WBD556 Audio Transcription

The Quantum Threat to Bitcoin Revisited with Richard Murray

Release date: Monday 19th September

Note: the following is a transcription of my interview with Richard Murray. I have reviewed the transcription but if you find any mistakes, please feel free to email me. You can listen to the original recording here.

Richard Murray is the co-founder and CEO of ORCA Computing. In this interview, we discuss the spooky and baffling effects of quantum mechanics, how ORCA is harnessing these effects to build quantum computers, and why success will be our generation's moonshot.

“On the encryption side a lot of people are worried about ‘how long do you want your secrets to remain secret for?’. And if you’re really worried about things staying secret for a long time, there are some things that people won’t want to get out for the next 100 years, then that’s the reason for people moving on to other types of secure platforms today.”

— Richard Murray

Interview Transcription

Peter McCormack: Morning, Richard, how are you?

Richard Murray: Yeah, great thank you, how are you?

Peter McCormack: Not bad.  You've probably had the shortest journey of anyone we're recording with in this next couple of weeks, so that's nice.  Yeah, thanks for coming over.  I know you've been talking to my brother quite a bit.  In the world of Bitcoin, we probably get an email, well we get lots of emails, people asking different questions, "Can you make a show about this; can you make a show about that?"  But every couple of months we get one where it's a concerned email saying, "But what about quantum computing, is this going to destroy Bitcoin?"  So, we made a show about this, what, three years ago?

Danny Knowles: Yeah, it was a good while ago.

Peter McCormack: Yeah, it was a good while ago.  It's time to revisit it, people don't really always go back to the old shows.

Richard Murray: Yeah, there's a lot happening, a lot going on.

Peter McCormack: Well, I try and follow it.  I read Focus and I read New Scientist, and I see the odd update on what's happening in the world of quantum, and I'm aware that it's like a space race.

Richard Murray: Yeah, and I suppose from the outside, it's difficult to follow, because there's so much science, and quite techy science going on as well, quite hard to differentiate what's science and interesting, but what's actually going to make it in terms of it being a useful product in the near term.  So, there's just a ton of conversation happening, so it's quite difficult to figure out what's real and what's not so real.

Peter McCormack: So, I really want to get into as much detail as much as possible with you, really nerd out on this.  But most of the people coming on our show are people from the Bitcoin world, so when people see a show coming out they're like, "I know that person".  People may not know you, so it would be good to do your background, why you're the buy we're talking to about this.

Richard Murray: Yeah, where do I start?  Pretty much everyone in this field of quantum computing has started with a PhD, so it is a deeply technical field.  So, I did that, went to university, got the PhD in quantum physics, but that was back in 2010 when quantum was just a weird fundamental science thing.  In fact, I still remember my supervisor.  I wrote loads of stuff on how to make it a commercial product, and my supervisor at the time just redlined through it and he just wrote, "This isn't scientifically interesting".  So, I pretty early on worked out that I wasn't a scientist, and I'm not really a scientist, and I spent the rest of my career being an entrepreneur businessperson.

So, I left the PhD in quantum to be a consultant, so I did sort of cool, technical-related consulting projects.  I don't know if your audience are interested in this; I helped Oakley work out new types of optical films that go on sunglasses.

Peter McCormack: Because they're based in, is it Letchworth?

Richard Murray: Yeah.

Peter McCormack: Well, they were until they sold out.

Richard Murray: Yeah, all over the place.

Peter McCormack: Yeah, but they had a head office up that way.

Richard Murray: Did they?  Oakley?

Peter McCormack: Oakley sunglasses?

Richard Murray: Yeah.

Peter McCormack: The only reason I know that, you know my friend, Sam, he used to work for them.  So, he moved up to either Hitchin and I think he worked in Letchworth, but then they ended up selling out to that huge sunglasses brand.

Richard Murray: Well, so this was in California, so got flown over to see -- if you've ever seen pictures of.

Peter McCormack: That sounds better, yeah, much better!

Richard Murray: I'll tell you, it's the funnest meeting I've ever had, turning up at Oakley.  I mean, their building just looks like a spaceship from the outside, it's unbelievable.

Peter McCormack: So, what were you helping them with?

Richard Murray: Looking at nanostructured coatings, and they were looking for them to go on the outside of Oakley sunglasses.

Peter McCormack: What, is that like a protective coating?

Richard Murray: Yeah, for polarising, for filtering colours out, and all sorts of weird and whacky -- they just love looking at the future of sunglasses, and it all related to fundamental physics, which is where I came in.

Peter McCormack: What is nanostructure?

Richard Murray: So, this was looking at -- this is going to be a bit of a distraction.

Peter McCormack: Yeah, fine.

Richard Murray: If you look at a moth's eye, bear with me, if you look at a moth's eye under a microscope, it has all this really weird structure.  And the structure is so small that it gives optical effects, like it makes the colour difference.  So, there's colour that comes from the fact that this surface has structure, has a form.  And so, Oakley were looking at basically printing this moth's eye structure on the surface of their sunglasses; I don't know if that's lost you.

Peter McCormack: Nice.  I'm with you.

Richard Murray: What was really cool is that their vision was that you have different types of colour filters for different types of sport.  So, if you're playing golf, you need a blue filter to help the ball look easier to see; you're surfing and you want a different type of filter.

Peter McCormack: To see the sharks!

Richard Murray: Oil on the surface, I don't know.

Peter McCormack: Danny, you're a surfer, what do you need?

Danny Knowles: I don't want to see any sharks!

Richard Murray: So, I did a lot of cool, science stuff, but always translating it into business; that's where I carved a bit of a niche, I suppose, like being not really a scientist, but a bit of a scientist, but really applying it to future products and things.  I spent a bit of time in government.  So, midway through my career, about 2013, everyone suddenly realised that quantum was a thing, so there was lots of academic research that all of a sudden, quantum computing all of a sudden looked like it was possible and real.  And I think around the world at the same time, everyone suddenly realised, "Hang on a minute, this might actually be a thing that we should start looking at more seriously".

So, I worked in government a bit to run different programmes investing in companies looking to get into quantum, and then eventually, in 2019, had enough of government, always wanted to start my own company, and then met two other really awesome academics, and they were looking for someone to basically come in and run the business.  The rest is history.  We formed ORCA Computing.

Peter McCormack: So, what does ORCA do?

Richard Murray: So, stop me if I get a bit technical.

Peter McCormack: Okay.

Richard Murray: We build quantum computers, so we build new hardware that we think people will plug into their server systems to accelerate --

Peter McCormack: You build quantum computers?

Richard Murray: Yes!

Peter McCormack: I thought quantum computers were still in the research phase, still to be proven, still to be stable?

Richard Murray: No, I mean we build quantum computers.  The things is, you might say they're still in a research phase because they're not big enough yet to solve really powerful problems.

Peter McCormack: Right, so the breaking of cryptography?

Richard Murray: Yeah, right.  So, our systems have about 10 qubits, quantum bits --

Peter McCormack: Yeah, we'll come into that.

Richard Murray: To break any type of encryption, you're going to need millions or billions of qubits.  So, that's why we're still maybe in the research phase, but why everyone's starting to get interested about it.

Peter McCormack: Okay, so that's already something I didn't know.  My awareness of quantum computing is that in China, they're building a quantum computer, and they've had some breakthroughs.  And is there one in France and Germany?  I'm not sure.  But that's my surface-level awareness.  So for me, it's like quantum computing is something that's coming down the line if these issues of stability can be solved.  I didn't realise there were actual quantum computers now out there and being used.

Richard Murray: Oh, yeah, definitely.

Peter McCormack: Did you?

Danny Knowles: No, mind blown!

Richard Murray: They exist; we've sold two of them already!  So, we've been in the news, we were on the BBC for selling our first quantum computing system a few months ago to the UK Ministry of Defence.  So, there's real activity happening.  So, I was just saying earlier, it feels a lot like The Imitation Game, so the origins of computing, because these systems aren't very big, they're not very powerful, so they're a bit like Alan Turing's machine back in the 1930s; they might be good for certain things, but you've got to really take a deep dive into the problem, I don't know if you remember The Imitation Game?

Peter McCormack: I do, yeah.

Richard Murray: So, if you imagine he's running that big machine, it's not solving anything; you're basically running out of time.  I think they'd got a 24-hour window to solve the problem, and every 24 hours it fails to solve the problem and they start over again, that's the sort of theme.  The way they got over that was eventually, they found some structure I think in the message they were trying to decipher.  They worked out that the first word is always going to be, "Hello" or, "Heil Hitler!" or something like this.  So, they worked out some structure in the problem that made the complexity a lot simpler.  So, they found a way, a neat trick to incorporate what they knew about the problem to make it a lot simpler.

That's sort of what's happening in quantum at the moment.  So, the computers we've got, the quantum computers we've got at the moment, they're not powerful enough to just solve every problem, they're not running quantum spreadsheets, or anything like that.  But what they are potentially being used for, experts come in, they analyse the data, they see certain types of structure, they work out that the problem looks a certain type of way; and then, that reduces the resources you need, the power of the quantum computer you need, to be able to solve problems sooner than you might need if you hadn't have done that.

Peter McCormack: Okay, before I ask you about that, just a quick fun fact for you, Danny, you might not know this; so, anyone listening, the Turing machine was in Bletchley Park, which is 30 minutes from here, where we are now in Bedford.  Hut 8 Mining, I'm pretty sure, we'll have to factcheck this or Jaime Leverton will kill me if I've got this wrong; but Hut 8 I think is the name of the building which the Turing machine is in.

Danny Knowles: That's very cool.

Peter McCormack: Yeah, I'm pretty sure it is, because she told me that.  She said, if she comes over, we have to take a visit.

Danny Knowles: That's very cool.

Peter McCormack: Yeah, and Andrew Poelstra also wants to go there.  Andrew Poelstra wanted to organise us to make a show in Bletchley Park, but they wouldn't allow us.

Richard Murray: I can't believe people haven't heard as much as I thought about Bletchley Park.

Peter McCormack: Well, I don't want to be rude, but the more nerdier types will have, because all the programmers that I've talked to in Bitcoin, they've all heard of it.  Danny is a surf, beer-monster, Man-U-supporting neanderthal!

Danny Knowles: That's my new bio!

Peter McCormack: And also, he was born a lot -- he's never seen Back to The Future; fact.

Richard Murray: That's not true?

Danny Knowles: No, it is true.

Peter McCormack: He's never seen Back to The Future.

Richard Murray: Please change that tomorrow!

Peter McCormack: We're going to have a film night tonight.

Danny Knowles: Retro film night.

Peter McCormack: Retro film night.  So, you can't rely on Danny for anything, apart from being the best producer in the world.  Okay, wow, so I've got questions I didn't know I was going to have because of this.  Okay, but how much does one of these cost?

Richard Murray: It ranges, so most of them cost tens of millions of dollars.

Peter McCormack: But yours?

Richard Murray: Well, what we're trying to do is do it a bit differently.  We're a plucky, British start-up, trying to reuse a lot of telecoms optical components that already exist.  So, our system, I won't say exactly how much --

Peter McCormack: Give me a range.

Richard Murray: Less than $1 million for the first systems, to allow people to explore.  So, still on the punchy --

Danny Knowles: And what's the physical footprint of them; how big are they?

Richard Murray: So, all I'll talk about is what people generally think a quantum computer looks like, and maybe your audience have seen, you get these golden cylinders, they're all cryocooled, so they're at super-low temperatures, and they need all of that to work.  Those are the systems that cost tens of millions of dollars.  Governments might be interested in that, but not maybe most businesses.

Our system looks quite different, so it costs a lot less than that.  What we're trying to do is, because our technology is different, we use light, we use photonics as a platform for quantum computing, rather than electrons and electronics, so it makes it different, our system is sort of a rack-mounted -- we try and make our system look like a server.

Peter McCormack: Do you call it a system, not a computer?

Richard Murray: Yeah, because I think -- I mean, maybe that's a bit of a technicality, just because it's trying to solve -- I mean, it's a computer, yeah.

Peter McCormack: Okay!

Richard Murray: So, we really try to push the message that quantum computers don't need to look like they're from space, which is how -- a lot of quantum computers look a certain way, because that's how they need to look to work in the way they do.  But a lot of people in the normal computing industry can't really see that working.  A lot of your audience, I don't know, they'll have server racks, they'll be mining Bitcoin and all this type of stuff, they'll know what a computer looks like and a quantum computer looks just completely different.  In our view, it doesn't need to.

So, our system, we've been deliberately trying to make our computer look similar to what's already out there, so that people can just plug it into their server, they can use it to try a few problems.  It just looks -- and people can work with it in the same way that they might work with just a normal server unit.

Peter McCormack: Are there many competitors out there making similar sized, let's say smaller business, not huge millions of qubits?

Richard Murray: Yeah, there's a mixture.  I mean in total, there aren't that many quantum computing companies.  There's probably 30 to 50 maybe worldwide, and they cover new types of software, and they also cover hardware, so there's not that many companies out there doing quantum computing altogether.  In terms of the types of companies that are similar to our stuff, it's very rare.  There's hardly any, probably a handful, maybe three to five companies who are trying to move beyond the $25 million machine, to build systems that are cheaper and more accessible.

Peter McCormack: But there are $25 million machines out there that can be bought, which aren't the kind of big Chinese -- there's sort of an intermediate?

Richard Murray: Oh, yeah.  So D-Wave is probably the first quantum computing company.  Any of your audience who've looked at quantum computing will probably have come across D-Wave.  Their systems cost $25 million, we believe.  They've sold a few; they sold one to NASA, they sold one to some US Defence agency, I think they sold one to Google.  But they're those companies who can afford to just take a massive bet on just one machine.

I guess in our view, for quantum computing to be as useful as everyone thinks it will be, like game-changing, it needs to be out there, it needs to be, I hate to use the word, but democratised; it needs to be accessible to people who are running normal servers and just need quantum on the side to give them a bit of a performance boost.

Peter McCormack: Yeah, so that was leading to my next question.  Why would you buy one of these?  You can give me a fake scenario so you don't expose a client, but give me an example of a client who would want one, where this would fit within their computing infrastructure, and what problem this is solving that other computers can't.

Richard Murray: Most of our clients are all working at the forefront of what you can do with a classical computer, so they're all companies that own massive data centres themselves.  There's one client in the energy industry that just owns one of the world's biggest supercomputers.  So, for their own purposes, they identity oil refineries and things like this.  So, those companies are quite good, because they already know what's possible and what's not possible with even the best classical computers; they're at the real forefront of what you can do.  So, they talk to us, they're interested in working with quantum computing to identify certain parts of what they're trying to do that can be accelerated with a quantum computer.

In terms of what an interaction looks like, it's pretty exploratory, to be honest, because no one really knows.  So, we can talk a bit in a while about future threats to encryption.  There are some things out there that everyone knows about, a quantum computer will be used for those types of problems; but those types of problems, like we talked about, need millions of billions of qubits, so they're not systems that will be available any time soon. 

The really interesting space at the moment, the sort of arms race, if you like, is taking these systems that aren't anywhere near as big as you might need to break encryption, or anything like that, they're much, much smaller, but they can be applied; you can find structure in a problem to apply them for short-term applications.  But to find those things, there's a lot of exploration.  We have to go through a ton of…

A big customer we've got in the energy industry threw at us loads of optimisation use cases.  So, they've got a big supply chain network, they've got oil tankers moving around, they've got goods moving between different ports, and all of that type of problem is at the limit of what you can do with a normal computer.  You basically have to put up with a good guess and you can't do a lot better than that.

Peter McCormack: What are the normal computers failing on; is it latency; is it number of things it can process simultaneously?

Richard Murray: Mostly it comes down to the complexity, so the number of parameters, but deeper than that.  Quantum computers are really good at very highly complex problems where there are a load of parameters, and all of those parameters are interlinked with each other.  So, take a portfolio.  One part of my portfolio might change and that will have a knock-on effect on many other areas of my portfolio; so problems that are very complex, there are lots of interlinking dependencies in your data. 

That type of super-rich data is really hard to solve with a normal computer, because you can imagine a normal computer just likes to go through, it has one parameter, it changes that, moves onto the next parameter changes that.  A system where if you change one parameter, everything else changes, normal computers really struggle to solve, because they like to do things linearly, like check one thing, move onto the next thing.  So, quantum computers in comparison have this weird ability to be able to solve many parameters at the same time.  I mean, if you want the physics…?

Peter McCormack: Yeah, I want everything, literally everything.  And just before we go into that, is quantum computer, quantum mechanics, are these all part of the same; the weird quantum world, is it all the same thing?

Richard Murray: Yeah, so if you want to go into the depths of it…?

Peter McCormack: Hell, yeah!

Richard Murray: So, quantum mechanics, quantum physics is like the new dawn of physics, so around the time of Einstein, these other guys like Dirac and Heisenberg --

Peter McCormack: Didn't Einstein say it was weird; he was freaked out by it?

Richard Murray: Einstein famously didn't really believe it.  There's this thing called hidden-variable theory, which says that all this quantum stuff isn't really happening, you just don't know enough about the system; it's just basically a good guess about stuff you don't know.  So famously, Einstein didn't believe in quantum.  But weirdly, because he didn't believe in it, he actually helped move the field forward.  He came up with a few experiments that he wanted people to do to prove that it was really real, and people did them and showed that Einstein was wrong.  So weirdly, through Einstein being wrong, he helped to show that quantum was real and helped move it forward.

Peter McCormack: What was that famous quote of his?  Find that quote of Einstein's.  He had that famous quote with regards to quantum, something about --

Richard Murray: Is it, "God doesn't play dice", or something?

Peter McCormack: No, that was something else.

Danny Knowles: Einstein said that, "If quantum mechanics were correct, then the world would be crazy"; is that the one?

Peter McCormack: No, there's another one.  He just said about it being weird, he just doesn't understand it.  I can't remember.

Richard Murray: I will say, when you get into the details of quantum, when I give lectures and stuff on this, I say to people, "If you think you understand this, you haven't understood it.  None of you should be sitting there going, 'Oh, yeah, that makes complete sense'".

Peter McCormack: And, is quantum a thing whereby people have figured out what it does, but there's no logic to why it works; or are they actually figuring out the logic?

Richard Murray: No, you get to the depths of quantum and at one point, you're into what makes up our Universe.  People don't know how quantum works, they just know that it does work.  It's been proven with tons of experiments.

Peter McCormack: Is this because we're just in a simulation and it's part of the supercomputer running us?!

Richard Murray: I will say, we've had a few moments in the office with a few beers where one person starts up and like, "What's really happening?" and we go down that road of philosophy and working out what the Universe is made up of, and stuff.  By the way, we try and not do that when we're trying to build a quantum computer, because you just get way too distracted by all the interesting stuff.

Peter McCormack: I think it's amazing.  I would do work experience there if I could.  This sounds fascinating.  Did you find the quote?  I'll dig it out later.

Richard Murray: He just didn't like how random it was and the fact that you don't know.  So, quantum all relies on probabilities, and I could try and give you a --

Peter McCormack: Shall we do a 101 on the physics, as much as you can do?

Richard Murray: Yeah, I can have a go.

Peter McCormack: Explain it like we're three-and-a-half!

Richard Murray: As long as you promise to interrupt me and tell me I'm not making sense.

Peter McCormack: I've built a career off it!

Richard Murray: So, the core thing about quantum is that, one thing that you don't think should be able to do two things at once can do two things at once.  I've probably lost you already.

Peter McCormack: No.

Richard Murray: I always describe it as light switches.  So, people in electronics, they all know about binary circuits; things are on or off, like light switches.  We all know a light switch can be on or off.  That's the normal world we live; things are binary, 1 or 0.  In quantum, the weird thing about everything quantum is you engineer these systems that can be 1 and 0 at the same time.  And this is the bit where, I'm giving a lecture on this and people will nod their head like, "Yeah, I understand".  You really shouldn't understand that something can be 1 and 0. 

So, I always describe it, if you're imagining a light switch and you're flicking the light switch backwards and forwards, or you're trying to hold the light switch in between, you're not really allowed to do that.  It should be as weird as imagining me to say, "There's a room and there's two exits to that room.  You as a person pick one exit or another exit.  That makes sense to us, that's binary, that's the world we live.  In quantum, if you were a quantum thing, you'd be able to leave through both exits".

Peter McCormack: At the same time?

Richard Murray: At the same time.

Peter McCormack: I've seen that, because that's the light slit experiment.

Richard Murray: Yeah, exactly.  So, that's the basis of all things quantum, so the fact that we use single photons, so single particles of light, and we do the Young's Double-Slit, we do the light thing where things --

Peter McCormack: Do you know about this, Danny?

Danny Knowles: Yeah, I remember reading about it.  I mean, I don't understand it, but I --

Peter McCormack: I don't understand it.

Richard Murray: Well, the main point is you really shouldn't.  None of us understand this, none of us take a look at this and go, "Oh, yeah, it makes complete sense.  Light goes through in two different directions at the same time".  All of us physicists working on this stuff, we all know that the light has to go one way or the other, it has to go out one door or the other door.  Whenever anyone tries to think about going out of both doors, because these things are not divisible, so what we're doing, the reason Einstein didn't like it, it doesn't make sense.

So, we take these single photons --

Peter McCormack: What is a photon?

Richard Murray: It's a single unit of light.  So, if you imagine lightbulbs, lasers, everything like that, that we think of when we think of light, it's all made up of trillions and trillions of photons, so single particles flying at us.  And this is now in the other weird quantum world of, maybe people have been taught in their upbringing about light being a wave; the other way of looking at it is a particle.  So, just imagine every lightbulb is emitting loads of these fundamental particles, these things called photons.

What we do in our company is just isolate one of those single particles, and you promise to interrupt me if this is starting to be not useful.

Peter McCormack: No, I'm fine, you're good.

Richard Murray: So, we isolate just one of those particles, and this thing cannot be divided, it can't go in two directions at once; it has to pick what classically --

Peter McCormack: So, when photons are hitting us, it's trillions of individual ones?

Richard Murray: Yeah, absolutely.

Peter McCormack: Okay, and how do you extract a single photon and hold do you take hold of that?

Richard Murray: Weirdly, I get asked this question all the time.  I mean, this is pretty much what our business is made up of; we spend most of our time just doing this.  You basically just take a really well-controlled lightbulb, you know exactly how many photons it's produced, and you just do a ton of filtering.  I guess that's probably the easiest way to describe it.

Peter McCormack: Until you just get one?

Richard Murray: Until you're just left with one.

Peter McCormack: How do you know that one has shot out; do you see it?

Richard Murray: Well no, because we'll get into the middle of weird quantum stuff.  You're not allowed to look at the photon, because if you look at it, it destroys all of the quantum effect.  What I was going to get onto is, when you send a single photon at a mirror, so imagine this mirror can reflect half of the photons and it let's through half of the photons, the photon doesn't do one or the other.  Remember, I keep going on about this, it can't be divided, this photon, it has to pick one or the other, or at least so we thought. 

But instead, the light, the single photon, becomes a quantum thing where it does both.  So, it both gets reflected and it goes straight through the mirror at the same time.  So, now it's a quantum thing, so basically our computer's built of loads of those operations where one photon gets split millions of different ways into this big maze-like thing.  It's still one photon, but one photon has been basically split up to cover all paths in your maze, if that makes sense.

Peter McCormack: So, has the photon replicated itself?

Richard Murray: No.  This is what I mean.  If you really understand this stuff, and I will say, any of your audience, you should not be sitting there going, "Yeah, it makes complete sense".  This photon is still one photon, it's not been replicated, it's one photon.  But because it's quantum mechanical, it can take loads of paths at the same time.

Peter McCormack: How do you know it's taking those paths?

Richard Murray: That's a bit more subtle.  This is back to the Einstein, him coming up with an experiment to prove it.  Basically, there are different outcomes, because if the photon is taking loads of different paths rather than just one path, it can do really weird things.  It can interfere with itself, so two possible paths of the photon might, at a later time -- imagine this big maze, the photon's passing through this big maze, sometimes two different paths of the maze will come together.  Then, they'll interfere with each other and produce an effect that you don't get if the photon is just going through the maze one at a time.

Peter McCormack: How was quantum mechanics even -- how did people figure it out, or what happened that made someone go, "Do you know what, there's some weird shit going on here?"

Richard Murray: Back at the turn of the century, so around the 1900s, there were just a lot of experiments that didn't really make sense.  People tried, a bit like the double-slit experiments, there's an experiment you can do which makes sense when you've got lots of light.  So, people tried to pass light through these two slits at the same time and you get this weird pattern.  And then this experiment, people slowly dialled down the light power, until they just had one photon, and at this point they were like, "Okay, the light will choose to go through one door or the other one", and they still observed the same effect as when they had lots of photons going through two doors. 

Peter McCormack: Okay.

Richard Murray: I've probably lost you.

Peter McCormack: No, you're fine.

Richard Murray: Anyway, there were loads of experiments that didn't make sense, and around the 1900s, Einstein and all those types of people suddenly went, "Hang on a minute, what we think we know doesn't work", and this whole new theory emerged to try and explain what they were observing, what they were looking at.  And that's what people -- in the field, people describe this as quantum 1.0, and all of those effects get built into how people design semiconductor chips.  So, you can't design a modern-day computer, a semiconductor, without understanding quantum.

Peter McCormack: Hold on; because…?

Richard Murray: Because it describes how all electrons pass through materials, it describes all semiconductors.  You can't really describe a transistor, the fundamental thing in a computer, without understanding quantum.

But what I was going to get onto, so now turning to quantum computing, so all of that stuff has existed for 100 years.  People have known that this really weird physics is there.  But slowly, over the last 100 years, people have been able to get better and better at actually making these quantum effects, so actually creating these pieces of light that are in two places at once.  So, that's the start of a quantum computer.

So, if we take an electron or a photon that's in two places at once, that's a single-qubit quantum computer that you can start using.  And now, we're at the stage of having, in our company, we've got a 10-qubit machine, so we've got 10 single photons, and all of those are in two places at once, or more places at once; and IBM have got 100-qubit machines, so all of a sudden they've got hundreds, in their case, of electrons all exhibiting quantum effects.

Peter McCormack: Yeah, I'm going with you for the moment.

Danny Knowles: And why does observing that break the quantum effect of it?

Richard Murray: I don't know if there's a good way to describe that, sorry!  Other than to say, all this stuff, it just can't be observed.

Peter McCormack: You mean, when you say, "Can't be observed", in that if you were looking at it, your eyes are what destroys it?

Richard Murray: Yeah.

Peter McCormack: But if you had your back turned, it would be fine?

Richard Murray: Yeah!

Peter McCormack: That's fucking weird, man!  I feel like I need a joint and I don't even smoke weed!

Richard Murray: Yeah, that's why a lot of physicists look a bit spaced out when they talk about this!

Peter McCormack: Hold on, but it's only if a human looks at it?  Or a camera?

Richard Murray: Do you know, no one really knows.  I mean, this is now at the limits of philosophy.  This is what I mean about no one really understanding, what is this?

Peter McCormack: But hold on, if no one's looking at it, it works, because your computing works, your quantum computing works?

Richard Murray: Yeah.

Peter McCormack: But if you were looking at the photon in your quantum computer, your quantum computer wouldn't work?

Richard Murray: Correct, yeah.

Peter McCormack: But if a camera looked at it and not a human?

Richard Murray: Yeah, it's as good as it still destroys the experiment.

Peter McCormack: So, even a dog, if a dog was looking at it, it would destroy it?

Richard Murray: I don't know; we don't allow animals in our lab!

Peter McCormack: I mean, to me, that's an obvious test.

Richard Murray: Do you know, someone must have done that experiment.

Peter McCormack: But even now, I'm like, "What do you mean, if I look at it, it destroys it?"  Is it because the photon wants to come to me?

Richard Murray: I'll be honest and just say, this is a whole massive area of philosophy that --

Peter McCormack: But hold on, I can't put philosophy and computing together, because computing to me is maths; it either does its thing or it doesn't.  Philosophy is like figuring this shit out.

Richard Murray: So, what I mean is that when we're building our quantum computers, we deliberately just don't think about all this stuff, we just know all this stuff works.  There are loads of rules that tell you how it works.

Peter McCormack: We don't know why!

Richard Murray: No, we don't know why.  I mean, the same -- actually, I'm trying to think, I guess there are loads of comparisons in companies where you don't really know what's going on, but you don't care, because you're just trying to build a product.

Peter McCormack: That's my whole life!

Richard Murray: Well, that's the same as what we're trying to do.  We're taking advantage of this 100 years of research of how this stuff works, without worrying so much about why it works, and that's good, because we can still build computers from all this stuff.  There's loads of weird stuff that happens.

Peter McCormack: How does that get turned into a computer?  What is the part of the computer that you're replacing?

Richard Murray: I mean, sort of fundamentally, it changes everything.  It changes the very concept of computing as we know it.  I mean, the nice comparison is, all computers are based on binary, they're all 1s and 0s, and quantum is just a whole new paradigm, a whole new way of building a computer that doesn't rely on 1s and 0s.  So, it doesn't use existing programming methods, it doesn't run Excel, it doesn't run Microsoft; it's a completely different way of looking at a computer.  Because it's not based on 1s and 0s, it's not based on how all other computers are built; it's a fundamental rethink about how you build a computer.

Peter McCormack: And why, what is the fundamental change you're getting here?  Because, 1s and 0s are sequential, so everything has to happen in a logical, sequential timeline; but in a quantum computer, things can happen at the same time?

Richard Murray: Yeah, that's right.  So, trying to describe it, a normal computer, I don't know, you're trying to add up two numbers.  So, you've got a 1 plus a 0 gives you a 1.  That's one thing you can do with 2 bits of a normal computer.  With a quantum computer, you can still do that, you can still have a quantum 1 and you can still have a quantum 0; but you can also have 1 qubit being 50% 1 and 50% 0, and the second qubit, the second bit of information being 25% 1 and 75% 0.  So, it gives you another, I don't know, another parameter.

Danny Knowles: It's almost like another dimension of it?

Richard Murray: Yeah, that's probably the best way to look at it.  It just gives you an extra, yeah, dimension of your data, so it makes every bit -- it gives every bit in your system more combinations of what's possible.  So, you've got 2 bits, I'm going to fail doing this, there are 4 ways you can arrange 2 bits.  With a quantum computer, there's still 4, but as the number grows, the number of different ways you can arrange 2, 3, 4 qubits is much greater than the number of ways you can arrange 2, 3, 4 bits.  That's getting distracted by the physics side of things. 

I'm just trying to describe it as completely -- you shouldn't really look at a quantum computer and say, "How does it compare with a normal computer?"  It doesn't really replace a normal computer, it's something completely different.  I mean, back to our customers we work with to try and look at this stuff, we spend half of our time asking them to come up with things that they had tried to do with their normal computer, but had just forgotten a long, long time ago, because it was just nowhere near possible.

Peter McCormack: Will there be a time in the future where our personal computers will be quantum computers?

Richard Murray: I don't think so.

Peter McCormack: Because, this is a hack for specific goals that you want to achieve?

Richard Murray: Yeah.

Peter McCormack: You found a way of creating things that you can process and solve problems in a different way?

Richard Murray: That's right, yeah.  And if any of your audience are into hardware, probably not many of them are, but there are GPUs, which started life, if you're a gamer -- GPUs started life, because it helps you run computer simulations, helps you render 3D objects.  But people worked out that these really specific pieces of hardware were quite good for --

Peter McCormack: Mining cryptocurrency!

Richard Murray: -- mining cryptocurrency, and doing one very specific operation, which is multiplying matrices.  That spawned a whole Nvidia, I forget what they are, $200 billion company just selling hardware that just does one thing really, really well.

Peter McCormack: Are you building a quantum ASIC?

Richard Murray: No, we are not.

Peter McCormack: Can we talk after the show?  I've got an idea; I've got a way of making us a lot of money!  Have you looked into this, because that is one specific task?

Richard Murray: Yeah, it's really interesting.  I guess on that side of things, most quantum hardware isn't -- we use ASICs to drive our hardware, but it's not the quantum hardware we use, because the hardware we use is just so completely different from everything that exists currently; it's different.  What's cool about this space is that basically everyone's looking again at how you solve computing problems.  Because they're looking at things, how you do it with a quantum machine, they're just looking at how you might do it completely differently, and it comes up with -- in fact, Fujitsu have got a really cool project working with ASICs.  It's not quantum at all.  They spent a bit of time looking at quantum, and they learnt a new way to drive ASICs.

Peter McCormack: Are we talking about different things with ASICs?  Are you talking about generalised ASICs?  I'm talking about Bitcoin mining ASICs.

Richard Murray: I'm talking about generalised ASICs.

Peter McCormack: So, another thing I learnt today; ASIC is not a Bitcoin term, it is a computing term.

Danny Knowles: Did you not know that?

Peter McCormack: No, I didn't know that.

Danny Knowles: Okay.

Peter McCormack: It's Application Specific… I mean, I guess it's kind of obvious.  But we have ASIC miners; can you grab one up?  Have you ever seen one of these ASIC miners?  So, companies like Bitmain, they produce them.  They're up to an S19, is their latest machine.  They have one job, okay.  So, their job is to mine Bitcoin, and they keep running essentially a random-number generator, essentially, to try and find the specific hash to find Bitcoin.  I say it like an idiot.  But that's all they do, and they try and do a gazillion attempts over and over, and once one finds it, a block gets found, and then we move onto the next block in the blockchain.  Can you build one of these for me, and how much, and when can I have it?!

Richard Murray: I'll tell you what we would do --

Peter McCormack: And, can I have exclusive rights?!

Richard Murray: I'll tell you what we would do; if you were a customer of us, or any of the other quantum computing companies, we'd take a look at the problem you're trying to solve, which is mining Bitcoin and what that looks like, and we would try and design our own quantum ASIC to solve that problem, taking advantage of all the quantum stuff; we'd design a quantum programme that helps you look at whether you can mine Bitcoin faster, more efficiently.

Peter McCormack: So, based on what I've just told you, there is enough there for you to go, "This is something we could look at"?

Richard Murray: Maybe, yeah.

Danny Knowles: The price of one of these is, what, $15,000?

Peter McCormack: Well, even less now, like $5,000 in the market.

Danny Knowles: There you go.

Peter McCormack: But if the quantum's so good, you might only need one.

Richard Murray: If anyone was serious about this, what we would do is we'd get a team.  So, we have a team of really excellent quantum people, who are also pretty good on the machine learning side, and we give them that problem, and it takes probably about three months to have a first look, see, to work out if there's anything there.

Peter McCormack: I am serious!  What do you charge?  No, I mean I have a feeling you might have people reaching out to you after this.  Do you charge people for that three-month research?

Richard Murray: It depends.  I mean, we're trying to run a business.  So, you can imagine what it's like.  We try and work out how strategic we think this is, how likely it will lead to something that's going to help us sell hardware and run a business, and we take a judgement on that.

Peter McCormack: What do you think, Danny?

Danny Knowles: I mean, if it's something that can work alongside a traditional -- like, could you have a quantum element of a Bitcoin mining ASIC?

Richard Murray: Yeah.  So, what I'll say is, I'm sure if there are any quantum people listening in the room, they might be screaming at me like "It's not possible!"

Danny Knowles: They'll be screaming at us, don't worry.

Peter McCormack: Somebody must have already looked at this anyway.

Danny Knowles: Yeah, I'm sure.

Richard Murray: People probably have, yeah.  But the field is changing all the time.  People are learning a lot more every day about what can be done, what's possible and what's not possible with machines today.

Peter McCormack: N-ossible?  You've created a new word!

Danny Knowles: I actually call that efficiencies!

Richard Murray: Yeah!  So, yeah, you don't know, and I'm pretty excited, going back to your point, the route to try and do something with these machines in the short term, bearing in mind the best machine's got 100 qubits, is to combine it with what you're trying to do already.  So, just imagine palming off very small parts of your problem to a quantum machine; so maintaining the current ASIC, whatever you're trying to do normally, but just maybe, I don't know, your computer programmers have worked out that one very small part of that problem is really, really hard, or the core of the problem that means that it takes longer than your bottleneck.  And the state of the art in this industry is looking at, if we solve that particular small part of the problem on the quantum system, can you do it faster?

So, that's the Holy Grail, that's the sort of really exciting short-term race in quantum computing at the moment; put high-end classical ASIC, or whatever classical computer, side-by-side with a quantum machine, allow them to pass information super-fast between each other; does that provide any benefit?  The answer so far, to be honest, has been no, so this stuff isn't providing any -- but there are applications emerging that look interesting enough.

By the way, as soon as anyone shows this, as soon as someone says, "Okay, this is something that is possible in the short term", that's it, you're already on the way to demonstrating value with a quantum computer.

Peter McCormack: If you solve this on the mining front, you present the risk then on being able to hack private keys, don't you, because it's a similar-ish risk?

Danny Knowles: I mean, you're asking me; I think you're asking the wrong person!

Peter McCormack: Yeah.  But then, I guess it could just speed up mining, it could just give you an edge?

Richard Murray: Yeah.

Peter McCormack: If the miners, they're essentially just spinning out numbers and trying to find a random one which essentially allows them to unlock the block, and I don't know what -- it's terahashes per second they talk about, isn't it?

Danny Knowles: Yeah, I don't know what we're at at the moment.

Peter McCormack: It might be a case, if you were doing this, you just speed it up.  So, you could make an ASIC, I don't know, you could add a 10%, 20%, 30% edge to all the other ASICs, which probably wouldn't justify the costs.

Richard Murray: Possibly, and this is the world that quantum is in at the moment.  And to be honest, a lot of our customers, the way it works best is if they throw at us loads of computational problems and we go, "No, no, none of these are possible.  That one looks possible, none of these are possible".  Or, they might be possible, but they would take a machine that's much larger than we've got today. 

Actually, to the point about decryption, that's something that everyone talks about quantum computing being useful for, the machines that are needed to break RSA encryptions, or even your hashing-type problems, you need billions of qubits.  So, you know, I've been talking about machines that have hundreds of qubits.

Peter McCormack: You've gone from 1 to 10.  Does it go in factors of 10, or are we just constantly talking about that?

Richard Murray: So, that's currently the thing that's debated and researched the most at the moment.  So, within ORCA, or within most companies, they're aiming at breakthroughs that give you a step change in the number of qubits that you might have.  So, everyone knows that at the moment, you're going from 1 qubit to 2 to 3 to, I don't know, to 64 to 128.  That's still a world away from billions of qubits or millions of qubits at the point where you're starting to solve some really juicy algorithms.

Peter McCormack: Going from 1 to 10, what is the complexity from getting from 1 to 10?

Richard Murray: I mean, on that side, these things are all hard, but 1 to 10 is fairly straightforward; a lot of companies have done that now.

Peter McCormack: What did you solve to do that though; what is the complex problem that's being solved?

Richard Murray: So for us, it all relates to the hardware, the machines we're building.  To be honest, we're just trying to make very high-quality, single photons, so what I talked about earlier.  For us, the reason why it's hard to get to 10 is sort of related to how easy it is to make single photons.  So, it's like an engineering challenge, "Can we make a single photon?  Yes.  Okay, can we scale that up to make 10 single photons at the same time?" it's that type of problem.

Peter McCormack: And when you say "at the same time", is it 10 different sources of photons, or is it the one source creating 10 photons?

Richard Murray: Yeah, 10 different sources of photons. 

Peter McCormack: And are those 10 photons going around the same maze, or a separate maze each?

Richard Murray: So, all of those photons will go through the same maze, but they'll all interfere with each other, they'll all overlap.  So, you know the problem when I described it earlier, I talked about 1 photon can take lots of different paths through the same maze?

Peter McCormack: Yeah.

Richard Murray: Imagine how much more complicated it can get if you then have 2 photons, and both of those photons can take every path through the maze.  But every time they meet, those two paths, something weird and quantum happens to make it even more complicated.  And now imagine 10 photons going through that same maze.  Sorry if I've lost you!

Peter McCormack: No, you're fine.

Richard Murray: Every one of those photons could take every path through the maze.  But every single time any of those photons meet, there's an interaction, it changes.  So, all of those 10 photons are all interdependent with each other.  And now imagine that you can programme that system, then that becomes the heart of a completely new type of building a computer, which is a quantum computer.

Peter McCormack: And how do they carry information, and how is information extracted from this?

Richard Murray: Good question.  So, you insert them into the system, so you've got a bit of information encoded in how and when you insert your single photons into this big maze system; you can change the maze, so it's a bit like, what's that movie?  Inception, I think it is, that really cheesy movie where they're building mazes and things.  Half of this job is to basically build -- so the maze basically, in an abstract way, kind of represents the calculation you're trying to go.  I've probably walked past you now.

Peter McCormack: No, no.

Richard Murray: Anyway, so you can design the maze, you can change the maze, you can change the maze in real time, and that maze structure basically reflects the problem that you're trying to solve.  And this is like deep, deep maths I'm trying to describe now.

Peter McCormack: Yeah.

Richard Murray: And then, what you basically do is you measure the outcome of the maze, so you can see where the photons appear, and at that point you can measure them, which stops them being quantum, and all of a sudden you can basically find out the very exotic and weird many paths that all of these photons have taken through the maze.  And if that's completely lost you, I'll just say that represents the way to carry out a calculation.  So basically, you can make your maze look like, I don't know, a Bitcoin mining operation, and the outcome of the maze -- if you successfully navigate your way through the maze, it represents a correct solution to mining a piece of Bitcoin.

Peter McCormack: Okay.  Yeah, I mean I barely understand most of what you're saying, but at the same time I think I'm take a slight step forward; okay, I think I understand a bit more.  So therefore, to go to 100 qubits is a factor of 10 complication from 10?

Richard Murray: Yeah.

Peter McCormack: And, sorry, I'm just trying to figure this out in my head.  Can calculations go wrong?  1s and 0s are pretty logical, it's binary.  Because this isn't 1s and 0s, can calculations go wrong?

Richard Murray: Yes.  Actually, that's one of the major challenges of quantum computing.  Because everything's basically analogue, it's not 1 or 0, which is noise-free; it's in between in this weird state, you end up introducing a load of noise, and that's the main challenge facing quantum computing at the moment.

Peter McCormack: Accuracy?

Richard Murray: Accuracy, and the fact that the bigger the system gets, like if you go from 100 qubits to 1 billion, there's going to be loads of noise that you're going to add into that system that you can't correct.  So, because you can't look at this thing, you can't correct for noise the way you would do with a normal computer, so you just have to do it in one shot, and you have to make your noise low enough; or you have to do something like error correction, which is a whole other field in itself.  So, you have to do all this stuff in one shot without looking at it, and that is the toughest -- that's the reason a 1 billion qubit quantum computer doesn't exist today.

Peter McCormack: Somebody, I can't remember, somebody on our show once said, "It's probably unlikely there will ever be one.  It is too complicated, there are too many issues to be solved that it's very unlikely, certainly in our lifetime".

Richard Murray: I'm a bit more of an optimist than that.  I think it's not without risk.  People describe it as the Moon Landing of our generation.  I mean, these systems are so complicated, it's that type of big deal.  I mean, when we landed a person on the moon, it took a huge engineering feat to get them there and return them safely.  Building a quantum computer requires the same level of engineering, probably the same amount of money as it required to put a man on the Moon.  But at the same time, if you could build a quantum computer like that, then you could do some tremendous things that just aren't even possible today, like you could find cures to cancer, you could break encryption.

Peter McCormack: That's not a good thing.

Richard Murray: Well, okay, yeah.

Peter McCormack: I think it's a bit like Bitcoin; we were talking about Bitcoin, it's censorship resistant money, it's freedom money.  Every time somebody talks about a criminal or a terrorist using Bitcoin, we talk about an activist or somebody living under authoritarian regime, who has the benefits from it.  So, there's this scenario where you can see the good side and the bad side.  I think the breaking of encryption is basically bad, because people need privacy.  But I think the curing of cancer is obviously good. 

But that's quite a bold statement to say, "Having a quantum computer could cure cancer".  Is it because of the volume of research it can do?  Why is that?

Richard Murray: Well, I guess it was maybe a bit too far of a step to make.

Peter McCormack: Yeah, but you've got to stand by that now!

Richard Murray: You're going to hold me to, "Where's my cure to cancer?" in two years' time.

Peter McCormack: "Hitchin scientist cures cancer with machine"!

Richard Murray: I guess what I mean is, when you get into the depths of trying to design a new drug, so if you're trying to design a new cure for cancer, and by the way I don't do this, day job's busy enough as it is trying to build a quantum computer!  Yeah, people tell us you're trying to design a new drug, that's quite a complicated thing, it's probably quite a big molecule, it's working inside of your body, it's probably quite complicated to work out what's going on there.

So, the way that we get around that at the moment is, we do a reasonable job of working out what it will do on a computer, and then you basically feed it to a human.  So, you do a lot of trial and error in clinical trials, by actually trying out drugs and seeing how they work, and things like that.  That makes it quite hard for us to work out, are there new molecules, are there new drugs out there that we haven't even begun to explore, because they're so different from everything else out there that we just would never have thought of it, and we certainly wouldn't have been confident enough to give it to a human.  And that all stems to the fact that it's very hard to model big molecules, big systems, using existing computers.

But quantum computers are much better at doing that.  So, the reason why you find it hard to model a big molecule is --

Peter McCormack: Lots of variables.

Richard Murray: Lots of variables.  It's a really quantum-mechanical system.  If you've done chemistry, if anyone cares, there are lots of interactions between molecules, it's all quantum stuff.  So, a quantum computer is much better at modelling all of those many variables, all of those quantum effects, so you can imagine a day where you can design a drug from start to finish just inside a computer, and then you can therefore hit a button and say, "Okay, this last drug did this one thing.  Let's just explore, just using a computer, how a drug will work and what types of things it might do".

Obviously then, the leap to curing cancer is a bit of a big one, but if you've got that type of ability to just design drugs from scratch to be solving certain problems, it just enhances drug discovery no end, and all the rest of it.  So, yeah, no one's identified a cure for cancer that you would be able to solve if you had a quantum computer, but you could still imagine if you had that type of power, you'd be able to do that type of really impressive calculation.

Peter McCormack: How much synergy is there between the world of AI development and quantum computing, and is it something that will eventually come together as kind of a combined science?

Richard Murray: I mean, the jury's out at the moment.  When I mentioned about quantum computing having lots of near-term applications, mostly what people talk about is applying quantum systems to machine learning and artificial intelligence.  So, there's quite a lot of active research at the moment.  It will take one breakthrough to finally find the computing system that can be applied to machine learning.  But it actually is quite easy to compare a quantum computer; it's quite easy to imagine it being used for machine learning, even though no one's had that big leap to show that a quantum system could provide speed up.

So, the whole field is wrapped up in this quest, if you like, to show that a quantum system is better than what's already out there, like a classical Nvidia GPU, or something like that.  No one's actually shown that speed up yet, but there's a lot of potential for it.

Peter McCormack: And, in the world of these massive, super, quantum computers, like I say, I've seen the big cylinders, I like the tubes coming off them.  They look like something out of the 1940s or something, it's weird!  But in that world, like I say, I'm very aware of the Chinese one, because I think everyone wants to know what they're up to, because they're probably very secretive.  Yeah, here we go.

Richard Murray: Yeah, I call it the Golden Chandelier!

Peter McCormack: It's actually quite beautiful really.  What is going on here; can you explain this image to us?

Richard Murray: Yeah.  So, I will say that the only quantum bit -- this is all just one big fridge.  So, a lot of quantum computers rely on being super-low temperatures, so close to what's called absolute zero, the lowest temperature you can go.

Peter McCormack: Because…?

Richard Murray: Because that basically isolates your quantum thing from everything else in the world. 

Peter McCormack: Your "quantum thing"?!

Richard Murray: Yeah.  Your qubits cannot interact with anything, they have to be perfectly isolated to preserve -- they can't be measured.  So basically the reason it needs to be supercooled is because, if it interacts with anything at room temperature, or anything else like that, that basically ruins the quantum state; it acts as a measurement, which destroys everything.

So, all this stuff needs to be cooled down to super-low temperatures, so most of what you see here is just one big fridge.  The only quantum bit, the part of the system which is quantum, right at the bottom of the picture, you can see there's a little square, there's a little red bit there; that is your quantum chip, that is the thing doing the quantum mechanical processing, that's the bit that contains your qubits.

Peter McCormack: Well, if you're spending tens of millions, they've rightly designed the rest of it to look interesting, because only you and a few of your friends will know that.  If they just showed you that little box, people would be, "Well, that's not fucking cool"!

Richard Murray: Yeah!  All the rest of it is feeding all of the information down.  So, all the vertical little coils you can see and all the cool little things, that's just basically your wiring feeding all the information down into your quantum bit just down at the bottom.

Peter McCormack: That is a very cool design.  So, anyone listening who's not seen the video, just go onto Google and search for "quantum computer" and have a look at these machines.  So, do we know whose machine this is, or do they all look like this?

Richard Murray: So, the title is, "First quantum computer to pack 100 qubits enters crowded race".  So, there are a few, IBM, Google, some of the big guys.  There's a company called Rigetti that's a start-up.  Those are the companies that have reached this 100-qubit-type point.

Peter McCormack: Danny, can you search up, "China quantum computer", because there's another one I've seen.

Richard Murray: Cool, so this is what our system looks like.  So, these are systems that use lights, so now all of a sudden you can't see that fridge system.

Peter McCormack: Remind us what they use instead?

Richard Murray: So, they use photonics, they use light, similar to what we do.

Peter McCormack: But you said before, there's another option, you could use something else instead of photons?

Richard Murray: You can use electrons.  So, the previous systems that we saw -- basically there's loads of different ways to try and do the same thing.  You can either rely on supercooled electronic circuits at really, really low temperatures, which is the picture we saw before.  That's what Google and IBM and those types of companies are doing.  This system relies on using light, so that's what this looks like, that's what our system looks like on the inside. 

The reason why it doesn't rely on super-low temperatures is that light, it doesn't really interact with anything in the same way that electronics do.  You can pass a laser beam across a room and it doesn't interact with anything.  So, it doesn't need to be cooled down like that, it can work at room temperature.

Peter McCormack: So, no one's trying to solve the 1,000, 10,000, 1 billion qubit thing right now; is it step changes?

Richard Murray: So, it's a combination.  There are a few companies looking at, "How do we get from 100 qubits to 200 qubits to 300?"  I think most companies also have an activity to look at how you get from 100 qubits to 10,000 qubits to 1 million qubits.  You've got to sort of work out both at the same time almost, so you've got to have a bit of a short-term product roadmap and then a longer-term one.

Peter McCormack: What type of qubit level does it get to the point where it starts to solve crazy problems that could never have been touched before?

Richard Murray: So, some quantum computers are starting to do that already.  So, there've been three demonstrations of what's called quantum supremacy, which is where a quantum computer does something that a normal supercomputer can't do.  There have been three so far, four actually; two by the Chinese, one by a company called Xanadu in Canada, and one by Google.  So, these have all built a quantum computer that's big enough that all of a sudden they can do something that a normal computer can't do.

The weird thing is, the problem is, that those problems at the moment aren't useful problems.  They're weird mathematical -- you've got to imagine that a lot of this is being driven by a bunch of mathematicians.  They've come up with a problem that's very specific that a quantum computer can do very, very quickly that a normal computer can't do.  The problem at the moment is that no one cares about that problem, it's not useful, but it's still a stepping stone to something that is useful.  

So, the race is on at the moment to be the first company or government, or however you want to do it, to show that link between a quantum computer doing something with a not-useful problem, and then to a quantum computer solving a problem that can't be done with a classical computer, which is also useful, if that makes sense.

Peter McCormack: Yeah, it does.  How much of the conversation amongst the community is with regards to the ethical side of this, and any potential dangers or risks of quantum computers being used for nefarious means?

Richard Murray: Yeah, it's big, I think.  I mean, I think we share your worries really about this being not that democratic, let's be honest.  At some levels, this is a sort of arms race happening, mostly between the US and China at the moment, so I think all of us are aware of that.  I mean, to be honest, the lucky thing for us is that all of the really good applications for these systems, rather than the bad ones, emerge first. 

So, the first quantum computers to be out there, they won't be useful for the really scary stuff, they'll be a world away for that; they won't be useful for breaking encryption.  They will be useful for machine learning and other types of problems, which more widely can be understood to be useful.  But I think what several companies, including ORCA, are starting to do is to get together and discuss the ethical side of quantum computing.

Danny Knowles: And there are quantum-resistant encryption methods, aren't there?

Richard Murray: Yeah.

Danny Knowles: Is that becoming a more widely used thing?

Richard Murray: Yeah, I think so.  I mean, I don't know if it's widely used at the moment; it's being researched.  So, there's a NIST, which is NSA, so various different US security agencies are publishing recommendations for these new algorithms, new encryption algorithms, and I think it's still hotly debated.  So for example, this recent publication by a US Government department, it didn't just give one solution, it gave five solutions, and what are called post-quantum algorithms; so, encryption algorithms that will survive beyond quantum computing.

So, I think in terms of people that are worried about security, they're not quite sure which solution to pick at the moment, because everyone's taking a look at how secure are these different alternatives.  Sometimes these algorithms that replace what's already there are a bit slower.  So maybe particularly in the case of what might be relevant for Bitcoin, obviously the speed in which you can encrypt your data and decrypt your data matters a lot.  And if someone could say to you, "This data is much more secure, because it's a better encryption algorithm, but it's going to take you 10 times, 100 times longer to encrypt, decrypt", obviously for a lot of customers, a lot of people using encryption, that's not very good.

So, there's this whole debate happening about what the right new platform, what the right new algorithm needs to be.

Peter McCormack: Yeah, I mean that's the biggest risk to Bitcoin.  I guess one of the risks with quantum computing and the risk to Bitcoin is that, you talk about these milestones being hit, quantum supremacy, but I would have no doubt there were certain governments, if they achieve certain milestones in their quantum development, they're not going to actually make that public, because there is a benefit to being able to use it for access to information, hacking, etc.  So, at the point where maybe SHA-256 can be broken, it might already have happened.

Richard Murray: Yeah, I mean no one knows, you're right.  I mean, China are investing a huge amount of money into this space.

Peter McCormack: Of course, as ever.

Richard Murray: I mean, they've got a strategy about building AI and also quantum alongside that.  Yeah, it's not possible to know if/when they'll have a system, or any of us will have a system that works and can do this type of thing; it's also not possible, in my view, to really know when a breakthrough might happen.  So really, any of us, any company, ORCA is looking at a few things that could very rapidly accelerate the number of qubits that we've got.  

So maybe not quite overnight, but pretty relatively quickly, all of a sudden you could have this new capability that no one has planned for before.  That's something that's for good, and also it's maybe not so good.  That's why a lot of companies are investing and looking at this now, because they want to prepare for the point where out of the blue, one of these systems that can be used for something suddenly emerges.  That's the really interesting side of things.

I will say as well, not so much on the SHA side and on the hashing side, but in terms of encrypted data, what you've got to worry about, and again this doesn't really relate to Bitcoin, where everything sort of turns over every ten minutes; but for normal encrypted data, some of that data might be being recorded today in an encrypted form, so that at a later date, when a quantum computer or some other system becomes available, that person can then look back at that data, which was encrypted, but now you've got a way of decrypting that data, and all of a sudden, an email that you sent ten years ago, or an encrypted file that you sent ten years ago, which has been recorded by someone out there, all of a sudden they've gone back to it and applied a quantum computer and decrypted that.

So, a lot of the argument on the encryption side, and this actually doesn't relate to vulnerability for Bitcoin so much, but on the encryption side, a lot of people are worried about how long you want your secrets to remain secret for.  And if you're really worried about things staying secret for a long time, there are some things that people won't want to get out for the next 100 years, then that's the reason for people moving onto other types of secure platforms today.

Peter McCormack: Yeah, I mean are you talking about things like messaging platforms as well, Signal, etc?  I mean, maybe we just need to go to that Black Mirror world where everything is public!

Danny Knowles: Let's hope not.

Peter McCormack: What are you hiding, Danny?

Danny Knowles: Everything.

Richard Murray: Yeah, I mean I don't want to paint too scary a picture; it is sort of cool.

Peter McCormack: It's exciting and scary.

Richard Murray: Yeah, and I don't want to dismiss it too much; that's true of any disruptive new technology.

Peter McCormack: Of course.

Richard Murray: And I will say, the thing that we are most excited about, the thing that I think adds the most benefit to society as the most democratic, is all of the stuff that will happen in the near term.  I think the ethical side of things is really important, and I'm not dismissing it, but we're forming views and we'll get onto that.  But in terms of all of this decryption worry, it's not something that's going to happen overnight, because even with a breakthrough, it takes a lot of engineering to go from 100 qubits to 1 billion.

Peter McCormack: And, is this something that if it is possible, it still won't even be within a decade; are we talking many decades?

Richard Murray: It depends who you talk to.  I mean, I believe that breakthroughs are happening and will continue to happen to make it a reality sooner than a decade.  But then, what do I mean by that?  I don't necessarily mean --

Peter McCormack: Well, for the people listening, if there is a risk to their Bitcoin, they want to know how long.  Is it a decade, five decades, that kind of thing?

Richard Murray: So, I think for the types of things we've talked about for Bitcoin, I do think you're talking about a ten-year, or longer, risk.  I honestly can't see a quantum computer emerging that can -- so, the secret is something that can run error correction.  That's what you need to support 1 billion qubits and therefore break SHA, or encryption.  I can't see that happening within the next ten years.

But you never know.  Particularly risk-averse people, I don't know if you've got a massive Bitcoin wallet and you're really worried about that, maybe you want to take steps, I don't know.  But I wouldn't worry, at least not now, for Bitcoin.

Peter McCormack: Okay.  Just another thing I wanted to ask you about with regards to quantum mechanics, it's completely random, but that thing I read about whereby you've got a particular in one place, quantum entanglement, and you turn that particle and there's another particle somewhere else that turns.  What the fuck is that all about?  Seriously!  And how do you know those two entangled?  And can there be two entangled from different ends of the Universe?

Richard Murray: Yeah, well I guess a good way to describe it is, you know when all the way back when I was talking about these photons, or whatever, being able to exist in two places at once?

Peter McCormack: Yeah.

Richard Murray: So, say your photon hits a mirror and it half gets reflected and then half doesn't, so it's in this quantum state, so half of that could go off to one side of the Universe, and half could go off to the other side of the Universe; they're linked.  So, if you do something to one of those, if you measure one of them, it either appears on one side of the Universe, or it appears on the other side of the Universe.  So, this quantum effect, if you like, is very long distance.

Peter McCormack: And it's not just photons, it could be electrons too?

Richard Murray: Yeah.  So actually, the best way to describe it is, it's anything that's super-isolated.  So, you basically want an object that's not in contact with anything, so it can't be bouncing off -- if it's a single atom -- so, there are ways of making qubits out of single atoms.  Those single atoms cannot interact, they can't collide with another single atom, because that basically acts as a measurement, like we described earlier, it destroys the quantum stuff.  Any quantum effect, any quantum mechanical effect will be destroyed if it comes in contact with a non-quantum thing. 

So, if it's a single atom and it collides with another atom, then it's destroyed.  If it's an electron, it has to be supercooled, because if it interacts with another atom or an electron, it will be destroyed.  If it's a photon, and photons are good because they don't interact very much, but if it interacts in a strange way with something else, the quantum effect will be destroyed.  So, this is why all of these platforms, they're all weird.  They're either supercooled, or they're in very high levels of vacuum, so you're sometimes talking about single electrons or ions in very high levels of vacuum.  All of this is just designed to have very, very isolated systems.

Peter McCormack: I think it might be quantum entanglement that Einstein talked about, and said it's super-weird, it doesn't make any sense.  In your beer sessions and you're discussing this, is anyone trying to understand, even though you say they can't, are people trying to figure out why this is happening? 

Richard Murray: Oh, yeah.

Peter McCormack: And have they made any progress?

Richard Murray: There are a bunch of -- am I allowed to -- I'll probably call them weird and whacky.

Peter McCormack: You can say anything you like.

Richard Murray: Well, I will say this point about, one of the main questions is when you measure these quantum objects and they become non-quantum, that doesn't make sense to any physicist, they don't get it.  And this spawns all of this thought about, I don't know if you've ever heard of multiverses?

Peter McCormack: Yes, I have, a lot about multiverse, yeah.

Richard Murray: So, physicists try and come up with answers to, "What on earth is going on?"  One answer might be, every time a quantum thing is measured, it doesn't stop being both things at once, because an extra Universe is created.  So, every time you're measuring something, before, it was both places at once, and you measure it and to us, it looks like it's become classical again, it's lost its quantum-ness, it becomes just one place again.  Maybe the answer to what on earth has happened there is the Universe has created a double, it's created a split-off, and the other outcome of that measurement exists in that other Universe, if that makes sense.

Peter McCormack: So, there's almost an infinite number of Universes being created constantly for every single scenario?

Richard Murray: That's what multiverses are, yeah.  Every time that anything makes a decision, so I mean I talk a lot about single photons or quantum objects choosing whether to be reflected off a mirror or transmitted through a mirror, each of those decisions creates another Universe.  The choice about whether you go to the shops or not, or stay at home and watch TV, that's another type of choice, that's another Universe that's created.

Peter McCormack: There's another Universe in the multiverse, Danny, where United aren't shit, where they didn't lose 4-0 to Brentford!

Richard Murray: Not many Universes though!

Peter McCormack: And they aren't bottom of the league!

Danny Knowles: Einstein said that quantum entanglement was, "A spooky action at a distance".

Peter McCormack: There you go, that was the quote I was looking for.

Richard Murray: Yeah.  I mean, it's sort of the right way to describe it, as much as I hate it.  And we did spend a lot of time talking about all this spooky stuff.  It is spooky, no one understands it.  To be honest, I try not to talk about all the spooky stuff, because it's interesting for sure, but when we try and build quantum computers, it's not useful, if that makes sense.  So, we try and avoid all the spooky stuff and try and focus on all the useful stuff.

Peter McCormack: Is it because it's a distraction, or because you --

Richard Murray: Yeah, because the team sit around, and if they've got a beer, all work stops and everyone scratches their heads and are like, "What on earth is happening?"

Peter McCormack: "What the fuck?"  Yeah, but that's the best bit, this spooky stuff.  Are there any other spooky things?

Richard Murray: Well, I will say there are a ton of things in our lives that we don't understand, one of which being consciousness.

Peter McCormack: I was about to say, women!  I mean, my daughter, God help me!

Richard Murray: I'm on mic, so I can't talk about mine!

Peter McCormack: Fuck, are we going to have to edit that one out?

Danny Knowles: I think I'd probably say so!

Peter McCormack: My brother would say, "Pete, you can't say that, you sound like Jeremy Clarkson"!

Richard Murray: Women may be one.  Consciousness, why any of us, we're all things, but we all have thoughts, we all have feelings, we all have a soul, whatever.  All of those things, we don't understand.

Peter McCormack: Something came out about consciousness this week; what was that?

Richard Murray: Yeah, I don't know.

Peter McCormack: Search Google News for "consciousness".  What was it I saw?

Danny Knowles: Maybe this?

Peter McCormack: Here we go, "Quantum Physics Could Finally Explain Consciousness"!

Richard Murray: There we go!

Peter McCormack: There we go!  I saw that this week.  I never actually read it, I just saw the headline.

Richard Murray: So, I would say the headline says it all, "Professor of philosophy weighs in".  Because, to be honest, at a certain point I don't really think it's the job of a physicist, because physicists are weird people, they don't get out of the lab a lot, they shouldn't really be thinking about consciousness, like what makes us people and things. 

But anyway, this is to say -- and by the way, quantum physicists hate me for talking about this, because no one really knows, but no one really understands how our brains work, and no one really understands how quantum physics work.  Maybe some of this stuff is related to --

Peter McCormack: Are our brains quantum?

Richard Murray: -- are our brains quantum?  There are some cool things in, so for example, I think starlings; some birds, they can't explain how they're so good at navigating.  So, they fly from one part of the world to another part.  They're amazingly good at being able to do that, in a way that if we had our own sensing devices, they're not as good as a bird, a sparrow, at doing that.  So, some people are thinking, maybe that's because, and I think there might even be some proof to say, inside of a bird's beak, I think it is, is a magnetic sensor that allows it to locate magnetic north; but it's a quantum sensor, it's something that relies on quantum mechanics to be better than any normal sensor could be. 

So, there's all this stuff; basically, at the edge of what we think is possible, quantum turns up.  And this comes back to, I guess, maybe the reason people think it's so powerful.

Peter McCormack: Can you scroll down a bit, Danny?  "The subatomic world.  Early quantum physicists noticed that the double --" oh, yeah, we talked about that.

Danny Knowles: This article has given us our title.

Peter McCormack: "Quantum Physics and Consciousness are Weird"!  It is all so weird.  Okay, what other weird stuff?  Let's just give some weird juice to the listeners.

Richard Murray: What else?

Peter McCormack: What blows your mind the most with this, when you think like, "That's just fucking insane"?

Richard Murray: To be honest, when I was a student and going through this the first time, we spent a lot of time with quite a few beers trying to work out this whole measurement thing, and we just came out with some crazy ideas.  We were like, "Well, hang on a minute --" and this is going to sound a bit weird but, "-- maybe there's another creature out there that's another level of consciousness".  So, we think we're making measurements, we think we're watching these quantum systems and quantum changes, because we're observing them.  Maybe there's another type of measurement, maybe there's another type of being.  We're just existing in what another being would come and look at and be like, "Well, that's weird".  Do you know what I mean?

Peter McCormack: The simulation, maybe.

Richard Murray: Well, I don't want to sound too weird, but maybe we are living in a simulation.  It sort of blows your mind to imagine that we're all living in this strange, quantum mechanical world, and none of us have a clue what it is even, let alone how it works.

Peter McCormack: So, the simulation stuff to me, I was like, "Yeah, it just sounds like bollocks", until, who was it who explained it?  Maybe it was Elon Musk who explained it.  He said, "But if you look at how far computers have advanced in 30 years.  Maybe 30 years ago, we first played SimCity and we could just build a road.  And now the crazy shit you can do with them".  But that's a short amount of human time span.  We've been here for thousands and thousands of years, the planets millions of years. 

Trying to imagine the advancement in 10,000 years in computing, you could see a scenario where you could have the ability, maybe with an ORCA machine, to create a simulation of the Earth.  And if that's possible, then the most likely outcome is that we are in a simulation.

Richard Murray: Absolutely, yeah.  We work with one bank and the guy we work with in the bank basically advises all of his incoming graduates.  He says, "If you don't learn how to programme a quantum computer now, you should start, because in the course of your career, it will change the face of computing.  So, make sure you learn how to programme a quantum computer, because at some point we're going to find that it's useful, and at some point it's going to completely dominate everything you're doing.  So, start learning now".

It's difficult to say exactly what's going to happen, but it is quite exciting.  I spend a lot of my time talking about how bad classical computers are at lots of things, like I don't know about you, I try and use Siri and I get really frustrated.  Siri's terrible.

Peter McCormack: You do that.  I've noticed Danny does that.  You book appointments and things.

Danny Knowles: Yeah, I make it remind me of things.

Richard Murray: Maybe your voice works better than mine!

Peter McCormack: "Remind me it's my wife's birthday!"

Danny Knowles: That's in there, honestly!

Richard Murray: Would you trust a driverless car?

Danny Knowles: I mean, I've been in a Tesla that's driven itself a little bit.

Richard Murray: And you were nervous?

Danny Knowles: Not really, to be honest.  It was on a highway.  I mean, it would be different if you were on a country lane, I guess.

Peter McCormack: I mean, you don't drive pilotless planes, essentially.  I mean, the pilot's there, but the autopilot --

Richard Murray: Yeah, but you still see a lot of these YouTube videos of Teslas just crashing through things that are made to look like people.

Peter McCormack: You also see videos of morons crashing!

Richard Murray: Oh, that's true.  Don't get me wrong, I love the thought of driverless cars.  I think often, you just come back to what it needs to implement AI to be actually good at driving a car, all the time, without any risks.  And commuters are quite bad -- I mean, it's amazing how far they've come to be able to do that.

Peter McCormack: But I think what it comes down to is, if there was an accident, you would rather blame yourself than a computer.

Richard Murray: Yeah, but maybe quantum computers could potentially offer much better AI, so give you much better trust in either the driverless car, or handing over trust to a driverless car.  It might be less likely to crash, or mistake a person as a shopping trolley, or something like that.  I hope it's not been too technical!

Peter McCormack: Okay, look, I'll level with you.  When you get into the details of how it works, I've got no idea what's going on, but I don't know really how computers work.  But honestly, this is my favourite interview that I've made in I can't even think how long.  I just find it fascinating, because it's weird, and yet it's useful.  And Einstein thinks it's spooky.

Richard Murray: Yeah, he didn't get it.

Peter McCormack: Fucking idiot!

Richard Murray: But that, what was it, 15-minute segment where we went in deep and you looked at me like I had two heads for a little while!

Peter McCormack: My confused face and interested face are exactly the same!

Richard Murray: Excellent, okay, that's good.  I didn't know that at the time.  Well I'll just say, if anyone doesn't -- you don't need to know, just trust us people building these systems.  Fast-forward past that 15 minutes, wait for us to build these awesome systems, and then give us time.  If you're interested, get involved. 

I do think there's an interesting future-proofing activity that people are doing today.  A lot of big FTSE-100 companies have all got a quantum strategy to prepare and plan for the time when these things are available.  And none of those senior execs get into the details of what multiverses look like, or anything like that, because they're just like, "Okay, show me the business advantage".  So, that's what we spend a lot of our time doing.  It makes my life less fun, I'll be honest, than this conversation, but maybe simpler.

Peter McCormack: Well, if people want to find out more, where do they go, tell them how to find out more about ORCA and yourself?

Richard Murray: So, contact us, we've got a website and things like that, you can just log on.  We're all over LinkedIn and things.

Peter McCormack: ORCA.com, ORCA something?

Richard Murray: ORCAcomputing.com, that's right.

Peter McCormack: And you're basically the Bill Gates of quantum computers?

Richard Murray: Yeah!

Peter McCormack: Hell, yeah.  Fucking let's go!  Man, I loved this, this was great, this was fascinating.  I wish you all the best.  I will be keeping an eye out for ORCA.  I'll be interested to see if you get a few people reach out and say, "Hey, can you build me an ASIC for mining?" let me know, because I'll be interested.  And if you can, come to me first.  Me and Danny are hot on this.

Danny Knowles: I don't know if we've got the budget!

Peter McCormack: We'll find the money.  No, all the best, man.  This was very cool, really appreciate it.  Good luck with everything you do, Richard.

Richard Murray: Loved it, thank you.