Episode 1: Dr Yihui Quek
How do we understand the process of learning, and what does it have to do with quantum computers? Take a listen to Episode 1 of insideQuantum to find out!
This week we’re featuring Dr Yihui Quek, a postdoctoral researcher and Humboldt Fellow at Freie Universität Berlin. Dr Quek did her undergraduate degree at MIT, followed by postgraduate studies at Stanford, and is currently in her first postdoctoral position.
🟢 Steven Thomson (00:13): Hi there and welcome to insideQuantum, the podcast telling the human stories behind the latest developments in quantum technologies. I’m Dr. Steven Thomson, and I’ll be your host. In this episode I’m really happy to be joined by Dr. Yihui Quek, a post-doctoral researcher and Humboldt fellow working in the field of quantum information. Yihui, thank you so much for joining us today.
🟣 Yihui Quek (00:33): Thank you for having me here.
🟢 Steven Thomson (00:35): Okay. So you’re here as a researcher working broadly in the field of quantum information, is that correct?
🟣 Yihui Quek (00:40): Yeah, that’s right.
🟢 Steven Thomson (00:41): Okay. So what first got you interested in quantum physics?
🟣 Yihui Quek (00:45): Well, I guess I was, I went to a math and science high school when I was growing up in Singapore. And so I was always interested in physics and math and my high school had this requirement that we had to do research in order to graduate. So I did a small research project with a, with someone who’s currently a professor in Singapore, Dr. Ng. And even though my research project wasn’t on quantum and she was mostly working on quantum, I kind of like, was always interested to hear about what she was actually working on. And so I guess just kind of by diffusion, I learned a bit about quantum and it sounded very interesting. And when it was time for me to study physics or when it was time for me to go to university, I chose to study physics and I went to MIT. So MIT is pretty strong at everything to do with quantum science and technology. So it was a pretty natural choice for me.
🟢 Steven Thomson (01:36): So what was this research project that you did in high school, can I ask?
🟣 Yihui Quek (01:39): Oh, it was in compressive sensing.
🟢 Steven Thomson (01:41): Oh, interesting.
🟣 Yihui Quek (01:42): Yeah. I built a single pixel camera. Probably the only time I’ve ever successfully done an experimental research project.
🟢 Steven Thomson (01:49): That’s amazing. I wish I’d done things like that in my high school.
🟣 Yihui Quek (01:52): Yeah, it was, it was super interesting. I learned a lot about signal processing, which, I mean, it, isn’t totally unrelated to physics. There’s a lot of matrices involved. And so I got a taste of linear algebra at that point as well.
🟢 Steven Thomson (02:04): And it didn’t scare you off?
🟣 Yihui Quek (02:05): No, it was actually, it was actually super cool. I think it was just right mix of being hands on and being theoretical. So I could see that the single pixel camera I built was reconstructing images very well. And I was curious about how we could do that with just a single pixel. So it was really an excellent project to introduce me to the fundamentals of signal processing.
🟢 Steven Thomson (02:24): Nice. So then from Singapore to MIT, why did you decide to go there? Was it just the reputation or were there any specific opportunities that MIT offered you that other places didn’t have?
🟣 Yihui Quek (02:36): So I think my answer to this comes in two parts. The first part is that just going out of Singapore was a life changing opportunity in itself. And the reason for that is that I had already spent by that time, 18 years in Singapore. And when you live in one particular place for so long, you become so optimized for that particular environment that you lose your ability to adapt when your environment changes. And this is all okay when you’re young, but as you get older, I tend to believe that you have very little control over the kinds of events that could be that could be sudden changes in your environment and you have to gain the ability to adapt to them and make corresponding adjustments in how you behave and your psychology. And so I feel that moving to a different country was a very good way for me to learn that kind of skill.
🟣 Yihui Quek (03:29): So that was why I chose to leave Singapore. And as for why MIT in particular I think that something I learned from MIT, which I find very invaluable even today is the idea that I actually have a great degree of control over my own environment. And I can do things that change my environment for the better. And so what I learned was this whole idea of ground up initiative. I think growing up in Singapore, I was fortunate to be in a place of not only a lot of material privilege, but also I think that I happened to be lucky to be in an environment that priced the qualities that I already possessed. Namely, I think that in Singapore, as in many Asian societies having the ability to get good grades in school is actually, it actually puts you very far ahead in life.
🟣 Yihui Quek (04:18): And because of that I felt that I was always able to kind of cruise without thinking too much about what I would like to change in my environment. But when I got to MIT, of course it was still the case that it was important to maintain good grades. But I also came into contact with a lot of people who had a lot of what you would call side hustles in the sense that they were not only getting good grades, but they were also kind of using their skills to make a real impact to their communities. And I was very inspired by this. For example, one of my dorm mates is Thai, and at the time that we started at MIT, Thailand had just experienced some really bad floods. And I saw that he had actually made a difference to these efforts by programming an Android app to enable users to share their geolocation data so that in the event of these floods, they could help other people and also be helped. So I was very inspired by that and I think I’ve come away now with the with a lot more willingness to kind of be the change I want to see in whatever environment I’m in.
🟢 Steven Thomson (05:24): So moving to another country then, it was a really life changing opportunity.
🟣 Yihui Quek (05:28): I think it was a really life changing experience and I would recommend it to everyone else.
🟢 Steven Thomson (05:33): At what point did you decide that you wanted a career in quantum physics rather than just studying it for interest? At what point did you decide? No, this is it, this is something I actually really want to pursue. And I really want to work on properly long term.
🟣 Yihui Quek (05:45): I wouldn’t say that there was one particular point. It was more of a gradual transition. Actually, when I started my PhD, I was going to do classical information theory. And, but this was already after a bit of experience in my undergrad with quantum information. So I had taken a class at MIT with Peter Shor and as the final project for the class, I wrote a review of part of the Neilsen and Chuang textbook, basically the parts to do with quantum information measures. And I had found that very interesting, but when I started my PhD, it didn’t immediately occur to me that I wanted to do quantum information. And in America, the PhDs are very freestyle, I would say like, you don’t have to commit to something or you don’t have to commit to a particular group before you start. And so it was definitely a conscious choice for me to go into quantum information or rather to go back into quantum information. And I think I’m still refining my subfield within quantum information. I think at the start of my PhD was more about quantum Shannon theory, and now I’m going more towards quantum learning theory. So I, I think along with this refinement is also the fact that I’m finding myself enjoying what I do more and more. So that’s probably a good sign. But to come back to your question, I don’t think there was one moment when I told myself, okay, I’m gonna be a quantum physicist from now on.
🟢 Steven Thomson (07:10): So for you, it was a gradual…
🟣 Yihui Quek (07:12): It’s a gradual learning process and a gradual increase of my enjoyment levels as well.
🟢 Steven Thomson (07:16): Oh, nice. That’s fantastic to hear.
🟣 Yihui Quek (07:18): Yeah. And I hope it continues increasing.
🟢 Steven Thomson (07:20): Let’s hope so! Okay. So you mentioned a couple of aspects of what it is that you do. So before we dig too deeply into what any of these subfields are, can you tell us what, what’s the big picture, what’s the area of research that you work in and what’s your contribution towards this big picture?
🟣 Yihui Quek (07:35): Mm-Hmm. So I think there’s the big, big picture, which is to build a quantum computer or working quantum computer. And then there’s the ‘big but not so big’ picture, which is kind of like an intermediate picture, which is to get a working prototype of something that resembles a quantum computer, and that’s the stage that we’re at now. And I think a lot of what I do is working in this intermediate picture. So what we have right now are not quantum computers, but we have quantum processors consisting of on the order of a hundred qubits. And of course, for a working quantum computer, you would need like much more qubits than that. I think to run Shor’s algorithm, you would need on the order of a thousand logical qubits, which is like a million non error corrected qubits or something like that.
🟣 Yihui Quek (08:24): But the point is what we have now is much fewer qubits than what we want. And so the question is what can we already do with so few qubits? And a lot of what I do is revolving around that question. The thing is that with our current quantum processors – I’m gonna call them quantum processors because they’re not actually quantum computers. They’re so small and they suffer from a lot of noise. So it’s… I think when people were writing all these early quantum algorithms, they definitely did not bank on there being so much noise. And so a lot of quantum computation is built on the assumption that you’re running an error corrected quantum computer. And so my, the question I’m trying to answer is what can you do, even if you’re not running an error corrected quantum computer. And more than that, if you’re running a super noisy and super small quantum processor. So I use concepts from classical learning theory, classical statistics and information theory, both quantum and classical to kind of answer this question. And it’s a really interesting question that sits at intersection of physics and computer science, because you need physical concepts to understand how quantum mechanics works in the presence of noise. And you also need computer science to kind of like figure out how to make that into an algorithm or what kinds of limitations you can expect in this kind of regime. Okay. So I really enjoyed that.
🟢 Steven Thomson (09:48): Okay. I see. I see. So the question really is what can we do with current generation processors?
🟣 Yihui Quek (09:55): And what can we not do? That’s also very important.
🟢 Steven Thomson (09:56): What can we not do? Yeah. So we’re, we’re on the way towards a quantum computer, but we’re not there yet. We’re taking the first steps. Okay. So you mentioned there a couple of times, noise and error correction. Where does this noise come from? Why are quantum computers noisy? What’s the problem with having a thousand qubit processor?
🟣 Yihui Quek (10:13): Well, I think a lot of the noise basically comes from an effect that qubits like to entangle themselves with their environment. And I think, you know this a lot better than me, but I think what I work with are mostly models of noise, but the intuition behind these models is the fact that we don’t really have fine enough control of our qubits to make sure they are behaving in the way we would, we would want to operate them as you would need to do if you were really building a quantum computer. And so, yeah, maybe you can talk more about why these things are, why these quantum processors are so noisy.
🟢 Steven Thomson (10:55): Well, I guess the point here is that it’s difficult to control things on a quantum level, isn’t it? That it’s, it’s easy enough to work with classical electronics and to assemble them all into bigger, bigger structures, more and more complex structures, but these quantum degrees of freedom, these quantum bits, qubits, they’re so small that it’s kind of hard to get them to do what you tell them. They like to..to yeah, entangle with their environment. They like to pick up environmental effects and decohere. And it’s very difficult to get them to stay in the state that you want them to stay in. Is that a fair statement?
🟣 Yihui Quek (11:27): I think that’s a fair statement. Yeah.
🟢 Steven Thomson (11:28): Okay. And then, so your work is, is about understanding the limitations of the equipment that we have now, maybe overcoming these limitations and figuring out what can we really do with this, this kind of first…first generation of quantum hardware?
🟣 Yihui Quek (11:43): Yeah. You could call them that, I think. Yeah.
🟢 Steven Thomson (11:45): Okay. So we’ve covered a little bit about what it is that you’re doing at the moment. If you weren’t doing this, if you hadn’t gone down this route of quantum information, quantum computing, what do you think you would’ve done instead?
🟣 Yihui Quek (11:58): Well, I think I enjoy communicating a lot. I enjoy writing. And so I think I could be a science journalist. I particularly enjoy reading the articles from Quanta. Do you know about Quanta?
🟢 Steven Thomson (12:10): I do, they write some excellent articles.
🟣 Yihui Quek (12:12): They have some amazing articles. Yeah. And I’ve learned a lot from reading these articles and I think it’s like, well, at least in my experience, it’s one of the best science journalism outfit. So I would want to be a science journalist for an outfit like Quanta, but also I think I could be a data science journalist and what I have in my mind now is this political website FiveThirtyEight.
🟢 Steven Thomson (12:32): Ah, yes. I’ve heard of it.
🟣 Yihui Quek (12:33): Yes. And I think it could be very cool to work for such an outfit because you not only spend time crunching numbers, but you also kind of make a political commentary out of that. And I think that would be a really good combination of things that I really like numbers and writing.
🟢 Steven Thomson (12:48): So it’s not just about the maths or about the physics. It’s also about the meaning behind it, of what meaning can you extract from this big pile of statistics and data?
🟣 Yihui Quek (12:56): Oh, definitely.
🟢 Steven Thomson (12:57): And how can you communicate that to an audience who might not be so literate in, in data sciencebut want to know the story behind the data.
🟣 Yihui Quek (13:05): Yeah. I, in fact, I think as scientists, it’s easy to kind of get lost in the, I guess, math or the beauty of math, but I think it’s also important to communicate your findings to a wider audience. And I think that that’s not only good for getting funding, but it also…often when you try and form a narrative around your work, it makes, it makes it clear in your own mind. So I really enjoy that.
🟢 Steven Thomson (13:35): Yeah, definitely. I think that makes a lot of sense. I think, to explain a concept properly, you need to understand it yourself and sometimes until you try to explain it, you don’t realize that you don’t understand it.
🟣 Yihui Quek (13:45): Yeah, exactly.
🟢 Steven Thomson (13:46): You can be so used to just talking in jargon with other people in the field. And then when you have to explain one of these technical terms, you realize, wait a minute, I, I don’t know what that means, in simple words,
🟣 Yihui Quek (13:56): In fact, I think it was Alan Guth at MIT who, who once gave us a piece of advice when he was teaching one of our classes, he said the best way to learn this material is to teach it to someone else. And if you don’t have someone else to teach it to just pretend there’s someone else there.
🟢 Steven Thomson (14:11): That sounds like some pretty good advice. Okay. So we covered a little bit about what you did at high school and going to MIT. Can you tell us a little bit about the type of work that you did while you were at MIT, the kinds of things you’re interested in and then how that led onto your PhD and what you worked on during your PhD?
🟣 Yihui Quek (14:29): So, as I mentioned at MIT in my last semester, sorry, the last semester of my junior year, which is my third year, I took a class on information theory and it was taught by Peter Shor. And as part of that class, well, it was a class on classical information theory, but at the end of it, I had the opportunity to write a, sort of a review article on quantum Chan theory, which is like a quantum version of information theory where instead of manipulating probability distributions, you’re manipulating density matrices, which are like a quantum generalization of probability distributions. And at the end of my time at MIT, I also had the opportunity to write a senior thesis with Peter and that was on studying something called super quantum correlations.
🟢 Steven Thomson (15:20): Super quantum. Wow. That’s a term I’m not familiar with.
🟣 Yihui Quek (15:22): Yeah so even more quantum than quantum correlations.
🟢 Steven Thomson (15:25): More quantum than quantum. Nice. Okay.
🟣 Yihui Quek (15:27): And so we managed to find an application to a communication protocol. And so all of that was very interesting. And when I went to Stanford I knew I was interested in information theory. And so I guess I started out trying to work on the link between information theory and biophysics. But then I realized that Ithere wasn’t really anyone at Stanford who was doing that. So I then went back to what I was more familiar with, which was quantum information theory. And then I wrote a couple of papers on studying classical feedback over quantum channels. And I think I asymptoted to quantum computing in the end because…well, this was not related to my paper, but but I felt that quantum computing was an area that had tremendous growth, potential, and a tremendous number of opportunities.
🟣 Yihui Quek (16:24): And of course while all of this was happening, Google had just made their announcement about their their very first quantum supremacy experiment. And I think that really opened my eyes to the number of new questions that would be enabled by, by this new technological breakthrough. And so that’s why I decided to eventually work more on quantum computing and …oh, something that’s very interesting! Why do, why is it that I like learning theory so much? Well, I think it’s because I went to Stanford for my PhD and at Stanford, everyone – and by everyone, I mean like 90% of people, including humanities majors – take the machine learning course. Machine learning is a very big thing in Silicon Valley.
🟢 Steven Thomson (17:10): Oh, yes, I can. I can believe that.
🟣 Yihui Quek (17:12): Exactly. And so I also took the machine learning course, even though I was in the physics department. And I mean, I took one of the many machine learning courses that they offered and I think this was my very first introduction to kind of…the field of learning theory. And I realized that learning theory is not just useful for explaining why classical machine learning works so well, but it’s also kind of a more fundamental kind of theory that can explain or can be adapted to many more computational situations, including what I’m currently using it for, which is to characterize quantum systems in quantum states.
🟢 Steven Thomson (17:54): I see. Wow. That’s pretty comprehensive then. There’s a lot of different concepts all mixing together in your research all coming together in a really interesting way. Okay, so you briefly mentioned there quantum supremacy, which I think is sometimes also called quantum advantage. Can you say a little bit about what that is and why it’s so important?
🟣 Yihui Quek (18:12): Right. So I think a lot of it’s important it’s more that it’s a benchmark for quantum computing because it’s the first time that a quantum device has been able to perform a task that a classical device cannot. And even…and the thing to note is that the task that this quantum device has performed is a completely useless task. It’s basically sampling from some kind of probability distribution that you wouldn’t be able to sample from so quickly, if you only had a classical computer.
🟢 Steven Thomson (18:45): I see. Is it a controversial statement to say that it’s a useless task?
🟣 Yihui Quek (18:49): No, it’s not. It’s widely accepted that it’s useless.
🟢 Steven Thomson (18:51):
🟢 Steven Thomson (18:54): But it’s the first time a quantum system has been able to do something a classical system could not. So it’s still a landmark in that sense. I see. I see. Okay. And then part of what you’re doing with trying to understand how these quantum computers work and what they can be used for, I guess, is finding more cases where quantum systems can have this advantage over classical systems, more cases where they can do things that, that other systems until now have never been able to do?
🟣 Yihui Quek (19:19): Exactly. And I think one of the most promising candidates for quantum advantage, or like the candidate that seems to have convinced most people is basically simulation of other quantum systems. So I, I think it was Feynman who said something like…oh my gosh, at this point you have to insert the Feynman quote!
🟢 Steven Thomson (19:37):
🟢 Steven Thomson (19:38): We’ll put it in in the edit…!
🟣 Yihui Quek (19:38): Something about simulating nature and by golly it’s a wonderful computational problem.
🟢 Steven Thomson (19:43): Yes. I know the quote that you’re referring to. Yes. I know the one.
🟣 Yihui Quek (19:48): Yes. And I think it’s an open problem and a very interesting one to figure out if even now with our current quantum processes, we can already simulate a small part of nature.
🟢 Steven Thomson (19:58): Yeah. I mean, that certainly seems like a very worthy goal. So coming from more of the many-body physics side, if you want to exactly simulate a quantum system beyond about, I don’t know, 20 or 30 electrons, suddenly you need more memory than even super computers have. By the time you get to 40 electrons, you, you just give up all hope. So you know, classically we’ve been able to simulate these large systems seems entirely hopeless. So being able to use quantum devices to directly access this quantum physics and directly simulate the things that we’re interested in. Yeah, I…you’ve convinced me I’m sold. I think it’s, it’s a very worthy cause, indeed.
🟣 Yihui Quek (20:33): Great. You wanna write a paper together?
🟢 Steven Thomson (20:35):
🟢 Steven Thomson (20:38): So I guess one thing I would ask not being in this field, you’ve used the phrase learning theory a couple of times. Can you maybe give us a kind of quick definition of what you mean by learning theory?
🟣 Yihui Quek (20:47): Ah yes. Very good question. So it’s been observed that. Okay. So I guess we’re at a period of time when machine learning and AI is at the forefront of our consciousness, but the thing is that they are extremely effective and we don’t know why. And some, some people have even gone so far as to term this the unreasonable effectiveness of machine learning and learning theory is kind of an attempt to build a theoretical framework. That explains why it’s easier to learn some kinds of, well, it’s easier to learn some things than others.
🟢 Steven Thomson (21:25): Okay. I see. Okay. So machine learning really is just a black box then.
🟣 Yihui Quek (21:31): Yeah. And learning theory is one of the candidates for opening this black box.
🟢 Steven Thomson (21:35): Okay. I see, I always had the impression – being slightly outside this field – that people who worked on machine learning knew how it worked, and it was just people like me who heard some of the buzzwords and were aware that it existed…I thought it was just, you know, outsiders didn’t know how it worked, but the people who worked on it, they had a good idea. But you’re telling me that actually no, no one, no one knows how it works?
🟣 Yihui Quek (21:58): I think that’s been a lot of progress in understanding how it works.
🟢 Steven Thomson (22:00): Okay. I see.
🟣 Yihui Quek (22:01): I think in fact, one criticism of learning theory is that it’s too theoretical and it makes certain assumptions that aren’t really met in real life, but it’s still a useful theory.
🟢 Steven Thomson (22:10): Does it only apply to, to things like machine learning or does it also apply to, for example, you know, children learning in school or to, to the way that humans learn things?
🟣 Yihui Quek (22:18): Oh that’s a very good question. I think learning theory is a theory that was built in the image of machine learning, but it can definitely be applied to other kinds of learners. But I think if you wanted to make it applicable to children learning things in school, you would have to relax some of the assumptions. Like for example, one of the assumptions of learning theory is that it studies the behavior of learners, but a learners is modelled as something that collects a sample of the kind of thing that it’s trying to learn. And instead of saying “thing that’s trying to learn”, I’m gonna call it a concept. So statistical learning theory – which I’m calling learning theory – models the learner as an entity that is able to observe the unknown concept in a certain very structured way, and this way of the learner observing the concept is by observing…
🟣 Yihui Quek (23:10): So a concept is a function that maps from a domain to a range and the learner kind of picks or, well, not necessarily picks, but let’s say picks, okay. The learner picks certain points in the, in the domain and then it observes what the unknown concept maps to on those points, but it doesn’t observe the unknown concept itself and it’s supposed to figure out what that is. But anyway, my point was that this is a very structured way of learning. And it may be the case that when you want to think about a child learning a concept, that’s not how a child actually observes the concept.
🟢 Steven Thomson (23:51): Okay. So kids are messy and unstructured, is the takeaway from this.
🟣 Yihui Quek (23:54): Yes. And they may, yeah.
🟢 Steven Thomson (23:57): Okay. Okay. So I see the theory is developed for machine learning, may have more general uses, but to get to those more general uses, probably we have to relax a lot of these assumptions and abstract away a little bit from the current theory. Okay. Okay. You’ve talked about working in quantum information and quantum computation, you’ve talked a little bit about learning theory. What do you think the biggest outstanding challenge in your field is at the moment let’s say in the near future? Okay. The end goal is probably to, you know, construct a quantum computer, but let’s say five to 10 years. What do you think the biggest outstanding challenge in your field is?
🟣 Yihui Quek (24:30): Well, I’ve talked about this a little bit just now, but I think the biggest intermediate scale challenge is to find applications for, or find use cases for noisy quantum computers. So if we could find such an application that would give us more confidence, that we should continue investing in quantum computing research and build bigger ones and better ones.
🟢 Steven Thomson (24:48): Okay. I see. I see. And what happens if we don’t find any signs of quantum advantage?
🟣 Yihui Quek (24:53): Well, I think that even if we don’t find, or I think that different potential applications have different levels of plausibility, I think it’s quite likely that we will find a quantum advantage in simulation. It’s less clear to me that we will find a quantum advantage in the other use cases. I’m particularly interested of course, in the quantum advantage in machine learning, but I think there’s still a lot that needs to be really hammered out before we can conclusively say that there is a quantum of advantage there.
🟢 Steven Thomson (25:24): Okay. I see. And if if we were to suddenly find this quantum advantage, what would be your hope for the impact that this would have outside of the research community? Because I could see, for example, quantum simulation, being able to accurately simulate quantum systems is something that would interest a lot of people in research. It certainly interests me, but for people who are not maybe working in quantum technologies or, you know, not at all interested in the math and the physics of it, what would you hope that the impact of a discovery like this would be on, on the wider world?
🟣 Yihui Quek (25:56): Hmm. I’m not sure that it would have a direct impact on most people. I think what I can envision is a future in which maybe part of a computer that people are using is quantum, but it’s only, it’s kind of like only a part that, that serves the specialized purpose for which the quantum computer is good for. I don’t think we would be able to use quantum processors to like run Microsoft word for instance.
🟢 Steven Thomson (26:24): Okay. So a bit like we have a…our computers these days have you know, CPUs, we have GPUs, we have specialized components, you can think of having a…a QPU I guess a quantum processing unit? Something like that, that just sits in the corner and does specific quantum tasks, like, I don’t know, cryptography or something. Yeah. Something like this. Okay. That’s interesting. I’ve never heard a suggestion like that before. I always had the idea that quantum computers would be very large things living in labs, somewhere solving very specific problems. I never kind of had the, had the impression that you could yeah, just…just take quantum computers for what they’re good for and then build that into real technologies. That’s a really interesting thought. Okay. So we’ve talked a little bit about your journey so far. So you started in Singapore, then you moved to MIT for your undergrad and then to Stanford for your PhD.
🟢 Steven Thomson (27:14): And you’re now here in Berlin. You’ve been here now for how many months now? Three months, six months? Five months. Okay. You’ve been here for five months. So I guess it’s safe to say that you’ve worked in quite a few different countries and very different cultures. You’ve worked in several different countries to this point, and I think it’s safe to say that they’re all very different to each other. Have you experienced any form of, let’s say culture shock, any surprises in moving from one country to another and anything that you’ve discovered that you didn’t expect before you move to one of these countries, be it in academia or just in the broader culture of the country?
🟣 Yihui Quek (27:49): Yes. So I’ve already talked about how there’s a lot of technological optimism in Silicon valley. And I think the social attitude towards new things or wild claims is very different here in Germany. I think people here tend to exercise a lot more skepticism when they encounter something that seems too good to be true. And I think this doesn’t just apply to technology, but it’s an attitude that’s kind of prevalent, even in daily life. For example, I’m currently writing a paper with some colleagues here in the Eisert group, and I’m always surprised by how much time we spend agonizing over every sentence to make sure that we’re not overselling our results. So I think this kind of, so I think there are upsides and downsides to both approaches on the one hand, I think the American enthusiasm can act as a sort of inoculation against the fear of failure, whether it’s in their work or in their personal lives.
🟣 Yihui Quek (28:50): But on the other hand, I think the German attitude of sobriety and kind of more moderation also comes in handy when you want to evaluate the claims made by someone else and yeah. See whether it’s true. So ideally I would like to combine both of these approaches. I would like to have the American exploratory spirit when it comes to doing experiments to figure out certain things about myself or my environment. But I think when it comes to reflecting on the results of those experiments, I would like to have more of the German sobriety.
🟢 Steven Thomson (29:29): That sounds like a good and very necessary balance to have, I suppose. It wouldn’t do to be completely depressed about all future technologies and insist everything be perfect all the time, but equally, yes, I guess you can have you can have crazy ideas flying high without having something to really underpin them. So yeah, that sounds like a very necessary balance.
🟣 Yihui Quek (29:51): And in terms of social attitudes, I’ve also noticed a difference between America and Germany. So I think in America, or at least in California, there’s a lot of talk about diversity. And the reason for this is that I feel that in America, demographics are quite central to people’s identities. So people are always very conscious of like what race they are, what gender they are. And they’re very conscious about the history, about the historical background of their ancestors, as well as the challenges that people of the same identity as them are currently facing. And this comes up a lot in conversations in America. Whereas I feel that it doesn’t so much here in Germany, and I think this may have something to do with the fact that many people in America have immigrant backgrounds. And so I think it’s fair to say that it really is a more diverse place.
🟣 Yihui Quek (30:47): And so that’s why there’s more opportunities for these different identities to kind of rub against each other and potentially cause conflict so that’s another difference I’ve perceived. And I think it would kind of be, I mean, I would prefer to see more, more considerations of people’s identities in Germany as well. So yeah, for example, I think this is one of the other questions you were going to ask me but I noticed that, especially in, in physics, it tends to be very male dominated here in Germany and I think that I didn’t really notice this when I was studying in America because like I think there’s more efforts to kind of promote diversity in the workplace in America. I mean, it was definitely still male dominated, but the generation wasn’t as extreme as it is here.
🟢 Steven Thomson (31:41): And do you think that this is due to efforts in America to promote diversity that are not quite so prevalent here in Germany?
🟣 Yihui Quek (31:47): Well, it certainly isn’t talked about as much here, and I think part of this is due to the same kind of skepticism that I mentioned earlier where I think maybe the German image of a stereotypical Yankee is like being all talk and no action and maybe that can be an impediment to people even starting the conversation. But I just wanted to point out that it’s actually important for efforts to promote diversity to be part of an ongoing conversation. And personally, as someone who has lived in both kinds of environments, I think that it makes a real difference to my own experience when this sort of thing is common knowledge. And when I say common knowledge, I mean, it in the philosophical sense where you say a concept is common knowledge if not only everyone knows it, but everyone knows that everyone else knows it.
🟣 Yihui Quek (32:36): And everyone knows that everyone knows that everyone else knows it. And so on ad infinitum. And the reason I think it’s important for these kinds of things to be common knowledge is that when you know that everyone else knows it, then that becomes the foundation on which you can begin to start discussing how to correct it. And so I think that maybe the very first step that needs to be taken is for these efforts to become common knowledge even here in Germany. And I would say that that is a necessary, but not sufficient condition for real change to happen.
🟢 Steven Thomson (33:09): So in your experience so far, I guess, across all countries, as you say, the physics in particular has been a very male dominated field for a long time. To me, it feels like things are improving, but we still have a long way to go in your experience. Do you think things have changed over your career at all?
🟣 Yihui Quek (33:28): Well, I think maybe things have changed because I was, I myself was moving towards environments that I felt were better and more nurturing for my own personality. So at Stanford I was actually hanging out quite a bit with computer scientists. And actually I felt that theoretical computer science, the generation is really good in Stanford. And I don’t know if it’s because Stanford has managed to hire a lot of really excellent female faculty in computer science in the Stanford computer science department and its affiliates. I know of at least two women who are, I think doing a fantastic job in increasing representation and making women feel really at home. So these people are Mary Wootters and Tselil Schrammis. Tselil is actually from the department of statistics, but just before I left, they had started this initiative called the women’s theory forum, which was a monthly get together for women in like the theoretical computer science and like related fields to kind of share about their research. And it just felt very, it just like…it just felt like a very natural thing to do, and I really enjoyed these gatherings. So…And also not only women were invited to these gatherings, of course, men were also very, very welcome. So I feel like these are the sorts of things that we need to have here in Germany as well.
🟢 Steven Thomson (34:52): Okay. So I guess you’ve talked a little bit about gravitating towards environments that you felt were particularly productive and nurturing for you. What would you say to any other women in particular, wanting to get into a field like physics, computer science? One of these fields that looks very male dominated and indeed is very male dominated. Would you have any advice for women looking to get into these fields about where they could look to find a similarly nurturing experiences you’ve had?
🟣 Yihui Quek (35:26): I think what comes to mind are two pieces of advice. The first is that you have to be unoffendable. So I think I’ve noticed at least in myself, that I used to be very conscious of or self conscious about how I was perceived. And this would kind of be a barrier to me, really expressing my thoughts on a particular thing. And it was a very stupid barrier because it prevented me from learning properly because I would feel, oh, this question I’m about to ask sounds so stupid. And I don’t wanna ask the question and because I didn’t ask the question, I didn’t learn, but I think nowadays I’m becoming more and more unoffendalbe, which means that I no longer have this barrier. And I say what I’m thinking, even if I think it sounds stupid.
🟢 Steven Thomson (36:10): And as a consequence you’ve…
🟣 Yihui Quek (36:11): And as a consequence, I think I’m learning more.
🟢 Steven Thomson (36:14): Brilliant.
🟣 Yihui Quek (36:16): And the second piece of advice is kind of always be very conscious of the environment that you’re in. I think I did not understand the importance of placing myself in a good environment when I started my PhD, but I slowly learned that applying a strict filter on my environment was one of the most effective and efficient ways of improving myself as well. Because when you’re surrounded by people who are positive, hardworking, and who enthuse you, then you find that you become one of these people as well. So if you just put in the effort at the start to select the right environment for yourself, then that can save you a lot of effort later on, namely the effort of kind of like making yourself excited about your own research, which can be very difficult when you’re kind of in a place where no one else cares or in, or if you’re in a place where like there’s a lot of hubris and people are kind of like always talking over each other and it’s very antagonistic. So always try and avoid these environments and go to positive and constructive environments instead.
🟢 Steven Thomson (37:19): That’s a really nice thought actually. Yeah. That’s a really nice way to look at things, to seek out positivity that makes you feel better about what you do makes you actually better at what you do surrounding yourself with people who actually support you and encourage you.
🟣 Yihui Quek (37:34): Oh, there’s actually another thought that I just had. So I think that I’ve also benefited a lot from having many positive female role models in my life. So as I mentioned, just now during my high school times, I was fortunate enough to meet Dr. Ng and she was a positive female role model and it made it, made it a lot… I think meeting her was kind of the catalyst that drove me to realize that there was an alternative career path to, that I could pursue besides being a doctor or a lawyer, which are the two most popular career paths in Singapore or something. And I think that having a role model makes it easier for you to visualize what you could be and when you can visualize something, then that’s already half the battle won. So I think that something else we could do for increasing representation of women in science is to kind of more regularly highlight positive female role models. And I feel that I kind of do this too, I regularly engage in this exercise where if I’m feeling uninspired, I actually look up…I have a YouTube playlist where I save interviews with my favorite role models. Male or female, but the point is that both of them are included and whenever I’m feeling in need of inspiration, I just like choose a random video from the playlist and play it. And so I think that has been very helpful for visualizing what I want to be.
🟢 Steven Thomson (39:05): That sounds like a really nice constructive practice that inspiring yourself when you’re feeling down.
🟣 Yihui Quek (39:12): Yes. Thank you.
🟢 Steven Thomson (39:12): And who knows?
🟣 Yihui Quek (39:14): I think it’s a good idea.
🟢 Steven Thomson (39:15): Maybe someone out there will listen to this advice and maybe they’ll put this interview with you on a playlist themselves.
🟣 Yihui Quek (39:20): Oh, I mean, if, if I could help someone, I would be really happy.
🟢 Steven Thomson (39:24): Okay. Well, if our audience wants to learn a little more about you, then where we could find you on social media or anywhere else on the internet?
🟣 Yihui Quek (39:32): Right. So my Twitter handle is @quekpottheories.
🟢 Steven Thomson (39:36): I love your Twitter handle, by the way.
🟣 Yihui Quek (39:37): Thank you. I’m very proud of the pun. I made it up myself. So ‘Quek’ as in my last name and pot and then theories and it’s all one word. And anyway, I think you’re going to link it.
🟢 Steven Thomson (39:49): I will, we will make sure to leave a link to your, your Twitter profile on our website, which is insidequantum.org. We’ll also leave it anywhere that there is a transcript of this podcast. So anyone who listens to this or reads a transcript will be able to track you down online and give you a follow on Twitter. Thank you very much, Dr. Yihui Quek, for your time talking with us today, it’s been an absolute pleasure.
🟣 Yihui Quek (40:12): Yes. It’s been a pleasure for me too. And a great honor as well. This is my first podcast.
🟢 Steven Thomson (40:16): Fantastic.
🟣 Yihui Quek (40:16): I hope it turns out okay.
🟢 Steven Thomson (40:18): I hope so, too. Well, thank you. Thanks also to the Unitary Fund for supporting this podcast. If you’ve enjoyed today’s episode, please consider liking, sharing and subscribing wherever you’d like to listen to your podcast. It really helps us to get our guest stories out to as wide an audience as possible. I hope you’ll join us again for our next episodes. And until then, this has been insideQuantum, I’ve been Dr. Steven Thomson and thank you very much for listening. Goodbye!