Hello, and welcome to the Physics World weekly podcast. I'm Hamish Johnston. Today, many creative industries are reliant on cutting edge digital technologies, so it's not surprising that this sector could easily become an early adopter of quantum computing. In this episode, I'm in conversation with James Wootton, who is chief scientific officer at Moth Quantum.
Based in The UK and Switzerland, the company is developing quantum software tools for the creative industries, focusing on artists, musicians, and game developers. James joined Moth in September 2024 after working on quantum error correction at IBM. He also has a long standing interest in quantum gaming and in creating tools that make quantum computing more accessible. He joins me down the line from Basil.
That interview is coming up. But first, a message from IOP Publishing, which brings you Physics World. As part of IOP Publishing's ongoing support for the early career researcher community, it's created a comprehensive guide to assist researchers in publishing their work. This guide includes valuable information on funding, peer review, ethics, accessibility, and strategies for maximizing the visibility of your work post publication.
Search for IOP science researcher publishing guide, and sign up to receive your copy, which will be emailed directly to your inbox. Hi, James. Welcome to the podcast. Hi. Great to be here. So, James, can you give us a brief in introduction to Moth? What does it produce, and how many employees does it have? Yes. So Moth is a quantum computing startup, and there are many quantum computing startups. Some of them focus on hardware. We're not one of those. We're one of the quantum software startups.
Often, quantum software startups will focus on a particular industry. You have some looking at health care. You have some looking at finance. But we are the only one who focus on the creative industries, so things like games and, and art and music. So what we produce is is quantum software. We produce, software for people products for people to be able to use quantum computing in their creative practices.
That could be from no code interfaces where people just have to twiddle some buttons and some magical quantum things will come out. Or it could be to, like, higher level programming languages that people could use to create their own, quantum software. As for how many employees we have, I don't know exactly. I'd say 15 to 20 is a sort of level where that we are at.
I see. And and are those people mostly sort of quantum people, physicists, engineers, or or do you also have, I don't know, for for lack of a better word, creative people, musicians, artists, etcetera, on your staff? Yeah. Well, as a small company, it's good, that if people can do multiple things at once. So, we would like to have people who are quantum people and creative people.
But, yeah, we have a research division and that's based, here in Basel in Switzerland, where people come from a a background in quantum, but they also bring expertise from various creative areas as well, either with a long standing passion for music or someone who, has has made their own quantum educational board games. We have team members, covering a full spectrum here. We also have another office in London.
And in that office, we have more the the engineers, and those engineers are are usually from creative backgrounds. So we have one engineer who's just finished her master's degree, at, the University of Arts London, but is, deep in the code of of quantum computing as well.
I see. So, James, before we we sort of focus in on some of the projects that you're working on at Moth, Can you give us a a broad stroke description of why quantum computing technologies can be useful for creating music, video games, visual art? Yeah. Sure. I think this is probably a question people asked at the beginning of computing as a whole, why people might wanna apply it to creativity. But, actually, games and music were some of the first things that people experimented with.
Because the creative space is such a a large one, there are so many computational problems, that one could try to tackle. And so quantum computers coming and bringing their unique ability to solve certain kinds of problems, it's inevitable that we'll find problems within the creative space that, are relevant for that. But also, quantum has also always been something that is inspirational for people.
And so there's also the aspect that it's quantum as inspiration, as something that people can can use to get sort of new insights. And so this is actually one thing that's really exciting about working in this area is that you're not searching for a specific right answer as you are when you are doing quantum chemistry and trying to calculate an energy or doing, factoring and trying to calculate factors.
There is a realm before that in which it's just exploring the new technology and what it can do and finding unique insights from that. And procedural generation for games, That's one of the targets, that you have for quantum computing. What is procedural generation, and how could it benefit from a quantum computing approach? Yeah. So procedural generation is the algorithmic generation of content for things like games.
Although it could also include things for other areas, as well, such as in CGI, just, generating textures is used to make, things less plasticky sometimes. But, I think of it mostly as its application in games, and there is generating content, which could be, levels. It could be story lines. It could be, characters. It could be all kinds of things.
And, this differs from the current generative AI where you just, you know, you throw a million pictures of of cats and dogs into, an AI technology and it will eventually learn what a cat is, And it might then be able to generate new cats. But in procedural generation, it's more handmade. It's more tailored. You have to understand what it is you are generating. You have to know the rules.
And then the task is to try and generate something as new and as unique, within those constraints, as you can, while satisfying those rules. And so, essentially, it's it's a problem as constraint satisfiability. You have certain constraints, and you you want to find something that satisfies them while exploring the possibility space as much as possible. And this is is very hard because, constraint satisfiability problems are hard to solve. So people do it in different ways.
But because of that, they sometimes roll run into what's called the 10,000 bowls of oatmeal problem, where you have a generator that is technically generating something very unique every time, but just not doing it in a way that people perceptually care about. Just as every bowl of porridge is unique, but no one cares about the difference between today's and yesterday's. It's all the same.
So having ways to really drive at the constraints of the liability problems and the optimization problems, well, are now new tools to help people who are building these generative systems. I see. And, you know, the fact that quantum mechanics is all about, probability and and correlations,
does that does that sort of help? I mean, I I think I might be getting a bit sort of quantum woo y here, but I'm sort of thinking of, you know, the multiverse in a game, and, you know, allowing quantum states to evolve. Am I sort of out to lunch on that or is there something is there something there? Yeah. There was, an app a few actually probably about ten years ago called Universe Splitter, which, took, photons from Eddy Quantic, and measured them in one or the other,
basis. And and you use that as a as a decision maker. So you would put in, shall I have pizza tonight or not? And then you would get the universe splitter to do it. And it's quite, like, quite the most narrative there that in theory, if you believe in the in the many world interpretation, then in the other universe, you're having pizza, but in this universe, you're not. So there's there's things that people could be
inspired by in this area. But for me, the probabilities are more of a nuisance than they are a help because I think people often think of procedure generation as something very random even though you have to put a lot of effort into into constraining it. And you often do it in a seeded way. So you with a definite seed, you produce a definite outcome. So the randomness of quantum is not necessarily
a good thing. I much prefer when running circuits to be calculating expectation values and therefore be to be getting the nice, solid, predictable things rather than, relying on the randomness. That's not really what we wanna do. I see. Okay. And, using artificial intelligence for generating characters in games is another quantum application that, that the company's interested in. Why would quantum AI be better than conventional AI for character generation?
And and maybe you can start, by I mean, could you give us a taste of what what does character generation mean in a in a game? Yeah. Well, it's essentially, one can think of it as another part of procedural generation, and indeed people often think of procedural generation as a branch of AI. So it's just a continuation in some ways of what I said before. For characters, it could be from things like just generating what the character looks like.
But, maybe more pertinent for this question is is broader things like the, the storyline of a game. So you might be generating a game and that game has a narrative. And, and then you can procedurally generate that narrative. And for that, you need to have some way of encoding the narrative. Like with any procedure generation, you need to know what you're generating to, encode everything in the right way in your generator.
And so one can ask the question, what kinds of information are more naturally encoded in quantum systems than in classical systems in bits? And, of course, we know completely obvious that quantum states are more naturally encoded in quantum systems. You know, it's much easier to make a superposition than it is to write down all of the coefficients in binary. But that doesn't mean it's just restricted to,
quantum physics. There's also, evidence that other kinds of information are more naturally encoded in quantum systems as well. And one that, some of our, some people at, Continuum have been looking at is natural language processing, particularly using this framework called Discocat.
So this is a way of showing that, the kinds of mathematical structures that are good for encoding quantum states are also good for encoding the kinds of structures, the grammatical structures that you get in in natural language. And grammar is something that also, exists beyond language. So you can use ideas from grammar to kind of if you think of the the narrative of a game as essentially a very large sentence in which all other of sentences in the game were just
subsets on. That that structure can also be expressed grammatically and other things in procedural gen generation. So the idea of grammars is something that even in the classical world is is very important in procedural generation. And so, we can help port some of those insights from, things like natural language processing over to procedure generation and also music.
I see. And and and so is the is the idea here that using a quantum system, it will give you a computational advantage in the sense that you'll be able to, you know, compute, the things that you need for a game, much more quickly than a conventional com computer or even a conventional computer couldn't make those
calculations? Or or is it more of a, sort of, a quality thing in the sense that, the stuff that you can get out of a quantum calculation is more realistic or more desirable to users, or or maybe it's a bit of both? Well, absolutely, it's a bit of both. And also I think it depends when we're talking because, so we we believe that a quantum advantage will come when we have fault tolerant quantum computers, and we don't rule out the idea that a quantum advantage will come before then.
But we also are definitely not betting on it. So a quantum advantage being a time when a quantum computer can do can solve problems that conventional digital computers could not. So yeah, we we want to have things that are based on concrete speed ups. We want to be solving problems that are useful for the creative industries in ways that are are faster or more accurately, really, quantum computers aren't about speeding up things that already,
run at a decent rate. They're about speeding up problems that we would currently not even think about. And therefore, we we'll be able to we're able to get new tools, that we wouldn't have had before. So having those, computational speed ups is very important, but that's for the fault tolerant era. So we're planning towards that, but we don't plan to deploy any products with computational advantage anytime soon. Within the NIST era, we have devices. We can use them.
And for that, it is more about creating something that is, unique. And so it's more about the quality of the results than than anything else. So if we for example, one thing that we are quite interested in is looking at simulations of quantum time evolution because this is something that quantum computers are very good at. It's one of the things that quantum computers are naturally good at, so we we can see good results in the
NISC era. It's something that a lot of companies are looking at for its advantages in in health care and material design. So, there's gonna be quantum computers and quantum software dedicated to that. Now how can we use that in the creative industries? And, well, just as a very hand wave of the explanation, physics is is very important in all of the creative industries. Musical instruments
are physics. They are, you know, the the waveforms, that they are creating is all due to the physics of of vibration of strings, of of how it affects the, the air. In in games, you use simulations of physics in order to implement, certain game mechanics. A lot of a lot of games are based around, the how things fall and how things move. So physics is very important within game engines. And in fact, game engines are often used outside the game industry because of how good they are at physics.
I was at a games conference recently where there was a talk by people who worked from one of the formula one teams on how they simulate the tracks for the for the players to practice on. So, I think, physics and visual media is also interesting. You get a lot of artists now who are producing digital media, producing things like animations, and these are using some sort of simulation of physics as well. So then the idea is, well,
what about quantum physics? And in fact, in some of our artistic collaborations so far, they have been using ideas and simulations from quantum physics to generate things that they find very inspirational. I made a thing called Quantum Blurb back in my time at IBM, which, was it it basically intended as a a toy to for people to play within hackathons.
Because coming back to my hatred of, quantum randomness, I found that when there were quantum game jams, then often people who had no idea about quantum would come to the quantum game jam with the idea of making a game and learning something about quantum. This is what people often do at game jams. They make a game over a short amount of time and use it as an opportunity to learn some new piece of programming
that they haven't done before. So they came to learn quantum, but as far as they got into quantum was learning that if you do a Hadamard and then a measurement in a quantum circuit, then you get a random number generator. They made the random number generator, and then they made a game around it, which was entirely non quantum. So I thought it's good to help people start off not at the ground level, but at something a little bit more sophisticated by giving them a tool that does something.
So I made this thing based on simulations of essentially simulations of quantum time evolution, that could be used to generate the kind of textures that you can use in very simple terrain generation, in procedural generation. And this was picked up by an artist who has now been using it as his main tool of expression for the last five years, and he tells me that he could keep on going until he dies just basing his art on on quantum blur.
So, yeah, it's really become the the quality of the output is is what is important to him. He finds it inspirational, and this is something that we see with other artists as well, not just in visual media, but also in in in, music as well. I see. So so the company's interest in in visual art goes beyond, creating visuals for games.
You're you're working with artists. I mean, do do you work with people in film and television, for example, to create, visuals for for those media, or is that is that something possibly for the future? So well, we're a young company, so we have lots of plans and lots of contacts. And, we are actively pursuing those kinds of, activities at the moment. But it's not not anything we've got signed up that I've got a a nice official thing that I can, talk about it on a podcast and provide.
I see. Well, we'll have to we'll have to get you get you back to talk about that. And finally, James, music is also a a a very big, target, I suppose, for MOTH. Company's very active in developing, something called AKTIA. Did I say that right? AkTIA? It's AKTIAUS as an s. AKTIAUS. Oh, sorry. AKTIAUS. And that's a quantum synthesizer and sound controller. Can you can you talk about that? How does it use quantum physics to make music? Yeah. So well, the names of all of our products.
So this is maybe one thing just to take an aside to to emphasize. Yeah. We we are a company that wants to be building products based on quantum computing. Even though we don't think a quantum advantage is necessarily coming into a fault tolerance, we can still be building products, in the meantime. And so Acteus is kind of step zero, you might say, because it's entirely based around, simulations of of of quantum.
So you're emulating a quantum computer rather than actually running on a quantum computer. And it's essentially, you could think of it as an educational tool or or a way of getting your first taste of of quantum because it's basically a block sphere that you plug into a synthesizer. So a block sphere is a representation of a single, qubit.
So, you know, for I've been talking maybe throughout most of this assuming people know what quantum computing is, but, of course, it's built on the qubit, the the two level quantum system. And so, this means that you can you have these two number two complex values. You have these restrictions of normalization, which means that the mathematics of a qubit or just one qubit is essentially the same as the mathematics of a of a point on a sphere.
So there's this great jet, a visualization of what's going on in a qubit with what's going on in the point of a sphere. On on the point of a sphere, you have the x axis, the y axis, the zed axis telling you where your where your position is. So you could those are three real numbers that you could plug into various, knobs and dials on a synthesizer. So then you can be placing quantum gates and seeing how that, affects the point that you're at on the block sphere.
So you look at the block sphere, you get an idea, from that visualization. But, with by plugging into a synthesizer, it also gives you a way of sonifying it and giving you maybe a bit more of a hands on experience of what happens when you put those quantum gates in. Yeah. So that's that's what we're doing with, Actias. But also another thing that we have out currently is our quantum audio package. So this is a Python package, and it's it's also music based.
It's it's taking a few different ways that people over time have proposed encoding music in quantum systems and also taken a few different ways that people have proposed, encoding images in quantum systems and then hacked them to do music instead. And then that gives you a way of encoding music in quantum systems in a in a more sophisticated way than in Actias, but in also less of a a fun hands on way.
I see. And and just in terms of, you know, your customers, how would I mean, you mentioned that you are simulating quantum computers for for some of these applications, but would, you know, if you did have a a quantum algorithm that a customer was using to, I don't know, generate material for a game, would they would they then lease time on on a on a quantum computer from a hardware provider?
I'm you know, I'm just trying to understand how, you you know, the the the whole sort of ecosphere of of the industry and how how you would see that happening. Yeah. Yeah. So these these early things there in the or Actus at least is in the realms of simulation, but everything that we build in future, we want to be, moving into actually running on hardware. So so yeah, the infrastructure of that, how does that work?
Well, one sort of dream I have in some ways is that, IBM Quantum offers everyone ten minutes free access on their systems per month. And so what would be great is to build the software that people can plug it into their IBM Quantum free access, and use that and, you know, really use those IBM quantum systems so much that they they have to think about revisiting that policy. So that that's that's the dream I have.
But, more generally, you know, one of the main sort of the product of all products, the thing behind everything is gonna be the the infrastructure, the back end, where, all of our products point to this way that we are handling all of the quantum system. So the people themselves will not need to worry about, how they're going to run it on a quantum system, how they're gonna do the error mitigation or error correction when the time comes.
They're not gonna need to worry about which device they have to use or which qubits on that device are the best qubits. All of that we'll take care of, and they just have to come along and, and use the product. I see. And if you could, you know, sort of gaze into your quantum crystal ball, when I mean, when do you think I mean, is that something that can happen now, I mean, you know, in terms of you or a customer booking time on a on a quantum computer and doing something useful.
Or is that something for five years in the future or ten years in the future? No. It's absolutely now. I mean, we've got a reservation for one of the QPUs tomorrow morning, for example, where we are gonna run a bunch of stuff. But also, it can be done at any time, as long as you're willing to get in the queue. So, these things are already there. We can very much run things on quantum computers now.
And in fact, our our forthcoming projects, which are based around quantum reservoir computing, are much more solidly based around running it on real quantum hardware. There is a, secret demo of our upcoming game that that is out there in the world if people can go and find it.
But, once that goes beyond being a demo, it will be something that, could for example, we haven't figured out the details yet, but it's it's it's gonna be a a game in which levels are generated procedurally, and they'll be generated on a quantum computer. So it might be, for example, every day it refreshes. And so every day we're running it on a quantum computer to get the new level. Right. Any any hints for our listeners about, about about where they can find this demo?
Well, it, if they're a 10 year old, they already know. Let's let's get it like that. Okay. I'm sure we've got a few ten year old listeners out there. And if I can if you can indulge me, James, just by answering one more question. It's the International Year of Quantum Science and Technology. And so whenever I interview somebody, associated with quantum science and technology, I'm asking them one question, and that question is, what does quantum physics
mean to you? So do do do you think you could have a go at answering that question? And, you know, you can say anything. It doesn't Yeah. It doesn't have to be technical. It can be personal or whimsical. Or Mhmm. Well, I've been so deep in it for a long time now. So it means just, it's what I do every day. It's it's, so it's my job, but it's also
my hobby, I suppose. The the reason I got into doing these kinds of things, looking at games and art, was because I wanted to supplement my my job of quantum error correction with things that were gonna be more accessible to the general public. The the first thing I ever did in this space was in 2016 to make a citizen science game. And I think I've I had an article in Physics World. It might have been about that, but it if it wasn't about that, it was on about one of its successes.
So, yeah, that's what it means to me. It it's a job, but it's also a passion. Oh, that's great, James. And I will put a link to that article in the notes for this podcast. Thanks so much for joining me today, and, best wishes to you and all your colleagues at Moth for the future. Thanks a lot. It's great to be here. That was James Wootton of Moth Quantum.
It was one hundred years ago that Werner Heisenberg formulated matrix mechanics, which provided an appealing mathematical framework for explaining the puzzling observations of the quantum world. Heisenberg's inspiration came on the German island of Helgoland, where he was seeking relief from his severe allergies. But what if Heisenberg had never made the journey to Helgoland in the first place?
In the latest episode of the Physics World Stories podcast, the writer, Kevlin Henney, discusses his new flash fiction, Heisenberg, not in Helgoland. It was written exclusively for Physics World as part of the International Year of Quantum Science and Technology. The story spans two worlds, the one we know and an alternative reality in which Heisenberg never visits the island of Helgoland.
That episode is hosted by Andrew Glester, and you can listen to it on the Physics World website or at your favorite podcast provider. Just look for the headline, alternate quantum realities. What if Heisenberg stayed at home?