, that seems to be the direction we're going in.
Alán Aspuru-Guzik:Well, I mean, there's no quantum computers with the with the Yeah. Classical computers around them. I also have predicted together with Gerald, actually. Gerald I think Gerald, came up with the term QPU. I actually loved it when we when he showed me the draft of the the the draft of the variation of quantum ion solar paper.
Alán Aspuru-Guzik:Somebody can correct me if I'm wrong, but I think we are the first ones to call a QPU a QPU. Obviously, we were inspired by the GPUs. My group was also working on GPUs at the time. But QPU and the reason that VQE is very appropriate to call these things QPUs because they're like the GPU. They they they depend on the CPUs.
Alán Aspuru-Guzik:And, therefore, the quantum computer really has to be called a g a QPU. They will never be independent, and they are gonna be connected they are reconnected to, like, a massive supercomputer as we know or or the cloud or whatever. So so that's kind of gonna be the future, I think, for the from near time. And that's why, yeah, I think regardless of the era that we are on, if we're still in the relational era or new era that that that my friend John is is pushing, I am a I am a believer that and and some people are not. Sabrina Maniscalo just put out a paper in the archive with a bunch of colleagues about needs and tools about quantum computing.
Alán Aspuru-Guzik:And I paid attention to the one that says, we're gonna be able to do something before error correction. Yes or no? And most people believe no.
Sebastian Hassinger:Really?
Alán Aspuru-Guzik:Well, I'm still I'm still one of those 20% as agents. I mean, you have to be more creative, and that's kind of one of the things I wanna show. What can we do with an pre error correct quantum computer? I'm still very optimistic that with enough quantum, lemon, we can do lemonade.
Sebastian Hassinger:Yeah. And and you mentioned, you know, the the GQE, technique could use just about any form of generative, model, for the for that front end, for the generative part. Right? So is are you thinking of potentially, with with VQE, you your group created the tequila framework, which is a a sort of an ex extensible software framework for for VQE, experiments. Are you thinking about doing the same thing with GQE?
Alán Aspuru-Guzik:As a Mexican, yes. I like to work on tequila. Good point. Now
Sebastian Hassinger:Is it called mezcal?
Alán Aspuru-Guzik:Yeah. For former former postdoc and now independent faculty member, Jakob Kottmann in Germany. Although I'm I'm just he's from Germany, but I have actually is he based now in the Netherlands? No. He's German Germany.
Alán Aspuru-Guzik:I don't remember. I have to Google it right now. Just to tell you where he is, Jakob Kottmann. You can hear my keyboard here. In in Oxford.
Alán Aspuru-Guzik:Yeah. In the University of Oxford. So, so he's in Germany. Okay? So, anyway, Jakob, which is a rock star, he still is maintaining mostly, tequila.
Alán Aspuru-Guzik:So tequila, just advertising here in the podcast, is still a a a a mundane package with dream new stuff. We don't have the GQE connected to tequila, but, obviously, it should be a it should be a necessity to do that soon. And and, yeah, I think you correctly told me that people have not got on to the GQE yet. I get this from my friend, Shaul Mukamel, very famous theoretical chemist, that told me, Alan, you you kinda wanna work 10 years ahead of everybody. Yes.
Alán Aspuru-Guzik:20 years ahead is too much because then people don't know what to do. Yeah. But 10 years ahead is kind of what you should do. And, you know, like, if you actually look at my publications now in my scholar, Google Scholar, the VQA paper keeps keeping up steam. And 10 years ago, it was not that much cited.
Alán Aspuru-Guzik:And I think the GQE will be the same once people realize, and thank you. Thanks to your post podcast, Sebastian. Maybe some of your listeners will be like, oh, what is GQE? And I didn't see this paper in the archive. Obviously, there's so much stuff coming up, but Links in the show notes, people.
Alán Aspuru-Guzik:Exactly. You just you just, you just figured it out that you know, it's helped me out to kinda popularize this. And and yeah. So I am working on it in my lab, in collaborations with Kohei at NVIDIA and, collaborations across companies and across institutes, across, academic and national labs. And so it's it's it's gonna be fun to see what comes out of the of the GQE era.
Alán Aspuru-Guzik:But, yes, as you say, right now, the GQE is one paper and a bunch of little mini eggs that are about to hatch.
Sebastian Hassinger:Right.
Alán Aspuru-Guzik:So I'm gonna I'm gonna not gonna disclose what we're doing. But Fair enough. But I not in the interest of my students' and postdocs and quarters and
Sebastian Hassinger:Absolutely.
Alán Aspuru-Guzik:Respecting their privacy. But I I have to say this, you can remind many extensions of the GQE because one of the things I'm gonna say is that GQE is not only for the VQE type problem. You can use GQE for variation of quantum classifiers.
Sebastian Hassinger:Oh, k.
Alán Aspuru-Guzik:So you can you can you can create all sorts of all the variation of algorithms with this. So you can do it for machine learning. Returning to machine learning so you can imagine there's a lot of cool things you can do with the GQE type frameworks. So I think GQE would be the the perfect match between machine learning and quantum computing, and people will not remember quantum computing without any machine learning computation of the circuit, and this paper will be the the seed of that. I'm pretty optimistic that it will be my most heavily cited paper 10 years from now.
Alán Aspuru-Guzik:People right now haven't seen it yet, but it's okay. You're always gonna be 10 years ahead of the of the of the curve.
Sebastian Hassinger:That's perfect.
Alán Aspuru-Guzik:That's something I also learned from Seth Lloyd. Seth Lloyd is kinda like that too.
Sebastian Hassinger:Yeah. It's guru territory.
Alán Aspuru-Guzik:No. Not guru. Yes. First of all, Latin America will be called shaman, not guru. Okay.
Alán Aspuru-Guzik:That's true. Yeah.
Sebastian Hassinger:Yeah. Shaman. Fair. Yeah. Fair.
Alán Aspuru-Guzik:Wrestling shaman. Yeah.
Sebastian Hassinger:Okay. Well, that's fantastic. I am dying to know what your group is working on, but I will respect the the time that you're postdocs and, and students need to work on it. But that means we're gonna have to have you back or members from your group back to discuss that work when it when it does go public. So, you're gonna have to promise that.
Sebastian Hassinger:So that's fantastic. I I really appreciate your time. This is super interested. I interesting. I I'm absolutely agree with you that there won't be quantum computing without, you know, sort of that preprocessing that'll be machine learning based.
Sebastian Hassinger:I think there's, you know, AI for quantum as a category is going to be an increasingly important topic and important area for exploration. I think this paper will turn out to be foundational in in all of that area. So thank you so much for joining us.
Alán Aspuru-Guzik:Thank you, Sebastian, and everybody stay tuned about what's coming out. Very good news coming out from our side. And, thank you very much, Sebastian, and and and, yeah, we'll stay in touch at all with you and I. Okay?
Sebastian Hassinger:Absolutely.
Kevin Rowney:Okay. That's it for this episode of The New Quantum Era, a podcast by Sebastian Hassinger and Kevin Rowney. Our cool theme music was composed and played by Omar Costa Hamido. Production work is done by our wonderful team over at Podfi. If you are at all like us and enjoy this rich, deep, and interesting topic, please subscribe to our podcast on whichever platform you may stream from.
Kevin Rowney:And even consider, if you like what you've heard today, reviewing us on iTunes and or mentioning us on your preferred social media platforms. We're just trying to get the word out on this fascinating topic and would really appreciate your help spreading the word and building community. Thank you so much for your time.
Creators and Guests
