Pod Insights a16z Podcast a16z Podcast | Quantum Leap About this episode Quantum computing. The dream of the 1980s is alive in 2015. But this time is different, for a number of reasons. What are the applications, beyond cryptography? And, what’s the point of it all — are the benefits of quantum computing truly going to be a difference of kind vs. of degree? And if so, how will it change what we already know about computers, from languages to coding to UIs to everything?
While it’s counterintuitive for us to imagine something so infinitesimally small, Ilyas Khan of Cambridge Quantum Computing helps us make the quantum leap in our minds … in this special edition of the a16z Podcast, brought to you from London.
Transcript 0:00 Speaker 2 Welcome to the A16Z Podcast, London Tour edition. Sonal and I are sitting here with a gentleman we're very excited to interview Ilias Kahn, who is the founder and CEO of Cambridge Quantum Computing. He is the chairman of the Hawking Foundation. He is the patron of the Accrington Stanley football club from your home home region of Lancashire. And he's actually a huge rugby fan too. So philanthropist, technologist, mathematician, investor, you seem to do a lot.
0:33 Speaker 0 Ilias, welcome. Thank you. As I said to your colleague, I'm also quite old, so one manages to fit a lot into these years.
0:40 Speaker 2 So Sonal and I are going to discuss all sorts of things, and maybe we can start with the quantum computing side of things. I mean, we've been talking about quantum computing computing for for quite some time. And I know that that your firm, Cambridge Quantum Computing, has been working on an OS. But maybe we can back up and sort of tell us, you know, first off, what is the promise of quantum computing and kind of where are things, what's the state of the art as you know it and as you guys are working on it?
1:15 Speaker 0 Right, yeah. So it's it's my third startup at uh the grand old age of 51. And both of my first two startups were in technology as well, but uh in my late 20s and 30s. And quantum computing, as you rightly say, is something that's been anticipated, talked about, written about for a very long time, and I guess takes its roots back to the terminology that was used by Feynman in the early 80s. And effectively it's the way of trying to avoid uh the use of gates uh in the form of transistors and employ nature to compute for you. So Feynman's famous lectures on quantum computing, or in fact on computing, of which one lecture is on quantum computing dates back to the year 82, I think, or 83.
2:05 Speaker 0 And the promise of such a machine, the ideal of such a machine has been exciting people really for th more than 30 years. Right. But it's one of those technologies that whenever you read about it, um, it's always in the next 10 years. For the last 30 years,
2:24 Speaker 2 yeah. That's
2:24 Speaker 0 right. That's right. And my encounter with with quantum um computing, in in probably in in a form that is more meaningful, occurred in the early 2000s. I was involved in running a fund for the government of Hong Kong, and it was a technology fund. It turned out to be quite a successful fund. I ran that fund as an investment manager with a partner of mine for three years, and I had the good fortune of going to um Stanford, in fact, uh, and I think listening at that time to uh Martinis. And um, and that's where my my interest in this as a uh as a technology that might become a business takes root. The current state of affairs is um blessedly simple to describe, actually.
3:14 Speaker 0 Uh most of the technology reports um can be divided into two sectors. One is the work that is done in academia and that is largely theoretical. And probably the tipping point occurred late last year, when the work being done by corporate entities, which have a profit motive became more important than the work that was being done by academia. So, in many respects, if uh let's say that we were looking backwards from the year, let's just yeah, use 10 years. Let's say it's
3:50 Speaker 2 and just to be clear, those corporate entities at the lecks of I know Intel's putting money in, IBM, et cetera, but uh what are who are we talking about? It divides neatly into sort of three or four categories.
4:01 Speaker 0 And um the the the quick sort of elevator pitch version of the answer to your story is as follows. The giants on the, should we say the the Western world, um, Europe and America are very clearly dominated by Microsoft and Google, of course, IBM, you mentioned Intel, they're a relatively recent entrant. And in fact, if you look very carefully at the annual reports of people like Google and Microsoft, you'll see that this is not a recent investment. Very large amounts of money, both as a percentage of their research programs as well as in absolute terms, have been spent, particularly by Microsoft and Google on quantum computing. Uh Microsoft has a fantastic campus which is devoted to um quantum computing.
4:51 Speaker 0 The second block of corporate activity is that which has been sponsored largely by the state and state-owned entities in China. And two years ago, we were all uh privy to the news that as far as the Chinese were concerned, quantum computing is their number one priority. It's not water, it's not population control, it's not food, it is quantum computing. There is no budget, which means there's no limit to the budget. It means not that there's no budget, there is no budget. They will do whatever it takes to be a leader in this field.
5:29 Speaker 1 Which it could be actually quite fascinating because I don't know if the US government is investing that much. Oh, yes. It is. Oh, yes,
5:35 Speaker 0 oh yes. There's some amazing work also now being sponsored on quantum cryptographical communication on the infrastructure side. And the Japanese, uh Toshiba, for example, uh, have been really very prominent in this area. So this means that if and when we have true quantum computing, the means and the infrastructure through which we might be able to effect such secure communications needs to be deployed. Uh Chinese, of course, are there as well.
6:04 Speaker 2 For probably the same reasons. Cryptography has a lot to do with it, I'm sure, too. I think Securecoms
6:11 Speaker 0 is definitely the easy and low-hanging fruit in conversations like this, when we try and ascribe reasons or rationale for why governments would be involved. But but I think it's actually slightly misleading. If the promise that Feynman held forth in the early 80s is even remotely true, then the impact of quantum computing on our lives as we sit around this table will be greater, for example, than that that our forefathers during the industrial revolution had during their lifetimes. And certainly, if you looked at the transistor age, roughly from the time that um let's say 1945 up to about 1980, before we got into the the my the the minutiation race, we've all I've certainly lived through that, and I can think of myself as being the recipient of amazing changes.
7:01 Speaker 0 My children live in a world which is very different from one that I lived in. However, those changes would be as if nothing compared to what will happen if Feynman's and other people's um aspirations are to be even remotely true. Alibaba is a represents what I would consider a th sort of a fourth category. These are not traditional technology companies, but they're people who've decided that they want to be in the game. Alibaba is just one of them. Well, Samsung has a very big program. Apple are very interested. And so these are new Intel we mentioned earlier, these are new entrants. Um Cisco, for example, Hewlett-Packard, uh, Ziemans. These people have businesses that if it is in fact true that quantum computing will become prevalent, and let's just say within 10 years, because 10 years is not that long a time, um, then these people have something at stake.
7:58 Speaker 1 Ilias, I want to ask you a question about what enables these players now to enter the game where they may not have before, but before we answer that, could you actually describe what the benefits of quantum computing are? Because you you describe this vision that Feyneman had, Richard Feynman had. And we understand that, you know, the problem it's solving is that going beyond the constraints of gate logic-based computing. But what does that give us? And what are the applications that are possible because of that?
8:28 Speaker 0 I think that the honest answer is that we actually don't know. And you know, I was a youngster when the first mobile phones came out, and I had no idea, no idea what a mobile phone would do for me, other than the fact that I could call my mother whilst I was on the motorway. And I think that that's although it's it can be trite, and I I don't want to descend into sort of aphorisms, but I don't think we actually know what impact truly quantum computing will have.
9:01 Speaker 1 When mobile phones first came out, you're right. Most people thought of like the application being the phone, when in fact the reality is that I think now most people use it for the computing than the actual phone telephony aspect of it whatsoever. I mean, I don't think people even expected the app ecosystem and app stores and all the other things that would happen, even though there has been this vision of ubiquitous computing for a long time.
9:22 Speaker 0 I'm glad you're supportive because I I use the analogy. I I speak about quantum computing a lot, uh often to business audiences. And I think one has to try and be accessible. There's no point using very long words that sound incredibly um complicated and impressive if you're not making a point. But it there's also a balance where it becomes slightly uh diminished. So that's point number one. Point number two, uh, now now to be a little bit more specific. Speed up in computation is what a quantum computer is about. Feynman and and and all the rest. Of the greats, uh Charlie Bennett at IBM, Peter Shore, uh Shore's algorithm, um, you know, Kitayev, all the people at Microsoft, of course, Martinis now at Google, you know, these are the giants of the of the of the industry.
10:17 Speaker 0 Um, my own partner, founder in in Cambridge Quantum, a guy called Takis. These people are very grounded, and their approach is that let's just focus on ensuring that the exponential speedup that is offered through pure computation can be achieved. Now, the applications of that we can't even imagine. You know, would it be the easy analysis of some genome-related uh aspect that is currently beyond us? Uh, for example, the thousands of years that it might take current computers to solve a certain optimization problem could be done very quickly. It could be, as we've discussed before, secure communications, which are truly secure, so that um we're not attempting to factorize large numbers, which is currently the case.
11:09 Speaker 1 The benefits of speed are not what help make quantum cryptography possible. It sounds like it's more a benefit of the fact that it doesn't have to be based on primitives of a certain type.
11:20 Speaker 0 So there are two aspects to cryptography. The current state of affairs is that when we communicate, more often than not, the public key encryption that we use is based on the inability of computers to find factors for large numbers. And so breaking or creating codes will no longer be possible using that particular mechanism. Now that that can get very scary because imagine if somebody had a secret quantum computer, um, and I don't know, JP Morgan, Citibank and Barclays and HSBC suddenly woke up one day and found that their accounts were empty. Yeah, I mean,
12:01 Speaker 2 the the NSA has recommended that people start using quantum encryption techniques, right? You know, even though in theory quantum computers don't exist, but
12:09 Speaker 0 still one aspect is cryptography in the form of codes, both code breaking and code creating. The second aspect is what Einstein um was famously uh responsible for being very skeptical about, and this is uh a spooky action at a distance or entanglement, or I mean there are many different ways of describing it for the average person on the street. But what we're talking about here is the fact that the mere act of observation in a quantum world breaks the quantum system itself. So the message disintegrates. So don't think of this. Isn't this the
12:49 Speaker 1 Schrdinger's box problem or
12:51 Speaker 0 uh that's one aspect of it, yeah. Yeah, the cat's alive and it's dead. But what the way I would describe it, and I can take no credit for this, this I heard from um Leonard Suskind at Stanford. Uh by the way, I should take that back, the great Leonard Suskind. And he teaches a course, which by the way is accessible to all is on the internet and Stanford make it available. But he defined something called the uh basic minimum, and he teaches courses on the basic minimum for uh classical physics, for mechanics, for quantum physics. And he describes this issue of entanglement um as follows. He says, in the classical world, the world of the big, we can measure things because if I measure the distance between the two of you who are sitting on the other side of the table from me right now, I could use a ruler.
13:45 Speaker 0 And the ruler would not disrupt the fact that you're sitting there, and it's relatively easy to say that you're 12 inches apart. Uh luckily we're in England, inches still matter. Now, it does the point is that it has a very um temperate and very low impact on its environment. I could even use a laser. I could I could uh have a laser ruler, and that would not affect the fact that it um is pretty accurate. Mind you, when you start
14:14 Speaker 2 then going into the world of the
14:15 Speaker 0 quantum, and remember we're talking about subatomic things, um, these are incredibly small. Um it it's really doesn't our we're not equipped to to to describe or imagine.
14:32 Speaker 1 Right, even conceive of it, right? We can't even conceive of it.
14:35 Speaker 0 We we we're we're hardwired to conceive of the big things. We can talk about the universe, but when you start going the other way around, our minds abandon um reality very quickly. Nevertheless, when we try to observe subatomic particles, the ways in which we observe and measure, which use light, or an electronic charge or some other such invasive instrument affect the actual state of affairs. And so that's what we have been referred to in the past uh when you mentioned Schrdinger's cat or action at a spooky distance. But there are ways in which communications can be safeguarded using something called, for example, quantum key encryption.
15:23 Speaker 0 And if um some bad people tried to um interfere with the message being sent, the message, a bit like Mission Impossible, it would sort of break up or dissolve or something.
15:37 Speaker 2 So this is where my head starts to hurt when we talk about quantum computing, and and you know, you just described this whole list of very large, very profit-driven, very in some ways rational companies, although Google you could you could argue whether it's rational or not, but they do interesting things. And you're saying that we don't know what it'll be used for, but and yet everyone is very optimistic about what.
15:59 Speaker 0 Oh no, I what I I think that's uh if if I've given that impression, then I I I apologize. What I said was I think it's very difficult to be specific about the applications that might occur. It is not difficult to realize what is strategically of great consequence. And the profit-driven uh businesses that you've mentioned, and ultimately I'm a profit-driven business as well. We can see things which do matter, and which a quantum computer, perhaps in the early days, aligned with classical computers because all of our classical systems aren't going to disappear in a puff of smoke one day. And I'll give you a couple of examples that are generically talked about even today.
16:45 Speaker 0 We've spoken about the fact that secure communications um will change in ways that aren't even incremental. They will just simply shift. Right. And there is a large infrastructure of people, whether it be governments, whether it be defense departments, whether it be banks, whether it be other types of companies who communicate all the time, and they will need products, and they will need machines. Now, is that is that 100,000? Is that a million, or is that 20 million customers? It's certainly a very, very, very large market. And so I think that it's only right and proper that we allow for the reality that there is a strategic defense and communication related application in the early days.
17:32 Speaker 1 Just like the early days of the internet, in fact.
17:34 Speaker 0 Yes, absolutely. And so many of our technologies. Right. Uh the the Apollo space mission program was shrouded in the kind of secrecy that was really a hangover of the Manhattan project. Now the other thing is, and this is very close to my heart, um, optimization problems. One of the great optimization problems, actually, is tracking signals from data. Now it could be stock markets, um, it could be people who visit Expedia. There are all kinds of things which currently take time, and in fact, not only take time, but become exponentially impossible to manage simply because of the sheer volume of data. Um
18:14 Speaker 1 because they're complex interactions, not just the size, it's actually the way that data interacts with each other.
18:20 Speaker 0 That's right. We know the maths, we know how to do it, but actually converting a piece of paper or uh an algorithm into a workable solution is where the machinery breaks down. And a quantum computer, even with a modest number of qubits, uh and and the th the analogy that gets bandied around by everybody who's been on the D-Wave uh website or who's read all these articles, is that a quantum computer with 40 or 50 entangled qubits would be more powerful than anything that we currently have in existence. Um is true. We're not there yet, by the way. But when we get there, these machines will have an amazing capacity to solve these types of problems.
19:06 Speaker 0 One of the applications from that would be the quick and careful analysis of our um DNA and the creation of designer drugs. So that's one example uh that is very interesting. Another, which is a little bit more hitting your pocketbook, is that the holy grail for stock markets has been predictive behavioral analysis. And at the moment, it doesn't exist. But it could exist. And it'll even the playing field. It won't take away from the fact that we still have to analyze companies and decide which companies are good and which companies are bad, but then investment will become closer to what um Benjamin and Buffett and all these people have been talking about.
19:53 Speaker 0 That's why I say it's holy grail. What is really relevant, which has eluded uh people forever is. Is to try and find out whether a stock or a bond or a treasury based on the collection of vast amounts of data is likely to move one way or the other prior to its movement.
20:12 Speaker 1 So very predictive in a completely new way. Just one interruption here, Ilias. I'm going to say something very blasphemous. Is it possible that we're also talking about something very academic? That's uh something that's a a technology that's seeking a problem. And the reason I'm asking this is because there are good enough solutions that can approximate some of these answers to these very questions. Why wouldn't a very powerful simulation system that computers can currently do address some of these same things? Like what does quantum computing really get us?
20:43 Speaker 0 It was a very fair question. I think that maybe up until a few years ago, you're right, but it's no longer correct, uh i.e. the this issue of uh being academic. And there are two or three reasons. First of all, self-evidently, classical computers aren't equipped. And even the simple example of the factorization, the only reason we have RSA and factorization-based cryptography is that classical computers take thousands and thousands of years. Even today, that little machine of yours, the iPhone that I've got or whatever, I don't know how many um transistors are in there, but two billion or something. Um and the fastest cray machines, and even if you wired all of them up together, they simply can't do it. It's beyond them.
21:28 Speaker 0 And so there are a class of problems which we know will become open. So that's not something that is merely speed up.
21:36 Speaker 1 So it's a difference of kind, not a difference of degree that we're tackling. So the whole
21:39 Speaker 0 Feynman, that's right. That's the whole Feynman thing. We're using nature. There's always a better, faster, smaller gizmal that comes around. And the better, faster, smaller gizmal then allows smart people to do something with it. So I don't actually I think on this occasion, it's not trying to solve a problem that doesn't exist. This is trying to solve a break in our development. And that break was that we back in uh, you know, when ENIAC came around and when Turing was talking about his machine, we started thinking of using uh the binary system, whether they were vacuum tubes or lamps or whatever to compute. And we got better and better and better at doing it. And silicon, of course, was a godsend.
22:25 Speaker 0 If we're able to harness nature, it'll just be a different way of doing everything better. That's all. I don't think we should think of it in in magical terms. The reason why it's no longer academic is that the advances in engineering, the same reason why we have the large headrun collider, the same reason we have well, there's so many advances in engineering, of course, now, because the I think the gate fidelity of an average transistor now is I think five or six thousand times thinner than than the than our hair. So that ability to engineer at that level is the only reason why we're now even thinking of being able to build machines. It's it's nothing
23:09 Speaker 2 else.
23:09 Speaker 0 I
23:09 Speaker 2 have a question. I mean, Sonal, you mentioned that it was a difference of kind, and I know technology when it emerges is not evenly distributed, but if it is such a difference, how do we how do you think this goes about being distributed? Does it go to Wall Street first? I mean, and and that honestly doesn't seem fair. It should go to, you know, genomics first. But how do you how does this power then get distributed into the world, do you think? And is there a a way to think about quantum computing that we haven't with maybe just the evolution of computing thus far?
23:40 Speaker 0 I do not know the answer. What I do know is what we all know, and so I'm stating the obvious, which is new technologies, whenever they emerge, are very quickly taken up. Right. But in the interregnum, the period when they're only selectively available, you either have government intervention or you have some sort of monopolistic Ayn Rand sort of um, you know, uh corporate that ends up uh for a while reaping the reward. So when the well, you uh in America, of course, you're familiar with the the railway barons. So I I don't know how this will play out, but in different parts of the world, I do think that is well, in fact, not only do I think, it has become self-evident that the race to build and manage actual quantum computing has become strategic at the government level.
24:34 Speaker 0 And that is why when suddenly Intel invests in quantum computing, because remember, this would cannibalize their semiconductor business in time.
24:44 Speaker 2 Oh, yeah.
24:46 Speaker 0 And when Google has made such massive advances, we are no longer talking about the the core expertise residing in universities. The budgets, I I'm connected with the University of Cambridge. The budgets that universities have had, even though they might sound large, are piddling compared to the corporate budgets. And in the last 18 months, that tipping point has happened when engineering solutions are being found. Now, I'm not on the hardware side. I'm on a different side of the business, which I find much more fascinating and where I think the great fortunes are going to be made. But I still can see that my counterparts who were at universities are no longer at universities.
25:37 Speaker 0 They're either sponsored by people like IARPA or they are being um employed by people like IBM and Google.
25:45 Speaker 1 You alluded to some of those four categories of players. You describe government, some of the big companies with RD facilities, you described other people who are sort of watching and engaging in the space. And then you describe sort of the new technology entrance. What about startups? Unlike the past analogues we have here when we look at the history and evolution of computing, where there was sort of this diffusion effect that happened from government to industry and then to startups, with sort of an Amazon web services like moment where it created this Cambrian explosion and the types of businesses that could be built. This is kind of all happening all together. Of course, the groundwork, as you've said, has been laid by universities. There are now a lot of big corporate RD players, there has been government funding. How can startups play in this world and what is what has changed that it has stopped being an academic to more of a practitioner-oriented enterprise?
26:34 Speaker 1 Like how can you do a startup here? How can D-Wave do a startup? Like, how can a startup, which by definition has has to bootstrap its resources actually engage and play in this space?
26:44 Speaker 0 I can give you my perspective, which is very, very limited. Um, first of all, I think that the big hardware game is absolutely not startup. The people that have got the resources to be able to attack this problem are by definition the big guys. And D-Wave uh is not a startup. D-Wave has been around for a very long time, almost 20 years. Uh and I, by the way, big fan of D-Wave. Um, shout out to Colin Williams, the CTO there. Uh, and so I I'm one of those that that admire uh the pioneering work that D-Wave have done, although I don't think their machine is a true quantum computer. The startup space, however, and this is you know why I'm so excited by this.
27:30 Speaker 0 All of these machines will just be refrigerators or or or or piles of junk if we don't have the quantum algorithms that can make them do things that we've been talking about.
27:43 Speaker 1 The applications.
27:44 Speaker 0 And if you look at the landscape, if you turn your eye backwards, the landscape is very sparse when it comes to actual quantum algorithms. Now, one of the reasons for that is that maybe there wasn't the commercial need, and it was dominated by academia. And so what would a scientist do? You know, he wants to publish a paper. He wants to be published in um in a journal. Startups are going to change the world, not the guys that make the machines. And the reason is that this is probably one of the purest forms of individual brain power. This is where the individual Einstein or the individual Picasso can make a difference. So we're on the mathematic side. We're gonna make these things work, and our customers are the people who will buy the machines and then say, Oh, hang on a minute, what am I gonna do with this?
28:35 Speaker 0 How can I make this thing work?
28:37 Speaker 2 I mean, what you guys have been described as uh as developing an OS for quantum computing, is that fair? It's more complicated than that I know.
28:44 Speaker 0 No, it's not complicated. I I I just I I have a real constraint um in what I can and can't say. But I just have to be a little bit generic here. So we the ambition behind Cambridge Quantum. So the question might be why did you start coming, you know, what's the the reason, the rationale. And that is um two years ago, I came across this uh amazingly talented individual, the uh one of these um uh uh Picasso types, and nothing to do with corporate um uh backing. And uh that time I taken a position of leadership on a pro bono basis for the business accelerator in Cambridge. So this was an attempt to help the University of Cambridge judge business school do what people like Stanford and Harvard Business School do in um commercializing university brains.
29:32 Speaker 0 So I was doing this, and I came across this guy who was absolutely brilliant, nothing to do with the accelerator program, just I met him in the math department, and our ambition was to try and develop quantum algorithms to create intellectual property to patent a whole bunch of these things so that eventually people would have to use or license our know-how. Right. And it's driven by mathematics and within mathematics, driven by certain aspects of mathematics. Which lends itself to the area of quantum computing, which happens to be graph theory. So that's how we started. I should say that in those days, I really thought that this would be something which would keep me occupied academically and of interest over the next eight, nine, ten years.
30:20 Speaker 0 Easter last year, 2014, was when everything changed. Everything changed. It's when Google disclosed the fact that they were so heavily involved. They went and bought the University of California, Santa Barbara team, I don't know, 50 people or something.
30:37 Speaker 2 Only Google could do
30:38 Speaker 0 that. And it's very easy to say, oh, they're doing self-driving cars and you know, spectacles that can compute, but this is a different order of magnitude. Artificial intelligence is what really turns these guys on. And we looked at each other and we knew that there has to be an operating system. So we then uh spent a lot of time building a simulator, a hardware simulator. Of course, a simulator in the classical sense will never be anything like as effective as a real computer, but it gave us the ability to start thinking of an operating system. Alongside that, what happened in that journey is what you guys have been asking me about. We suddenly started to realize that you could create applications.
31:25 Speaker 0 I
31:25 Speaker 1 was gonna ask you one quick question about this, just because um it's a theme that comes up for us a lot. You mentioned in your origin story, you mentioned that you guys originally started out trying to create IP intellectual property and patent, which is fascinating because you also said that part of the model that you were trying to help the Cambridge judge school with is this business accelerator. And one of the defining things of Stanford compared to a lot of different universities is it actually gives its IP away in order to enable startup founders to just freely create their businesses and then it expects to sort of reap the benefits when they come back and do more philanthropy as a way of actually so they have a long game in mind. How does that play out at Cambridge?
32:04 Speaker 0 Well, Cambridge is not there yet. Uh if I had anything to do with it, we would be there. So just to be very specific and accurate, I donated a relatively large amount of money to the judge business school, and that was the seed capital that the judge then used to back businesses. I was on the original investment committee that then helped choose these businesses, and this is a program that has become very successful. In fact, one of our businesses was just bought by Apple. When I say therefore pro bono, I meant that my involvement was not for profit. But I got so involved in looking at these youngsters with brilliant ideas, and these were hard tech, not some other app or well, but actually I shouldn't diminish that.
32:53 Speaker 0 They were predominantly hard tech. And the point you've made about Stanford, in my opinion, is compelling. My vision of Cambridge is that we would eventually be as far as possible on the same footing. Now, of course, we're not Stanford, and the Cambridge phenomenon has been well described and documented. It's been enormously successful in in Europe. And my modest involvement then led to my doing my own startup. So whatever we're doing is not a spin-out from the university, because we didn't take a technology, it didn't exist. But what I did is I I stumbled across this amazing mathematician, absolutely amazing mathematician.
33:35 Speaker 2 I I have a question, you know, in in computing, typically hardware has always outstripped the software. So more recently we had multi-core, multi-threaded, you know, silicon. And but nobody could program for it, you know, and and we're trying to sort of is the same. It sounds to me like you guys are neck and neck right now. I mean, and I don't know because you can't say too much, but is that the same dynamic going on in the quantum world as the hardware kind of racing ahead of the software? Yes, absolutely. Yeah,
34:06 Speaker 0 yeah, yeah. And I think that the if there is to be a reservation, if we are to pause for breath and and and and and and remain grounded, then the challenge for Cambridge Quantum and with any other startup that's focused on the software side is exactly the point you've made. You know, why is it that in this area we would be able to keep pace? The only thing is it's actually an empty landscape right now. I'm not aware, I'm sure there must be businesses that have started up that are looking at quantum um software, quantum computing software. And in fact, there are a couple that are very focused on financial services. But I'm not sure that I'm aware of anybody with the mathematical approach that we've taken.
34:53 Speaker 1 So actually, Ilias, I want to ask you to clarify that because when you described math when you describe the whole space and saying that we're not on the hardware side, we're doing the math side of this equation, no pun intended. Um it immediately, when I think of math, I think of software. And when I think of an operating system, I think of software.
35:10 Speaker 0 There's a whole layer upon which people can code in the classical world for 40 years, whether it's an operating system, whether it's Linux or whether there are um intermediary systems, whether it's your iPhone, you know, where it's all plug and play, if you have expertise in what we generically call coding, you can do software. Well, you know, in quantum computing, we don't even know how these bloody things are going to work. It is, and and where we have the advance, where in fact we have Nobel Prize winners. We have people at places like Cambridge and Stanford and MIT who know how the maths works, but they've never used it other than to do maths. So if you look at Microsoft, this topological Q bit, you know, the program that they're doing, they've got Fields Medal winners taken away from the blackboards, not people who code.
36:06 Speaker 0 The code guys can come later. So that's why the coding guys.
36:11 Speaker 1 So it's too early to answer the next question, which is what would the future of software development look like with a quantum computer? I mean, do we have any ideas? Are there people even thinking about it, or is it just simply too early?
36:22 Speaker 0 You have to imagine, and again, we're not hardwired to imagine this, but uh so we we re reduce ourselves to analogies. A machine of that magnitude would be more powerful than every single computer hooked up together. And if you had a power source, I don't know, that was a sun or something, and then you could power it, um, it still wouldn't classically be able to do that. And that's why people get excited. Now, that is a very now who knows whether that will ever happen. You know, I'm not uh so if you talk about small blocks of time, let's talk about the next three, four, five years. I can see that software expertise, people that are just magnificent and equipped to create the sorts of um products that we've been talking about, they will work more closely with initially the hardware than has been the case for a very long time.
37:15 Speaker 0 People who are currently software guys, they don't even care if they ever see a machine. So
37:20 Speaker 1 you're talking like more like bare metal programming revisited from scratch when it comes to the city. It
37:25 Speaker 0 has to be. Has to be. It depends on the machine. You've got to have access to the machine.
37:31 Speaker 1 Having you reinvent the language, the GUI for quantum computer. I mean, you have to reinvent everything.
37:36 Speaker 2 One of the things that you write and talk about and think about uh is the philosophy of mathematics. How does how does this sync up with you know that part of your brain? And and you know, is quantum mechanics and quantum computing right there with your philosophy of mathematics?
37:54 Speaker 0 Most mathematicians think that philosophers of mathematics are people who wish they were mathematicians. And I and I fall into that category. I I'm a very amateur mathematician, which is why at least in my academic career, I I have had to make do with looking at grappling with issues of uh the philosophy of mathematics. My should we say approach to this is very much influenced by uh 20th century philosopher called Wittgenstein, who has been waylaid by people who think he was a philosopher of many other things, but by his own admission, he was a philosopher of mathematics. When we look at the limits of our language, we actually are also talking about the limits of our knowledge.
38:44 Speaker 0 And of course, there's a very definite segue into not quantum, I wouldn't say quantum computing, but the whole debate that has been with us for over a hundred years about the reconciliation at some point of the laws of quantum mechanics and quantum physics with the laws of relativity and gravity, the so-called theory of everything. I don't know whether my children or whether their children's children's children in a hundred years' time will be any closer to a theory of everything. But I have benefited in my short time by exposure to some pretty gargantuan minds, non more gargantuan than the mind of Stephen Hawking.
39:30 Speaker 0 And I am fascinating. Oh, that's right. You were the founder of the Hawking Foundation. Yeah, I'm merely the chairman, I'm not uh Kip Thorne from Caltech of Interstellar Fame is my other colleague on the board, and Professor Malcolm Perry from Cambridge is the other um founding trustee. But to go back to the point that I was describing, I I cannot think of a greater prize for humanity than making progress. In unifying these approaches and thereby understanding the essential question that we've always asked, which is how did it all happen? Where did we all come from? Is there a beginning? Is there an end? I mean, today we know that there are, I don't know how many, there are uh 150 billion or so stars just in our galaxy, and our galaxy is one of hundreds of billions of galaxies that at least we know about, and then there's this whole issue of what we don't know about.
40:34 Speaker 0 And despite our ability to understand these things, there's some pretty basic things we don't understand. And that's where I think the philosophy at least my inspiration in looking at people like Wittgenstein and the limits of language with respect to mathematics comes along. And I don't think that quantum computing is necessarily directly related, but obviously quantum mechanics is.
41:13 Speaker 1 Um, and especially also with your role in judging some of the business plans at the Cambridge Business School. What are some of your views for how we should educate people in this coming age of machines and more?
41:24 Speaker 0 I I mean I actually think that the Western world is sometimes too harsh on itself. And I think the real debate that determines the difference between our approaches in society is the extent to which we're willing through our taxes to pay for efficient education. That's a start and the ending point for this conversation. And beyond that, we've actually done very well. We can always do better. So I think you know, we can sometimes be a bit harsh with respect to that question. Now, more broadly, society has many issues that we have, you know, challenges. Do we treat the most vulnerable in our society well? Uh do we treat the ill well, the aged, people like that. Now, these are issues that are political.
42:11 Speaker 0 How do you use the taxes that do you tax? So I think it's too easy to go down that route. Um, and therefore long-winded way of saying that STEM has to be the top of the list, in my opinion. It's remained the key to human progress. There would have been no industrial revolution. There certainly wouldn't have been Einstein and relativity had we not allocated resource, and long may it continue.
42:37 Speaker 1 So it's interesting when you say an emphasis on STEM, you didn't include philosophy in the liberal arts. What drew us to your background as well is that you're sort of uniting all these worlds and your work and the way you think. So I think
42:49 Speaker 0 it's entirely possible for all of us to be engaged and interested, but then to have a specialization that's different. And I think that the here in the United Kingdom, I think we've got a newfound respect for science and technology and for engineering. And it's to my mind a renaissance that actually the Americans are finding as well. Do we all have
43:15 Speaker 2 to become somewhat at least conversant in technology and engineering, or does engineering kind of come to those other sides of people's interests and people's minds? I think
43:26 Speaker 0 the latter. I I think the great success of anything that's engineered is that we don't worry about whether we know how it works. I mean, I'm fiddling with my watch. I have no idea how my watch works, but it works. Right. And so I think that great engineering is all about delivering the solution so that the user has no idea. And the iPhone is another example, the car is another example, and so that's my philosophy. However, in order for that to work, you have to have superb products.
43:56 Speaker 1 Thank you, Ilias.
43:58 Speaker 2 Ilias, we look forward to seeing you on the quantum side. Thanks for all your time. Thank you.
Previous Company Building in Crypto: A Conversation between Brian Armstrong and Chris Dixon Next What's Next for Education Startups Page updated August 25, 2026 · transcript and summary are machine-generated