Firgun Ventures Podcast

Transcript


Dr Kris Naudts 00:04

Well hello Mete and Zeynep, thank you very much for joining me on the first episode of this new show called Time to Talk Quantum.

Zeynep 00:11

No, thank you so much for inviting us on this first episode, very excited.

Mete 00:13

Thank you very much for having us on the show.

Dr Kris Naudts 0:29  

The question we get an awful lot, or that I get an awful lot, is, what is a quantum computer? How does it work? As I then start explaining, I realise people don't know how a classical computer works, and they asked me about qubits, but they don't know what bits are. So could you enlighten us a bit on these concepts?

Mete Atatüre 0:44  

I think that's a very good point. In the whole communication aspect of quantum physics with the public, I think we start on the wrong foot. We start explaining the guts of the principles of quantum computers or quantum gadgets, whereas public in general don't really know exactly how a semiconductor works, which also requires quantum physics to be able to for us to be able to explain it. So I think it's important to make sure we focus on what it delivers and how it really operates on a very deep technical level, can be a bit arms length from from public, that's not where you understand why you pick up the technology. But in very briefly, quantum computer, or the concept of a qubit, quantum bit, relies on the fact that the peculiar nature, what public considers as a peculiar nature of quantum objects, and that is the fact that they, their state, their property, can be, can include more than one entity at the same time, presence at the same time, just like in the case of bits, bits can only be zeros or ones, binary code, quantum bits can be combination of zeros and ones, where how they combine matters a lot. Is it zero plus one or a zero minus one? There are different values at the end of the day. And the most important thing is to appreciate that it's not just the fact that it can be a combination of zeros and ones, as we dictate what the language computing language is going to be, but the fact that benefit for a quantum computer comes when you have many of these together so that it can benefit from the parallelism they bring. So one qubit, there's often a mistake where people make one qubit is not more powerful than a classical bit. They are roughly the same, equal in terms of what they bring to the table. But 10 quantum bits as opposed to 10 classical bits, there's a big difference in the way it solves certain problems, but not all. So that is where the quantum computer aspect comes in. 

Dr Kris Naudts  2:42  

And people talk a lot, and you touched upon these subjects by explaining, as you just did, superposition, entanglement, spooky action at a distance. Could you say something about spooky action at a distance? Because people seem to be very fascinated by that Einstein quote. I believe it is.

Mete Atatüre 2:56  

Sure, it's a very interesting topic. It started out as a deep philosophical problem that quantum physics portrays. And Einstein, Podolsky and Rosen three, three authors decided that this element is, is an evidence that quantum physics as a theory is not complete and needs more work. So they presented this today. What we're doing is, yes, perhaps this is correct, but it is also very useful the way it is. It does function, it does work, and it can be quite useful. It has not been disproven yet, despite many, many attempts. So we take it as a reality, and what it offers is basically that the description of a physical reality for an object does not need to be localised., does not need to be local. In other words, if I take an entity and meaning multiple objects, quantum objects and spread over some distance, distribute them over some length, scale, there is nothing in the theory that relies on their separation, physical separation. So in principle, nothing in the theory says you cannot take two objects that are correlated with each other, quantum or just electrons spins, whatever, you, nothing in the theory says you cannot take them beyond a certain distance from each other, where they lose their correlations. Correlations exist regardless of distance. So this is the spooky action at a distance, then means what looks local, a single object actually is completely related and correlated to another object, by design, by the way we’ve done previous interactions, for example, that is far, farther away that you think falsely that it is not related to this one. That's where the paradox comes in. It is not a paradox. You should not look at these individual objects. You should look at the whole system, no matter how far they are from each other. That's the concept of quantum computer. 

Dr Kris Naudts  4:49  

Now these qubits, you need to create them. How does that work? Because there's various ways in which you can create them. Could you comment a bit on these various ways, and whether certain ways are better than others? And where is this going?

Mete Atatüre  5:04  

So when I give talks on the topic, I go back and I spend a few minutes just to highlight how important it was in the history of science and development of quantum physics, the first time we experimentally isolated a quantum object. All the communication, if you look at the 1930s discussions, letters to each other, Schrodinger, Heisenberg, all that,they always talk about, the Gedankenexperiment, the thought experiment, because they take it at face value that there's absolutely no way you could isolate a quantum object, an electron, a photon, a proton. These are not things that you find in nature alone. They're always part of a matter, microscopic objects. So the first time this was isolated was actually a trapped ion, a single atom, electron removed, one electron removed, so it's a charged particle, and using electric fields, they were able to localise it to inside a vacuum, and they could see the light that comes out of a single ion. This is profound. It's the first time we actually have a single quantum object that you can see isolated from the rest of the universe. So that turned out to be, the trapped ions turned out to be one of the very nice, promising, quantum bits, the internal degrees of freedom of this isolated object. And then there was this big race of developing other examples where you can more or less isolate it, not, in some cases, not even physically isolate it. For example, electron spins inside semiconductors. Physically they're not isolated. They're in the material. But energetically, conceptually, they're isolated there. They can be treated separately from the rest of the material. So that gives, gave us a big platform of qubit realisations, electron spins and semiconductors, very competitive. Photons themselves, the light, the constituents of light is, has been for a long time, one of the competing qubit candidates. Many companies today are pursuing photon based systems, trapped ions, neutral atoms also are relevant anytime, any physical system you can isolate and have access to the internal degrees of freedom, some internal property, whatever it might be, so that you can classify what a zero and a one is and create coherence between them, superposition between them. You’ve got a qubit, effectively. Superconducting qubits would be the most, I would say in this, this concept would be the farthest one from the original atom picture, because you physically make a device. It's actually an LC circuit in the quantum limit, and it still behaves in the same spirit.

Dr Kris Naudts 7:36  

Okay, now it seems hard to make qubits, well we can make quite a few, but then a whole bunch fall over. Why is that? And does it matter how many qubits we can make? How many, how many are we able to make these days?

Mete Atatüre 7:46  

How many we can make is, it's a loaded question. We can make billions. There'll be companies out there or researchers out there saying, oh, yeah, on a wafer, I can put 1 billion. That is how many you can put. Whether they work, is it a good qubit or not? Is the second question, right? So whenever we talk about qubit number, we should actually talk about the fidelity of the qubit. How good is that qubit, is a quantum object. So we talk more about quantum volume, how many and how good, combination. And it's interestingly, as perhaps expected, the very high quality qubits are hard to realise, so the numbers for those kind of computer realisations, qubit number is limited, fidelity is great. And then other examples, mostly semiconductor, superconductor realm, then,, there we have many more qubits, but their quality is not as good because they're integrated into a material that ends up being noisy in the environment. The key problem of isolation of a quantum object is not just physical isolation from the environment, it's also isolation from any kind of interaction, noise, magnetic field, electric fields, anything that's in the environment that could disturb these guys. These guys are delicate. Many of them not interacting with anyone except with each other is challenge across the board.

Dr Kris Naudts 8:57  

So we have tens, these days, hundreds, thousands?

Mete Atatüre 8:58

In the case of more chip based systems, we’re 1000, 1000 level.

Dr Kris Naudts 9:02

Good ones? 

Mete Atatüre 9:04

Good is relative. Operationally, let's say good enough quantum volume to demonstrate some sort of an advantage. We're at the level of few hundreds, and they're on the road to 1000s. In the case of the other limit, let's say atomic systems. We're on the order of, say, 100, very good ones, so they're in the lower end, but extremely good ones with similar, comparable computational power.

Dr Kris Naudts 9:32  

Okay. So people also talk, this is hardware, people also talk about quantum software, algorithms. People talk about middleware. Can you talk a bit about what that is? What can that be today?

Mete Atatüre 9:47

Sure, of course, I mean, the whole field, of course, started with mostly physicists. I would say, quantum physicists and some mathematicians coming in to the picture, algorithms, a bit of computer science, but only small fraction, and then it developed into this almost industry, emerging, emerging technology, so now I would say it's fairly recent that we realised it's going to take more than physicists to realise all these technologies, which is great. So all these expertise has to come together into one, one activity. So on the software side, I think we're, my personal opinion, we're lagging behind a little bit. We have many hardware options, and we have some clarity on which way we can go forward on each one. On the case of algorithms, I think there's a great room going forward to develop new algorithms. First of all, identifying applications where these algorithms are suitable for. And at the same time, new algorithms that can take a hardware, a physical system, and make it perform better than it is today. So our algorithm approach, typically, for a long time has been this quantum version of the Boolean, right? Not much more than that. So it's a simple exercise to say, what does a classical computer do in the old, original sense? What would be the quantum version? Out of curiosity, but this has been the main workhorse for many years, and now we have others coming to play, surface codes and measurement based systems all that, which is great, but I do think that there are, there are more algorithms out there to be discovered,

Dr Kris Naudts 11:16  

And middleware refers to things like error correction for example?

Mete Atatüre 11:31  

Yeah, exactly so, so, in a way, you have to work very hard on the hardware to make sure your quantum computer sustains this quantumness for sufficiently long time. But then you also have to do something to correct for things, when things go wrong, and error correction is a key problem.

Dr Kris Naudts 11:40  

How does it work? How do you correct for errors? 

Mete  Atatüre 11:52  

Right, so it is, it's also in a similar sense. It is kind of borrowed from the classical version. You do majority vote, similar concept. So instead of information being encoded per bit, you encode it redundantly to three bits, for example. And then if one bit accidentally flips for something, magnetic field or this and that, one bit accidentally flips, you look at the information from the three bits, and then you pick the majority of the two. In case one is flipped, and you say, oh, so with a good probability, is this one that's flipped. Of course it might be wrong, right? The other two might have flipped both, but there's a small fraction for error. And then if you worry about that, then you go to five. So this creates hardware complexity in your system, in return for better performance in the quantum world, now you have to do something similar. So this increases the qubit number again, the same way, and then in the perhaps the more advanced version these surface codes, the, you look at, essentially, you do certain operations between qubits and check for parity. Are they the same, or are they different from each other? Are they parallel, or are they anti parallel to each other, experiment.

Dr Kris Naudts 12:52 

Okay, okay. So when people talk about quantum computing, it's hardware, software algorithms, and they need the middleware to correct for the fragility of the qubits in the main, yeah, okay. 

Mete Atatüre 13:11  

I would say in the whole field right now, it's always growing, and we're at even more accelerated time, so this is fantastic time, but I would say that the bottleneck right now is the quantum error correction, being able to implement it in an efficient way to push things forward. But we're at that limit.

Dr Kris Naudts 13:16  

Okay, sure. The other big bucket of quantum technologies is quantum sensing, which suggests sensors, indeed. Could you talk a bit about how that could help with, say, GPS independent navigation, could help with climate change? Maybe you can comment on a few of those possible applications.

Mete Atatüre 13:43  

So referring to various applications ultimately boils down to making things accurate. So sensors can go both ways, right? You either, you can make things accurate or you make them precise. They're slightly different from each other. The timing, the push for timing, it comes from atomic clocks. One thing that people realised is that instead of having one physical system to do very accurate measurements, say of time or some tempo or something, if you connect them through entanglement, so they're quantum correlated with each other in a distributed system, and if you make a measurement on the collective system, the resolution you get is better. That means you have a better timing option. Then you can define what one second is with better, better precision than we've ever done in the world, in history. So that's the timing aspect of it. And the atomic clock is a timing concept actually started in UK. It was the NPL. The current state of the art right now, I think, is Jun Ye in US, Colorado, and it's really pushing the limits where, given the time, I'd like to spend a bit moment on this, because I find it fascinating. Perhaps we shouldn't define time that accurately, because when you get to the level that we are now, then time becomes a problem. Because space time coupling, due to Einstein's relativity, starts to come in to play. So when you say what time, Jun Ye will say, where, here or 10 centimeters above, because it does matter under gravitation field, which is fantastic that we’re at level of precision to be able to see experimentally changes, gravity changes. There are applications of being able to sense gravitational fields at the very, very fine precision. You can literally load up a an instrument, gravity sensing instrument on a truck, and you can drive through an area and identify gaps, holes…

Dr Kris Naudts 15:42  

Groundwater levels.

Mete Atatüre 15:46  

Exactly, works underneath without if you don't have any documentation, which is London, for example, the UK quantum technologies program focused a lot on this aspect, and there's some great demonstrations. Now on the sensor side, there's a different, bit more loosely defined. And there, what we're trying to do is the, almost the next step from identifying a qubit. Quantum sensor is also qubit, but a qubit you really isolate from the rest of the world, the rest of the universe. A quantum sensor is you create a qubit, you isolate it from the rest of the universe, and then you pick something and say, go and interact with that one and tell me what you see, right? A sensor interacts and then observes right. In the quantum world, there's, there's not really a concept of observing without affecting, right? This is, you do a collapse. So a quantum sensor operates by actually interacting with the object, with the entity, or trying to learn more about and then, and then you make a measurement on one and to understand what's happening on the other side. So it's very interesting from a science perspective as well, but it does have applications very broadly, anywhere from sensors, as we mentioned, gravity, all the way to life sciences.

Dr Kris Naudts 17:02  

Yeah, correct. So maybe you can talk a bit about that as well, because there's, there's various centers in the world. You're co director of one of them, Q Biomed Hub here in London and Cambridge. There's similar initiatives in the States. There's initiatives like that in Australia. So this is, indeed, what early diagnosis, molecular signatures. Could you talk a bit about that?

Mete Atatüre 17:19  

Sure, it leverages. I should start from why you want to invest into this. Why you would like to make sure that we put efforts, expertise and efforts into this area, one because, if it is successful, it really has a big impact. So if you can help healthcare system in the world with new, novel technology that you develop. Yes, it's worth it. So it's definitely worth looking into. The second thing is, quantum objects tend to be small. Small means integratable. Sometimes it's not the first thing you think of when you talk about an isolated qubit, usually in various publications, public, science publications. You see a dilution refrigerator, essentially a cold price that photo is a quantum computer. Quantum computer itself, rather, is a chip, of course, not the whole cooling unit.

Dr Kris Naudts 18:14  

Yeah the big channel area.

Mete Atatüre 18:17  

The impressive big unit, is just the cooling unit, of course. So you think that if you have such a system, it is hard to incorporate into more ambient conditions, like healthcare. This could be macroscopic ambient, human. You need to interface with a human. Or it could be microscopic ambient, that is say, inside of a cell, you can observe what's happening in the microscopic world without downsides of dealing with more conventional sensors. So in both cases, there are some physical systems. We're lucky, and we're determined to add more to the list, but there are some physical systems that sustain quantum coherence at room temperature, under border, ambient environments, messy environments. So they are the unique ones we push for, particularly to see if we can capture brain signals from outside without having to have a huge NMR machine attached to your head. This is especially important when the patient isn't fully aware of why they're doing this 

Dr Kris Naudts 19:05

A child.

Mete Atatüre 19:07

Yes, exactly. So a cute little helmet that they can wear with a little Viking horns so that you can identify the brain images and identify how recovery, for example, from a trauma, is proceeding, as opposed to an adult who can commit to a large scale, some discomfort for finite time. And also, it's important to be able to do this in a very fast, quick turnaround, so that you don't need these very expensive centers to be able to book a certain time to get, so you don't have to wait three months to get your service. You can do it in a week. A GP should be able to do it. It's a fairly flexible system, and it does depend on optically pumped magnetometers. Again, an atomic cloud that reacts to things that are happening in the environment.

Dr Kris Naudts 20:01  

Do you mean like quantum MRI scans?

Mete Atatüre 20:05  

Exactly, exactly, so if the environment is powerful enough to disturb the quantum state of this object that you're dealing with, then this can become the sensor, if you know what that environment is, and that's exactly what we do. We put an interface to the brain or spine, and we can make those measurements, or we put again, these ambient, operational quantum sensors. Diamond turns out to be very interesting material that hosts quantum objects, atomistic defects, and they're optically active. So you can see them literally with your detectors. So having them inside cells, you can measure ion concentration, local temperature, magnetic field, electric field in the cell, and also its viscosity of the environment, all simultaneously, more or less. So these are great advantages to support bio research.

Dr Kris Naudts 20:53  

Yeah. If we go back a bit to quantum computing, one of the possible applications that people get very excited about, including myself, is drug discovery. Novo Nordisk also takes big efforts to get there. The cluster run Basel with Merk is also working on that, I believe, or Roche. 

Mete Atature 20:57

Absolutely.

Dr Kris Naudts 20:58

What can it do, though, that AI can't do? What does, what does quantum going to deliver there? 

Mete Atatüre 21:19  

That, that's a very interesting language. I think I will refrain from answering exactly that question. I would like to turn around to say, AI and quantum can do better, because it's, it's been perceived in the past as quantum being an alternative to more conventional ways of computing. You will be, you will be delusional to throw away years and years, decades of fantastic research that that got us to this point, and particularly for AI, we're in a very exciting time of finding out patterns that we would not be able to identify, or having algorithms that are very good at identifying the solution than humans themselves. So, so what, what we would propose is that it's a combination quantum computers give compute power. AI needs a lot of compute power, and so I would certainly leverage all the developments of AI and classical computation interfaces hardware, but then have some of the parts that require the critical computing power to be QPUs. So GPUs are there and QPUs are there,

Dr Kris Naudts 22:22  

But they are indeed two somewhat separate tribes. Quantum is a tribe of physicists and AI is a tribe of computer scientists. So it sounds whereas indeed we both need each other, I suppose.

Mete Atatüre 22:34  

Exactly. And it makes perfect sense, because if you look at the conventional computation, compute, it really started from one physical system, silicon. No one doubted it. At some point, there were a couple of other elements that popped up, and then it kind of disappeared. I think we all agree that silicon is the main standard, gold standard. So while hardware improvements are continuously happening, it is never deviating or there's no doubt as to what kind of a physical hardware we're dealing with big improvements over the especially on the last 10 years, that's for sure. And now we're looking into even more, bigger jumps, but it wasn't the case for the last 30 years. So then you can really focus on the software part, if you trust that the hardware part roadmap is clear, for quantum hardware part, the roadmap is not clear, right? We have still many physical systems that are competing friendly with each other, and there might be algorithms that you can tailor for one better than the other. So there is a lot of exploratory work, horizon scanning still going on. So it's perfectly normal for AI to first stick to silicon based conventional systems, because that wasn't the bottleneck till now, but now we do have a limit, and even the conventional systems are looking at alternative arrangements defining the classical bit. So this is the time to start merging, and maybe companies have started things.

Dr Kris Naudts 23:58  

Yes, this is not to do with life sciences, per se, but this whole debate around AGI, and there's a whole fear mongering debate around that as well. If quantum meets AGI, what's going to happen?

Mete Atatüre 24:11  

In all of the developments that we ever have in the whole history of humanity, you never stop science. You never stop progress, because it might be dangerous. What you do is you do is you do your science and you do your technology, and then you put on your citizen hat on right, you take off the lab coat, and you put your citizen hat on. And then you work with the policy people. You work with industry. You work with all the stakeholders and the public, information to the public, to make sure it's used for the right purpose, for the benefit of humanity and the world.

Dr Kris Naudts 24:32  

Now you also have undoubtedly seen, like everybody else, these vast sums that get spent on data centers to indeed power the AI work. Now, Elon Musk is talking a lot about building data centers in space. 

Mete Atatutre 24:44

Because it’s cold to begin with.

Dr Kris Naudts 24:46

Right, exactly. Does quantum need to go into space? Is that a workstream? 

Mete Atatüre 25:02  

Technically, quantum is already in space. There are cold atoms in space that show quantum coherence over periods of time. People are considering, from a very fundamental research side, they're considering creating these spooky action at a distance experiments, but over very large distances to really check general relativity and limits of quantum physics in terms of proper distances, probably where the next physics lies, combining quantum and gravity, which is something that hasn't been unified yet, the only un-unified thing, field. So that's the fundamental science and knowledge generation part of it on the application side. However, yes, there is quantum communication, secure communication, distributed quantum computing, blind quantum computing over very large distances, all can be done through satellite links that have to be operating at a secure, that security level. We cannot simply rely on trusted nodes, that someone is there to protect your information for you, and accessing quantum computers, we have this concept of cloud access in general, given how expensive they are and they're not that many around. So usually we have these cloud access concepts. But if I have a question that is extremely expensive, even the question itself is extremely expensive, let's say profitable, then I might worry about asking that question at the risk of the question itself being revealed. So for example, this could be a pharmaceutical company or some others, then you need to make sure that that communication line itself is quantum secure as well. So that's why satellites could come in to extend the radius.

Dr Kris Naudts  26:47  

Okay, so we are in this room all very excited about what quantum can do for good and all these applications and science and etc, and there's so many fields It will revolutionise, we hope, and we expect. Now, a lot of the coverage that you read in the media is about fear. And one of the key fears that's being also rolled out time and again is the Q day where everything will be decrypted, and that we need to all be quantum ready, or companies need to be quantum ready, or all the sensitive data are going to be stolen, some of them already harvested now, I believe, is what they say, decrypt later. Could you comment on that?

Mete  Atatüre 27:20  

Well, if we, if we reach that level of capability for quantum computers, yes, that is a problem. That is something to worry about. I guess. I don't think individuals have to worry about it. I don't think citizens have to worry about it. I think it's more a challenge for probably governments and perhaps large corporations.

Dr Kris Naudts 27:39  

Should they get their skates on to do this? Because we need a fair amount of qubits in order to for this reality to happen.

Mete Atatüre 27:46  

So I would say, based on what we know, and there are many algorithms of problems to solve when it comes to applications, I don't think we're anywhere near the full range of applications, quantum computer might have an advantage. At the moment, the hardest one is Shor's algorithm. So hardest one is to decrypt this, these very text, let's say, communications. So, so I think that is still some years away, unless we solve proper quantum error correction, feasibly. Application wise, feasibly, this is not going to be an issue. So I, I suspect very fast progress, and I'm sure me and many others I’m sure will be wrong given the pace of the progress, but I suspect that there will be a lot of progress in the next, say, five years, about performance of quantum computers and the number of quantum computers out there, and how accessible we can make them for public but how powerful can we make them on their own? Can they actually break codes at that level? Probably not.

Dr Kris Naudts 28:55  

Because it seems to be so that things like drug discovery we’re sensing for sure, but also drug discovery is going to be much, much, much earlier than decryption, much earlier, which is the good news.

Mete Atatüre 29:05  

There is already a lot of effort is put in at that level. So when we say industry stakeholders engagement, that's actually this, one of the areas, quantum chemistry, can we actually simulate better? It's almost like think of it as being able to having the power to be able to ask the reverse question, instead of trying out different, this is what AI kind of does, trying out different combinations to see if it is useful, and being able to sift through many of these very efficiently. You can, you get to ask the reverse question, what would be the combination that allows you to solve this problem? So that's that's quite powerful, if we can get to that.

Dr Kris Naudts  29:42  

What role would a quantum computer play in, say, war or conflict?

Mete Atatüre  29:48  

It would play exactly the same role as any other technology we have available. There's a concept called dual use. Anything you do can be used for other purposes. Let's say. I'll just leave it at that. So. Again, it's policy, engaging with people, getting people on the table, around the table, as early as you can, so that you develop the right mechanisms to put in place so that no technology is harmful, or, say, defense or others are justified and it's highly maintained.

Dr Kris Naudts 30:20

But is this another Manhattan Project in the making? 

Mete Atatüre  30:23

No, in Manhattan Project. So people do refer to quantum computing as a Manhattan Project, given the scale, but the goal of the Manhattan Project was very clear, this is not what quantum technologies..

Dr Kris Naudts 30:34

And it won't be, and it can't be?

Mete Atatüre 30:37

It can be. We should never lose sight of the fact that any new technology can be used for both purposes. It is what you do with it. Ultimately we decide, not the technology. So it's never the technology's fault,

Dr Kris Naudts 30:49

Okay, what about the quantum internet? What is that? And do we need that?

Mete Atatüre 30:53

So quantum internet is exactly the name suggests. It's the, inspired by internet. That is the original internet, I should say computers connected to each other, right? Well, this, now today, is a very common, obvious, obvious thing, but you have to go back to the first time this was actually realised physically in CERN, and no one had connected computers together. You would have a computer, and you would run your code. It would be a compute machine. It will give you, give you an answer. If this answer has to go to another location, it's the human that has to go over there and say what the answer is, right. So when they actually connected multiple computers together so that one computer accessed information in the other computer, that was the breakthrough. That was the first demonstration of the concept of Internet. Today, the whole world is connected. It's amazing to see this progress, but we shouldn’t take it for granted. There was a time where it was only three computers linked together with wires. That was the Internet. The Quantum internet is in the same spirit. Can we make those computers Quantum? Can they be quantum objects? They can perform, compute, and still communicate with each other without losing their quantumness. So without using that human interface to say, to carry the information, create a quantum object carrying a quantum information, and then be able to transfer part of that information to the other side. So in other words, distribute spooky action at a distance and then be able to do operations more effectively, efficiently or in different locations. So its advantage is that we will be able to do distributed quantum computing. This I see is the biggest impact. That means you don't have to rely on one chip having all the power for quantum computer you need. You can connect many of these chips together, to work together.

Dr Kris Naudts 32:30

Do we ever have a quantum computer in our pocket? Do we have a quantum phone?

Mete Atatüre 32:36

Do you have a high PC system in your phone?

Dr Kris Naudts 32:38

I do not.

Mete Atatüre 32:39

Right. Do you need one? 

Dr Kris Naudts 32:40

I don't think so.

Mete Atatüre 32:42

Good, okay, so you don't need it. So the hardware at the moment certainly limits you from doing that. If you can make it smaller, perhaps that will be also be possible. Industry doesn't always follow technologically sane actions. We made things very small and then decided it's too small, and then we made things large again. If you look at the typical cell phone, it started huge among us, the 1980s with the antenna. And then it got smaller, smaller, smaller, and it got too small. And then we, the screen was tiny, and then we made it big again. So it depends on what… 

Dr Kris Naudts 33:16

Like hair.

Mete Atatüre 33:18

Exactly. So it depends on what public wants to a certain extent, or what is technically doable. But I think quantum computers will remain as problem solvers for bigger questions, rather than individual…but of course, you you may not even know at some point when you access some cloud computation for you playing a game, computer game at home, you are accessing a of course, a center, yeah, or when you access or when you Google something, you're accessing another center, right? Some of that might be a quantum computer, and you wouldn't know, and it's okay. You don't need to.

Dr Kris Naudts 34:10

You touch upon a good point. Quantum and gaming, quantum and art, I mean, quantum and music. What is happening there? Because you don't need that many qubits again, for doing meaningful stuff.

Mete Atatüre  34:20

Oh, absolutely, absolutely. There are, there are already some actions in there. It's still a ,there's still a barrier. In Cambridge, we have a strategic research initiative called QAMSS, quantum and advanced materials for sustainable society. It includes 16 departments in the within the university, plus non departmental entities like Kettle’s Yard, Kettle’s Yard is an art gallery, and they do a great job in promoting art in Cambridge. So we engage with them, because we think most obviously from the advanced material side, if you have new materials, art will pick it up. If you have a particular way art develops a technique, science will pick it up. Art picks up new materials, but science picks up lithography and starts doing it in a nano limit, which allows us to have the chips that we have. So communication goes both ways in a very inspirational manner. So it would be a shame if we didn't do that interface with quantum and see what happens. So we're at that stage now where we're seeing some examples and they’re very exciting. So, yeah, keep your eyes on it, there will be things coming up.

Dr Kris Naudts  35:19

So when a new technology comes around, banks are often first in the queue to try it out and to make a difference. What is going on in the world of finance and quantum?

Zeynep Koruturk 35:27

So I think finance has actually really been at the forefront of some of these developments in quantum and and I know that there's been a few of these large global international investment banks that have been looking at it for the last 10 years. Some have actually built huge research capabilities as well, and are really driving the the innovation on that side. But I think, I mean, there's all these optimisation problems, all these Monte Carlo simulations that can be really, really affected by quantum but from an investment point of view, if you're looking at it, the the entities in finance that are a bit more advanced are the hedge funds, and I think because they're really thinking about potentially getting an edge from a trading perspective. And there, I think the type of additional investments that they can make and access to some of these companies and potential algorithms that can be developed by some of these companies make a meaningful difference, and they're very lean. So from our perspective, I think that what we've seen is that, yeah, there are a number of large banks that have been at the forefront and that are becoming very big centers of quantum in finance, but the ones that are truly at the edge of it is probably the hedge funds. They've made quite a few investments in a number of companies as well.

Dr Kris Naudts 36:33

Okay, so if Mete wanted to invest in quantum stocks, where should he look?

Zeynep Koruturk 36:39

That's a very good question, because we do get that a lot. And I think there's been obviously the performance of some of these quantum stocks over the last year really, really brought quantum at the forefront, and it's been a dinner conversation topic for a lot of people. I mean, Kris and I have been invested in this space for the last almost 10 years now, and I think initially, when we were making these investments, a) there were very few companies that you could invest in. They were all private, roughly, all private and, and when we were saying to people that we made quantum investments, there was just no response. And, and that has changed a lot.

Mete Atatüre 37:08

What does it even mean?

Zeynep Koruturk 37:08

Yeah!

Dr Kris Naudts 37:10

Or pity!

Mete Atatüre 37:11

The investment was maybe here or there.

Zeynep Koruturk 37:14

They didn't want to engage with that much. They're like, Oh, that's good, and moving on. But I think that has changed so meaningfully in the last year, and partly, I think that has been driven by some of these public companies. Obviously, there's been a number of companies that have SPACed over the over the last few years, and the performance of these stocks weren't that meaningful if you look at the last five years, but in the last year, they have really, really been very, very positively impacted by the broader tech environment. And we've seen some stocks going from like three, 4 billion cap value to about 25 plus. And that, of course, brings it home for many people. And people have started to think about, is this the next AI, should we start investing? And the reality is that at this point in time for a lot of people, the only way to access it is through public markets. And in the public markets, there are still very few companies that are operating in quantum there are a number of larger tech companies, of course, like the big tech companies, have their quantum units, so there have been ETFs also coming up, incorporating some of these full stack quantum companies and and some of these big tech companies.

Dr Kris Naudts 38:18

What are some examples of such ETFs?

Zeynep Koruturk 38:20

Well, there was a Blackrock ETF, for example, that has recently come up that is following a quantum index. But I mean, I think it's a nice idea, but the reality is that 99.99% of quantum companies remain private. So today, I think the, this will change over the next few years, but today, the only way to access them is by making private investments into some of these companies. Some of these companies, and that at this point in time is quite hard for for individuals. But the space will change. Quite a few of these companies are on a pipeline to go public, either through SPACs or through normal IPOs, so I think it will be in a very different scenario in the next five years. 

Dr Kris Naudts  38:54

If you compare to AI. I mean, if today, a person of some standing founds in AI startup, it's worth a billion the next week, is that the same in quantum are these early stage startups extremely overvalued, like some of the public companies seem to be? 

Zeynep Koruturk 39:09

No, and actually, I mean, it's been, it was a consideration for us also, when we started investing in the space we were, I mean, we were wondering whether or not some of these private companies would would benchmark themselves to some of the public companies, whether they're in AI or in quantum and, and I think that has not been the case. I think the valuation on the private side has has remained reasonable and, and therefore I think we're quite comfortable leading with those types of valuations. But, but there have been a lot of big statements, including from, from like the founder, the CEO of alphabet, saying that where he sees quantum is where AI was five years ago. And that can give you a sense of where these might, valuations might go over the next few years, especially as the technical progress continues and the timeline shrink towards delivering proper commercial value. So we do think that is a very, very very interesting time to be involved in the space and either as an academic or as an investor. 

Mete Atatüre 40:03

I should say, if you go back 10 years, it would be an odd thing for an academic to get involved in a spin out and actually pursue it diligently. And usually academics are not, not the greatest when it comes to both being good as an academic and as a company person and as a fundraiser all combined. But now the ecosystem, the quantum ecosystem, understands how to do this, how to bring the right people together so that you give up the control, you spin it out, and it's out there with the right people running it, so it's a more mature approach. The timescales are shorter. 

Dr Kris Naudts 40:41

Yeah

Zeynep Koruturk 40:42

Yeah I think one point that you mentioned Mete earlier was also getting a broader range of people involved in quantum because I think that's the other thing that a lot of these companies are now realising. I think it just requires potentially a different group of a different skill set to bring a company out of a lab, versus bringing this company public. And we have noticed a lot of them really maturing in their way of thinking around this as well, and bringing the right potential C levels or potential board members that can help in that path. And the conversations we're having with companies today is very, very different than the ones we were having with companies three, four years ago

Dr Kris Naudts 41:19

From lab, from lab to market, has clearly happened from markets to public companies, hasn't happened just yet, but would be happening soon enough, you would imagine. Now, as you already mentioned, I mean, big tech in the US obviously takes the lead, probably in terms of amounts invested in quantum and pushing the boundaries of technology as well. But there is a great group of startups, not, not, not 1000s, about 500 or so, according to the Tony Blair Institute. Could you describe a little bit how that maps across the world, geographically?

Zeynep Koruturk 41:42

Yeah. So maybe the first thing to mention, and we have discussed it jointly in the past, is that, I mean, the number, the exact number of quantum startups, is like different people point out different numbers for this, but I think if we take Tony Blair Institute's result, which was, I think, 516 so let's say 500. We know that about 150 of them are in the US and and, and the great news for the UK is that about 64 of them are in the UK. So if you're looking at a number of like, number of startups per country, UK is second in the world. I think we are disregarding China. So I think China might distort this picture a bit. But if we're looking excluding China. And then there's about…

Dr Kris Naudts 42:21

Is this Brexit at work or Oxford, Cambridge at work?

Zeynep Koruturk  42:21

I think it's a longer term thing than Brexit. I think the caliber of UK academic institutions is just so high, whether you take Oxford, Cambridge or London, that I think it's not surprising that it's really a center of excellence from a quantum perspective as well. And then you have other hubs in Europe, which makes the other 1/3 of this group, which are also quite remarkable. I mean, I think France and Germany are not that surprising for many people, Holland. I mean, Kris, we discuss it sometimes. I think Holland Delft is an exceptionally good ecosystem. And I think there have been certain countries that have just been at the forefront of this versus others, and that have put meaningful investments behind it.

Dr Kris Naudts 43:01

For sometimes seemingly arbitrary reasons. Denmark, for example, okay, Niels Bohr Institute historically, but then Novo Nordisk leading the charge.

Mete Atatüre 43:08

Well, you need people who would pioneer it. Right? Yes, people to champion a topic. Really spend time and effort to go talk to people, policymakers, people who have money, to say this is going to happen, you have to spend that effort. I would really highlight, for the UK, we are punching above our weight when you say per capita. I would really highlight the success of the quantum technology program in creating the ecosystem so people are even aware of how to how to create an industry.

Dr Kris Naudts 43:33

Yeah, could the next big quantum company come from Paraguay? Is that possible?

Mete Atatüre 43:40

I think it's always possible. It's, but there is a reason why there is a Silicon Valley. It's not that, you know, Silicon Valley offers a better infrastructure or cheap design than anywhere else. It is an environment. You have to feel like this is the environment. You see how progress is being done. I think that's why ecosystem is the key word there. If you don't have the ecosystem, you're isolated and you're trying to reinvent the wheels at every step.

Dr Kris Naudts 44:06

But you need a deep, academic ecosystem to bring up these startups. Otherwise it can't happen. Hence, Paraguay is probably unlikely. 

Mete Atature 44:15

Probably unlikely.

Dr Kris Naudts 44:15

Whereas a strong AI startup could easily come out of Paraguay. 

Mete Atature 44:17

Sure.

Dr Kris Naudts 44:18

I think it's a different, different, different proposition, one that poses questions whether indeed, one therefore, per definition, creates a two or multi speed world within Europe, even from what we've seen, some countries are miles ahead, others are way behind, and how the ones that are way behind catch up, it's not so obvious.

Zeynep Koruturk 44:37

I mean at the European level, at least you have a European Union. If you are indeed a Paraguay, because that was the example given, that's even harder, because you might be properly left behind, but, but I think the additional thing to mention is that in the last few years, especially, there's just been a lot of attention from different countries and their policy programs to to kind of try to not be left behind. And we know that there are quite a few countries that are a) putting resources for academic, labs, universities, but also the ecosystem buildup, and we are in conversations with a number of them. And I think the future might not be that bleak, in the sense that there might be a few more Centers of Excellence outside of these big centers that we have mentioned, but, but that means that they have to take action immediately and put proper resources behind it.

Dr Kris Naudts 45:23

Okay, but often it's the usual suspects, Australia, Canada, UK, US, Germany, France. They are the usual suspects that generate the most...

Mete Atatüre 45:30

There is investment, you have to put in the investment to create the environment where people can find out that there is a problem that they don't even know yet to solve, that it's going to be a roadblock, so that they can then find out through interaction and then figure out a way to solve it. Solving is the second step. First is to identify what will be the roadblock if you don't have that support around you, particularly for quantum I call it the non quantum support. If I don't have the non quantum support, I won't be able to make that leap anyways, so this is where your being isolated actually matters.

Zeynep Koruturk 46:03

We were talking about this because it's a topic that both Kris and I actually are quite interested in. And I mean, technically, talent is equally distributed like, distributed across the world, in the sense that you can have someone coming from a country that does not have these technologies. If he's given the right resources, the right visa to migrate to another country and actually build there, you can still create some kind of ecosystem, but that so far, has not been translated in the policies, it seems.

Dr Kris Naudts 46:29

From your vantage point, Professor Atatüre, do you follow developments in China, Russia? Do we look away from it? Is it accessible?

Mete Atatüre 46:39

Of course. I mean, science is global. Science translation to technology is also global, especially now, more so than before. How easy is it to get to the information is, however, local, so it's the developments in some parts of the world, you have easier access than other parts. But nevertheless, you have to keep your eyes open, because we as scientists or researchers have to appreciate that we're all in the same, same boat, right? This is we're all doing research. Any discovery in New Zealand is immediately mined here in the UK, and vice versa, right? So that's the knowledge generation aspect of it. We should make sure that technology development or industry development, particularly around sensitive, sensitive areas as well, national security, whatnot, it should still not get in the way of the academic exchange, the global nature of academic world, or research and knowledge generation. So that's a key part. But it is, there is a risk. There is a risk where borders, boundaries limit communication, as it did in the Cold War era and previously.

Dr Kris Naudts 47:47

Exactly, if you were in charge of the CIA or MI6, what would you, what would you do, with your expertise?

Mete Atatüre 47:55

To address which question? I probably will be busy with everything else!

Dr Kris Naudts 48:01

How would you anticipate the coming of the quantum era? How and where should an intelligence service place its bets?

Mete  Atatüre 48:09

I wouldn’t do anything that is different than what is being done around the world in countries. As far as I'm aware, usually you're only aware up to a certain point. Again, it goes back to the concept of dual use. I need to know what it is, and I need to have it. I don't even know what it is, but I need to know what it is and I need to have it. So US was a very good example for this. They had the National Security Agency was key player in developing the very early the onset of quantum technologies, the concept of quantum computing, feasibility tests, likelihood of coming forward and funding some of the progress. It's not to to protect yourself against it. It is to make sure you have it. Someone else can have it too, but you want to make sure you have it as well, which is good technically. It does push forward, and it's it's down to the policy to make sure that it’s…

Dr Kris Naudts 48:55

But, but not every country can have it because they wouldn't have the resource.

Mete Atatüre 48:59

That’s right, and that's that's something we need to address collectively as a world.

Dr Kris Naudts 49:05

Is that time now that we need to start addressing that? Is this, is this UN territory?

Mete Atatüre 49:10

I think, given how fast the progress is, if you look at humanity, there's like hundreds and hundreds of years between this guy gives us as an example of them, between the discovery of glass and paper. You know, to all these things that you would say roughly around the time we start saying we're civilised, you should have both. That's hundreds and hundreds of years gap between them. Yes. And then, you know, the first man flight versus going to the moon is, what, 60 some years, right? So this timescale is shrinking and for AI, the big breakthrough is probably every other week or something. I mean, this is….

Dr Kris Naudts 49:43

It used to be every other generation, but now it's indeed multiple ones per generation.

Mete Atatüre 49:46

Exactly everything happens within one generation. So the speed is so fast that policy engagement, or thinking, trying to think ahead, is almost a challenge. So you have to have people around the table all the time.

Dr Kris Naudts  49:59

Once again, Mete Atatura and Zeynep Korturk, thank you very much for coming on the show. It was a great conversation.

Zeynep Koruturk 50:03

Likewise, thank you so much. I really enjoyed the conversation

Mete Atatüre  50:06

Thank you again for the invitation. I really enjoyed it as well. 

EPISODE ENDS