The Quantum Reality: Risks, Market Pressures, and Ecosystem Challenges

Um, my name's Andrew Lord. I'm leading, uh, VTS Optical and Quantum, uh, research. And, uh, I'm on the fourth section here of, of a tour through Quantum and trying to analyze, uh, its impact on telecoms.

And this final section is really short, and it's just me saying, what on earth? What are the risks? What are we gonna do?

Uh, what, what would it look like, uh, and what are the, the things we haven't solved yet? And I'm gonna start by showing you, um, a picture of an Andrew Lord, uh, in 1981. Um, and if you look at me now, you can see what 40 years of quantum physics has done.

Um, now Andrew Lord, in 1981 was coloring in, um, computer punch cards as, as that one at the top there. Have you ever seen those? You're too young, aren't you, you guys?

But this was how you accessed a computer because our county in where I lived in the UK only had one computer. It was too expensive for more than one, and it wasn't in my town, it was in a university somewhere. So I had to color these punch cards in.

I had to post my program to the computer, someone fed them in, and the results were printed and they posted back, and I got the results a week later. Um, can you see the analogy? Because currently this is how quantum computers are being used, um, D-Wave and others are giving or selling access to their quantum computers in a very similar way.

You buy time, you might buy a minute, um, you go into a web server, you, you enter what you want to do, you send it, it's provisioned, it's scheduled, it's run. The results come back and you, and it tells you how much time you've used up maybe three seconds or five seconds or something, and you get, um, time by, you know, you buy time very much like these large language models are, are selling their, their services as well. Um, so my question really is, is, is that how it's gonna be?

Are these quantum computers always gonna be so expensive, so rare, so infrequent that this is the only way I'm gonna get access to them? Um, and if that's the case, um, what does, what does a telecom operator do, but also a role because, um, all you need to do is go into a web server and, and enter some, some commands or some questions. Now, when I speak to a lot of people that they counter this with a common thesis, which is this, they say, Andrew, um, we don't know what quantum computers are really for.

What we do know is that it's important that everybody has access. And when you speak to governments, they say the same thing. That their job is to develop the technology for the benefit of everybody.

And they come back to what I said in section two. Who is everybody in this case? Who is the Andrew Lord in 1981?

Who is that person sitting programming quantum computers? Is it a high school student or is that unrealistic? Is this too difficult?

Um, that's, that's an existential question I think around how the technology will be used, but also what impact it would have on the overall network and companies like bt. So this is my real slide to think, and I guess maybe ask you guys some questions in the last few minutes. Do, do you think, um, everyone will want access your, your children, you know, your, your, your high high school kids doing science well, they want to benefit from it.

Um, currently there's already a massive program in mainly in academia of people developing quantum computer software tools. I have a PhD student doing exactly this. Um, so, so there are groups of very bright people figuring out how to write programs to run on quantum computers.

Um, but this isn't like a, a XX 81 or a a a spectrum or, or a PC that I would have at home at all. This is pretty tough, very hard stuff. Um, I, I think the people that are most likely to use quantum computers will be the owners of the computers themselves.

Imagine you built a computer costing billions and then someone else uses it and generates a game changer for the, for the planet and makes trillions, and all you did was make the computer and spend a fortune so they could do that. Does that make sense? I have an issue with that whole use case.

I think there's gonna be deep tech teams out there that are trying to find new materials, but I'm not sure what the, uh, the process is whereby, um, money is made in the right places. Um, there's this whole question of blind computing where actually it's possible you have a quantum computer and you don't know what people are using it for. Is that ethically found that right?

Um, would, would a quantum computer owner that's literally spent billions of VC funding, uh, want to hand that capability over to someone without knowing what they've done with the risk that they probably potentially changed the world and you would, you wouldn't ever know, um, and, and that you certainly wouldn't benefit other than from their access payment, which might be peanuts. So yeah, open question, open discussion. I've got no, no answers, but I'm very interested to, to know what business views are and maybe we can look at anana analogies from other industries of how this is kind of evolving path.

So my final slide then, um, to date, um, just really summarizing, essentially, do we think the punch card 1981 approach to classical computing is sufficient now? Um, I don't know. Um, is blind computing feasible or is it non-ethical?

And do we need better policing of this whole industry? And do we need to get that policing in place before it happens? Unlike ai, which was probably came pretty quickly, didn't it for us, caught us out.

Quantum computing, should they be networked or kept separate? Should they all be co-located in a nice part of the country where the energy is low? Um, and then do you kind of drive there if you want to access it?

Is that what we're talking about? And is then a national scale just impossible in terms of the distances? Um, or do we end up with kind of networks of centers of excellence dotted around with, you know, with some kind of network between them?

Um, and then there's a whole load of technological, technological advancements that we could hope for around quantum memories. We've seen holo or fiber already, and I'm very hopeful about that one. Um, so maybe the next 10 years we'll see a, you know, a revolution in some of those building blocks, which will make it easier.

Um, but I'm very interested in, in the thoughts of people around if somebody invents a new game, changing material, do they own it? Does the quantum computer that that invented it own it? How does, how does that work in the future?

And, and ultimately I don't, I'm not entirely convinced. I understand BTS and other telecom operators role in this. Uh, if we're just providing the internet access, there's, there's nothing there for us that's just data.

If we are providing, um, a trusting access to our customers who want to access, you know, compute resource, but dunno how to and need a, a controlled way of doing that, that then maybe we do have a role. And, um, I have a fi the final side is really just, um, to put out there that if people want to talk to me, uh, I want to communicate with me on the subject. Uh, I'm looking for collaborations always, because, you know, you cannot do this on your own.

Um, really appreciate the talk and we, uh, then the time and the questions have been fantastic so far, but we have seven minutes or more. So let me know what you think. Thank you.

Uh, I'll actually stand up here and kind of mc a little bit, uh, as we mm-hmm. Have a few more minutes here. Um, Scott, I actually do want to bring up something that you mentioned earlier when we talked about the fact that it seems to take forever for, uh, NIST to approve, uh, new quantum encryption.

Uh, I would actually say that's a benefit because as we found in the selection process for di lithium and crystals and everything else, we actually had a candidate key fall out because it turns out that it was stupidly easy to factor, uh, given advances in, uh, computing horsepower that have developed since the original candidate keys were put up. And so the laggard approach actually saved us from having to rekey everything. I also think it's very interesting that a lot of companies are already starting to create future proof quantum encryption schemes and integrate them into what they're already doing as a methodology for, um, effectively moving past RSA, at least for highly secured workloads.

Um, I'll, I'll throw that out to you guys. Would you even notice if someone started using a different encryption key scheme? Would it matter to you in the long run?

Unless you were doing something extremely specific? It's been happening for years. So this is, that's nothing new, like new encryption schemes getting pushed into production.

You Mean you're still not running triple des, Right? Let me, let me back up. And, um, as someone who does a lot of work with the federal government, my comment on time lag through NIST approvals, uh, it was not stated with judgment, it was just stated as it is a time lag.

And you'll notice I tried to put emphasis on the even longer time lag it takes to get things into deployed systems once it's actually been vetted. I mean, Andrew, what are your thoughts on this? Do you, do you feel like, um, What, go ahead.

What I've, um, noticed since the, um, NIST competition kind of reached, uh, a mature point, the proliferation in teams around the world trying to break existing, um, that, you know, the winners, the, the lattice based scheme, it's taken off and we have, we've heard various reports, the people claiming they've broken it that turns out they haven't. But it's now, it's now to be pot shot, isn't it? It is a target, yeah.

And so, you know, it is not gonna stop. And, and the problem is, um, I have a complex algorithm that I think is com computationally difficult to crack, but I can't prove that it's impossible to crack. And that's, that's an impossible proof.

Um, so, so, you know, we always have this threat, this risk that somebody somewhere can break into it and then there's back doors as well. So I, I am, to me it makes sense that if you really want us to secure your data, you probably want to lock the doors, lock the windows, lock everything else, and, and, and, and do more than just one layer of protection. And, and you know, we, some of our customers say that to us.

Um, I mean, PQC, having said that, it's still a big hope. Know we don't want it to be broken. Of course we don't.

I'm just saying we, we should be a bit more, um, yeah, flexible in, in how we, how we view this landscape. And, and I think that that's the key that everybody has to understand is that on a long enough timeline with enough resources, everything is crackable. It's just a, you know, every house is you, you can rob any house no matter what, how advanced the security system is.

It's just a matter of how much effort you wanna put into doing it. And, and that's actually one of the dangers of having this data just laying around is anything that we encrypted, you know, or here's a good example. Shaw one hashes were supposedly, you know, going to be very difficult to create collisions.

And now it is easier than you might expect. But I mean, I, who knows, like r remember for classical computers for a long time, you know, uh, prime, uh, uh, public key cryptography was almost impossible to break. And sneakers was a fantasy until it wasn't.

Uh, John, did you have a question? Well, I'm Wondering about the, you know, everything in a whole, we start looking at this, obviously quantum computing is gonna be useful for something mm-hmm. Pieces that we hit on today that I had not been thinking about at all.

Uh, it was combine the co-location and the amount of loss we've got on things. Um, lots of different things we've gone through co-location, electrical, you know, years. A hundred years ago you went to a specific place to run your electric, um, equipment.

Um, 50 years ago we were going specific places to do computing. Nowadays we think it's everywhere. Our basic networking is based on the idea of average power.

When you start shooting a laser down someplace, you lose some power to it. You get a chance to recover that. When you lose the photons, you lose the information that's there.

The amount of loss there is staggering. And I didn't think about that really before this. And so I'm just kind of curious how long it's gonna take us to be able to move this type of data, any reasonable distance, and what that's gonna really mean to networking from the standpoint of, you know, we have computing, obviously, like I said earlier, we're gonna have computing that is useful into the quantum world, but how far are we gonna be able to move it?

And I, I just, I don't, I don't think I understand enough of this to understand that at all, that I'm just starting to, you know, get a glimpse of it from today's talk. But we are agreed totally, but we're already seeing, um, both in the US and in Europe, um, metro quantum entanglement demonstrations over a distance of, of few kilometers that it's not impossible. But, but I agree with you that, that, you know, national scale is gonna be very difficult Currently.

Exactly. Yeah. And things changed so quickly, huh?

Wasn't that long ago that we thought a hundred meters was as far as we can move data. Well, if we're talking about using free space optics for satellite quantum key distribution, you can use, uh, free space optics horizontally. So maybe that'll be part of how we break barriers.