Quantum Computing 101 with Nicolas Alexandre Roussy Newton & Gavin Brennen

31:11So for instance, does light interfere with a qubit so that it collapses the qubit to a one or a zero? Yeah, so light is a very good qubit in a way. You know, when we look at starlight, that's coherent over most of its travel. It could be, you know, many thousands of light years where a single photon, you know, was in a certain direction and that's like a qubit. You could call this a zero, this a one, and in between a superposition. And that will travel for thousands of light years. So it's, in a sense, a very good qubit. It's hard to make light interact with other light. But there are companies that are building light-based quantum computers.

31:55PsyQuantum, Xanadu, Quandela.

32:01And that's their goal, to exactly use that kind of qubit. And it is absolutely one of the promising architectures.

32:24Nick, when you were initially conceiving of the idea of BTQ technologies, and again, you deal with, BTQ deals with quantum security, did you feel this was a problem that needs to be solved now if we don't even know if quantum computing will get to the point that it's actually useful within a certain number of years. Yeah, for sure. And we started working on the company about a decade ago. And this was about the same time that the U.S. government and the National Security Agency and NIST, the standards agency, started this process to upgrade post-COVID security standards on the internet. So after a decade of research and input from the community and large companies, this past August, there were a series of algorithms that were standardized, and those are thought to be quantum safe.

33:27And the reason why people are adopting these algorithms today is because, you know, one, there's this risk that adversaries can just store encrypted data and decrypt it when they have a working quantum computer. So let me ask you about that. So someone could basically hack into a bank, let's hypothetically get all of the usernames and passwords, not be able to, you know, not be able to decrypt any of the passwords right now because classical computing can't do it. But let's say 10 years from now, quantum computing can finally do it. Some of those passwords will still apply. They can now figure out 100 % of the passwords.

34:11Some percentage of those will still allow, will still be being used by those users. And so people who take snapshots of the world now, even if they can't understand what that snapshot says, will be able to understand it once quantum computing hits and they'll be able to then go into the accounts of anybody who didn't change their passwords on a regular basis? Yeah, sure. So I think there could be a lot of encrypted data that's targeted, especially in industries where it's very sensitive, like healthcare or financial or defense-related data warehouses and so forth. And also just the nature of transitions in cryptography.

34:56So historically, we've done this a few times. where we've upgraded standards. And that process typically takes a decade just because it's a lot of work. So that's already within potentially the time horizon that you're going to have a lot of innovation in quantum computing. And yeah, you also have other problems like hardware, for instance, right? So if you're manufacturing a hardware device today, hey, that's going to be out in the field for a long time. And that needs to remain secure against quantum computers. So yeah, these standards implemented by the US government, they're now mandated. So the Biden administration had an executive order.

35:48And you're already seeing a lot of internet traffic already, like it's already post-quantum. So Cloudflare, which is one of the largest CDNs, I think they already published that almost 20 % of internet traffic is using some sort of post-quantum cryptography. So you're going to see this get rapidly deployed. So can a classical computer have an algorithm or a program that can still be quantum safe when quantum computing is here? Yeah, yeah. Yeah. So that's what the community over the past 10 years has essentially looked at, is how do we develop classical algorithms that can withstand a quantum attack where there's no known quantum attack vectors, essentially.

36:36So let me try to understand that. So from what I understand of cryptography, a lot of it involves, let's say, the factoring of very, very extremely large composite numbers, like a one followed by 100 digits. That size number would be impossible basically for a classical computer to factor into its prime number factors. But a quantum computer could do it fairly quickly, again, with 15 million qubits. But how do you make something that's... If a quantum computer could basically look at infinite possibilities or almost infinite possibilities, how can you make a classical computer that fights that or that protects against that?

37:20Yeah, so the community, we've been looking at different types of underlying math problems that can secure cryptography. So we're going to have to move away from the RSA-based crypto that we've been using for the past 40 years on the internet. So this is a huge change. And I think we're still not sure like what algorithms will ultimately be secure. I think there was like recently, there was actually, so one of the algorithms in this US government process that was studied for years and years was ultimately just broken because one really clever researcher in Europe developed a scheme and published a paper and was able to compromise that algorithm.

38:10So I think there's a lot of research to take place. And I think, Gavin, maybe you can speak to some of the quantum algorithms as well that are emerging in terms of attacking post-quantum crypto because there's been some research in that side of things, if you want to touch on that. Yeah, so one thing is quantum computers can't solve all problems faster.

38:36and in fact like even some really basic problems you might think you can get a speed up and it really just can't like like what like for example the the majority problem like determining whether a string has a majority of zeros or ones you know that if you're close to being half zeros and half ones you have to wait until the last the last bits to figure out right classically quantum mechanically it's the same thing you can get a fact you can get a speed of a factor of two but that's it um so it's not any kind of major improvement that's fascinating actually because that makes sense right so as opposed to factoring a large digit into its divisors where you could look at all numbers simultaneously and sort of resolve it that way there's no way to look at there's only one state here.

39:31When you have just a large string of zeros and ones, there's really only one state. So there's no way to look at a bunch of different states simultaneously. It doesn't make any sense. And there are other problems that we know that are very hard and very unlikely. We don't have proof, but it's very unlikely quantum computers are going to give us an exponential speed up on. So it's sort of like when the solution is potentially a list of solutions, like you're listing the numbers that can factor into a larger number or you're listing the fastest road trips between New York and LA. That's a list. Quantum computers are great at building these lists out of an infinite number of variations.

40:20but when you have

40:24one object which you're trying to understand the properties of like this object, does it have more zeros or ones a quantum computer has a hard time with that Yeah, and it can get subtle too it gets into ideas from complexity theory but one of the main classes of problems is finding the lowest energy configuration so maybe you have a bunch of spins and you want to find the lowest energy configuration where they can interact with each other in little neighborhoods with their closest neighbors. And this is like a workhorse problem that's used for solving optimization problems. And some of those problems are thought to be exponentially hard to solve for both classical and quantum computers.

41:15so yeah it it's quantum is is only good on on like a few types of problems problems where you have big data uh not very good for quantum computers because it's it it takes a lot of effort to put a lot of data into a quantum computer and um you don't want something that has a lot of input or output of data like you probably weather forecasting would not be a good application for quantum computers. But if it's something where, as you were saying, you have something where you can encode the problem, where you can work on superposition effectively, and you only have a few likely outputs, then it's a very good speed up.

42:04So to make something classically quantum safe, first you have to come up with quantum algorithms to make quantum encryption that works. So you have to kind of find ways to encrypt things other than the classical public key cryptography or things like that, which quantum computers can break. You have to kind of have a whole new model of encryption. Yeah. And what have they come up with? Yeah, so there's a variety of them. Like one of them is lattice-based cryptography, where you try to find the shortest vector to particular points in some high dimensional lattice. We don't have a proof, but it's thought that quantum computers won't be able to solve that problem efficiently.

42:56There's others based on hashing, which is just a kind of a basic function you can apply to inputs that scrambles them up. It's thought that quantum computers don't have a way to solve those problems. Again, a lot of these things you don't have proofs for, but there's a lot of smart people that have tried and not found ways to get a speed up with quantum computers on those problems. But like with these optimization problems, isn't that something that a quantum computer can do very fast because you just try all possible paths to optimize at the same time and boom, you have a solution? Yeah. So it's, it's something that's caused a lot of debate in the community.

43:43Um, and exactly, that's the kind of the intuition that, you know, we should be able to sample the space. And, um, you often hear the words tunneling where you could, you know, sort of use the, the quantum mechanical effect of being able to explore spaces that you wouldn't be able to do classically tunneling over energy barriers to to find like the lower energy dips in a landscape um but it there are a lot of caveats there um so it turns out that's uh for some of these problems you can't actually get any significant speed up um but it's it's something that's still being researched. There are certain particular sub-problems you can get to speed up with optimization, but it's not sure, it's not clear exactly how much.

44:55so what is btq your company figured out that you feel should be applied to devices now for a future world where we're going to need quantum cryptography to keep data secure? Sure, yeah. So we've been developing a lot of product lines around hardware, so making sure hardware remains secure, but also looking at blockchain as well. So you mentioned Bitcoin. All of that is vulnerable to quantum attacks, So people need to build new signature schemes. And the issue with implementing these new U.S. government standards is that they're very large in size. So they're not compatible with existing blockchains.

45:46So you need to build compression technology. So BTQ, we've built the leading compression technology to implement those standards within a blockchain environment. And then what we've been working on with Gavin is how do we not just kind of protect against quantum, but how do we harness quantum for things like quantum energy advantage? And how do we implement these NISC era devices or near-term intermediate scale quantum devices to do something useful? And so one of the problems and one of the solutions that we've built is something called QPOW or quantum proof of work, where we've demonstrated using a commercially available quantum device today that you can use these kind of quantum miners to perform validation in a network.

46:47And the benefit being it's significantly, like, in terms of energy cost savings. So there are interesting applications for these near-term devices that we're seeing. So devices that might even exist in the near term, like Google's Willow? Well, they already exist. So like some of these companies, I think Gavin mentioned earlier, like Xanadu, they're developing these boson sampling devices. They're still expensive, but they'll be much more cost effective than these large scale universal quantum computers. So these are all in the near term regime that we're currently in. but no one's really I think looked at the applications for them within cryptography like BTQ has and I think one of the reasons for that is is when you speak with these companies they're very focused on this long-term vision of breaking like using Shor's algorithm and running that and running millions of qubits.

47:57But yeah, we've looked at the near term. And yeah, I think Gavin's also been looking at some other ideas together with our team on that. So are you trying to anticipate what future quantum cryptography will look like? And then building that classically into, let's say, the current Bitcoin? yeah yeah absolutely so we can take these standards that have been developed by the the u.s government and we've also been involved in the in the standards process uh and then yeah we can optimize that for a specific blockchain use case because the standards body hasn't been focused on on blockchain and i think it's just until very recently that blockchain developed an interest in safeguarding against quantum attacks.

48:53So yeah, this is technology where we can build replacements, but it takes a lot of work. You'll need potentially a lot of hardware acceleration as well. And then we can also leverage quantum technology in interesting ways like within quantum analogs for proof of work. that could be more efficient than the existing paradigm. But let's say you're trying to make something quantum safe. And as Gavin said before, sometimes people think something's quantum safe and then some physicist somewhere in the world writes a paper and it's like, oh, no, no, no, it's not actually safe. Do you think future security systems, even on blockchain, will be able to automatically update when there's new methods of encryption?

49:51Yeah, potentially. I think that's the risk and always has been with cryptography as you're in this ever, this race never ends, right? So there's always advancements in computing, advancements in algorithms. So it's a constant battle between the cryptographers and the cryptanalysists to basically do that. So it's always ongoing. But yeah, I think the real concern for blockchain is that there needs to be rapid progress here because it's not clear what's the best solution for post-quantum. We've proposed various solutions, but implementing that across these blockchains is a huge task. And, yeah, it's very much not clear that the community is going to have to move very rapidly to make sure that everything's secure before quantum computing technology is there.

50:57So let's say I'm a bank listening to this because really banks have the same problem as Bitcoin has, which is that they could be hacked and banks use similar kind of cryptography to keep all their systems secure. Almost everything on the internet uses some form of like RSA or public key cryptography. And what should they do right now? Let's say they call you up and say, you got to help us. We want to make sure we're quantum safe. Yeah. Yeah. So I think a lot of the finance industry and large industries have been aware of this threat going back years. And they've been paying attention to it. So they do have a head start over other industries.

51:43But yeah, essentially, they're going to have to be compliant and replace all of their RSA-based infrastructure with post-quantum schemes, whether that's in software or hardware. So they're going to have to work with vendors, make sure that they're compliant, or else there's a risk factor for their business. How aware are they that they need to do that? I understand that Biden said you've got to be compliant. But if I'm like the head of IT at J.P. Morgan, is my job, depending on whether tomorrow I change the whole bank to be quantum safe? Yeah, I think like all the banks have been very forward thinking on quantum.

52:27I mean, there's a reason why Goldman Sachs and J.P. Morgan both have quantum research divisions within their company. and they've been looking at this for years, both on the security side and then also looking at speed-ups for options, pricing, and other things in terms of financial instruments. So, yeah, I think they're all very aware of the threat, but now it's just a matter of finding the best solutions to implement, and that's where BTQ comes in. We've been developing these solutions and this specialized intellectual property that would be needed for certain use cases. And Nick, as we were talking before the podcast started, I sort of feel there's two types of quantum computers.

53:13There's the companies building the quantum computers and maybe allowing access to them through the cloud. And there's companies like yours, which are building solutions that will be needed for a post-quantum world, meaning those solutions are actually needed right now in anticipation of the future at some point being post-quantum. Right, right. Absolutely. Yeah, so it's, yeah, as you mentioned, like the security piece is very near term. It's already here. And you're seeing this adopted globally. And it's now very top of mind, just given the recent Google announcement. But there's also been other major announcements.

54:00A company, I think out of Boston, Qera, announced an announcement last year. So you've seen a constant string of quantum computers, you know, getting far better than just a couple of years ago. And, of course, you have a lot of countries, whether that's in Asia or in Europe, that are developing and investing billions of dollars into this technology. So safeguarding data is going to be critical. Now, is there any kind of Moore's law of quantum computing where, okay, quantum computers are going to have more qubits every 18 months? Or, you know, is there anything that sort of makes sense out there?

54:47Yeah, I mean, Gavin, we've developed some internal metrics for this, if you want to touch on that. Yeah, so there does seem to be something like that if you just chart the growth in numbers of qubits over time. We don't have a lot of data points because quantum computer companies have only been around since really 2010s. But it is showing kind of an exponential growth. another important metric is the quality of the gates we call that the the fidelity because you have to have really high quality gates uh to make things work and um that's also uh the the air rates is is going down uh in an exponential manner uh so there is that kind of scaling and you can make forecasts of when we would expect quantum computers could be big enough to say crack elliptic curve digital signatures for example that's you know used in signing for bitcoin transactions and when you you know place a proposed transaction in a block and you want to add a block to the blockchain and Bitcoin, you do require that digital signature.

56:21And there is a window before the transaction has been approved on the network of an average 10 minutes where if a quantum computer could solve Shor's algorithm for the discrete log problem, they would be able to find out the secret key of basically the owner of the signature and then could start forging signatures. So if your quantum computer could be big enough to solve that problem in that time window of 10 minutes, that threatens the entire integrity of the network. So you can make forecasts on when that event will occur. And so that's something we look at at BTQ. When will it occur? Well, so, you know, like I said, there's a lot of still engineering problems, like how you connect these modules.

57:18But you can get sort of optimistic and pessimistic estimates. And the optimistic in the sense of go quantum is, you know, early 2030s. maybe 2034, 2035. Pessimistic would be more into late 2030s, 2040s. But to your point now, which is that if quantum computing gets to that point then, if someone takes a snapshot right now of every Bitcoin transaction, even though it's totally encrypted and we can't hack into that, they will be able to look back at the past. In the 2030s, they'll be able to look at the 2020s and figure out every transaction that occurred. And that itself could lead to a lot of, you know, havoc and the world, you know, collapsing.

58:16Yeah, well, so, yeah, there is this idea where you harvest now and crack later. Not so much with Bitcoin because people can, you know, use different public keys. So the future transactions won't be affected by past events. But certainly with the U.S. government in protecting secrets or any government, they're using oftentimes public key cryptography. And once that's out there, if you collect that data now, you could crack it later. So, yeah, that's one of the real big incentives to move now to post-quantum cryptography. Because, yeah, exactly. You can just wait until the quantum computer is ready and then you learn those secrets.

59:07And, you know, what if the model of quantum computing, what if people start thinking about it differently? Like right now you're saying they're trying to reduce the error rate as you add qubits. What if I don't care about the error rate? And I just add tens of thousands of qubits, knowing there's going to be lots of errors. And I just run repeatedly and have some way of confirming solutions. So the error rate could be huge, but I don't care. Yeah, so that will not scale well. It'll turn out that just the number of ways for errors to occur will just accumulate to a point where it's just not going to work.

59:57And that's something we call the threshold theorem. There's an actual value of the error rate, which you have to get below in order to have any hope of making things work as your size grows. and what's really cool is actually we're we're below that threshold in a lot of the quant computer architectures now that was one of the really big things about the google announcement with the willow chip is they got below that threshold and many other companies are doing that so that's it's really exciting in in the something we've been looking to do for decades now and it's happened. And that's why there's a lot of excitement in the air.

1:00:41And for you guys, since you are a company and not just being theoreticians about it, do the Googles or IBMs, which have the money and resources to throw at this, do they scare you in terms of as competition? So I think, yeah, a lot of the companies that are developing QCs, you know, it's certainly an exciting space. Like at BTU, we're not currently developing quantum hardware. You know, we're focused on the quantum security side of things. And we've been building this intellectual property mode for a decade now and are a first mover public company in the space. And I've made two acquisitions in the past six months.

1:01:28And I think, yeah, we're kind of differentiated in terms of our talent and IP and market opportunity that we're seeing. But yeah, you're certainly seeing interest from all those companies and some of which we've been working with and see them as strategic partners. And let me ask you, I try to understand and look, I looked at kind of the quantum mechanics and math behind how qubits work and quantum computation happens. And it's really not understandable to someone who doesn't have a PhD in quantum mechanics, it seems to me. But do you think someone could develop a skill set as a quantum programmer or a programmer of quantum algorithms without necessarily understanding all the underlying quantum mechanics underneath i you know it is it is a talent war out there um but i i think yeah on the software side certainly you you could get away with not i'm kind of doing a phd in in the in the field um for sure but i think you're you're gonna see um a lot of attention on just like talent in this space, just given how hard it is to find good talent.

1:02:52But yeah, I mean, you can do like software engineering similar to computer science. I mean, you have the PhDs that are devising the very complex technology and algorithms, and then you have the software engineers that are higher level. Right, like you could be the best programmer, you can be the best programmer in the world of classical computing without understanding the nuts and bolts of how a computer works, how a semiconductor works, how a transistor works. I'm wondering if the same thing could occur for quantum programming. Yeah, absolutely.

1:03:31We use these abstract models to represent things. And the reason it took so long for people to come up with quantum computing is because it looks too simple. You know, like every single qubit, you wouldn't think of as just a zero or one back, you know, like in the 1930s or 40s. You would have to think about where exactly this atom is sitting and all the electrons and protons. And it would be a very complicated object, you know, would take, you know, tons of blackboards to describe what that atom is doing. But because of all the experimental progress we've had, we're able to simplify it and abstract it as just the electron is in one state or another, and that's a zero or a one.

1:04:19And same thing when kids will learn about quantum computers, they're not going to have to dig into all the physical details. They'll be able to think about it more as a game. And I think it's really exciting, especially with AI, which can be a tremendous teaching tool where you can use sort of natural language queries. You don't have to be hyper pedantic in how you ask a question and get a reasonable answer. So yeah, I'm actually really optimistic that we're going to start to change the ways we think to be more quantum. It'll be just part of school education. You know, you guys should write, just for like even like marketing purposes, you guys Guys should write a book, Quantum Computing 101, just to explain what it is, how people are going to start using it, how people could program in it, what sort of problems can be solved now, your version of the quantum wars law, like all these things.

1:05:22You should make a name for yourselves doing the Quantum Computing 101 book. We have a lot of plans on the marketing and communication side. So we'll definitely keep in the loop on that, James. That's great. And now I'm going to ask the dumbest question of all, just to close this out. So given that one can't understand kind of the, let's say, location of a particle until it's observed, how did the Big Bang happen?

1:05:55Oh, just an easy question here. Please tell me.

1:06:03uh yeah uh i mean these are the big bang is is is a a way that is a consistent description of the the physics we know um and uh and it it it is a way to describe you know how the universe is expanding, why we get a certain ratio of matter to antimatter, and all these big questions. But why it happens, well, there are probably some people that have better ideas than I do, but I think these are still big mysteries. All right. Well, at that, you have answered a lot of my questions. I feel like my understanding is a lot greater. I also want to add to anybody listening to this, if you want to tweet out questions that maybe we could answer later, what's the Twitter account of BTQ?

1:07:10I think it's just BTQ underscore tech. So yeah, BTQ.com is the website. And what's your stock symbol? Yeah, just BTQ or BTQQF in the US. Yeah, because you trade in Canada, right? Yeah. Well, Nick, Gavin, thank you so much for your time. I really appreciate it. I feel like I finally am getting a grasp of what this is, what the timeline is, and why there is a particular urgency now for companies like yours that are doing solutions as opposed to building the actual computers themselves. So I appreciate you spending the time to come on here. And should I be pessimistic or optimistic about the universe, given that quantum computing will eventually happen?

1:08:02I feel like we're already too late to save things. Optimistic. Optimistic, yeah. All right. I'm there with you. Good luck, guys. Thank you for your service.

1:08:21Thank you.