Blond & Quantum Episode 6: The Race Nobody Is Talking About | Bill Vass (Booz Allen)

Bill Vass, CTO of Booz Allen Hamilton, joins Eva on Blond & Quantum to separate signal from noise in quantum.

Bill built Amazon Braket. He runs thousands of engineers working on quantum, AI, and national security. And he has some surprising takes on what's real — and what's not.

Listen to find out:
📍Why a motion sensor and a ceiling light once broke a quantum machine
📍One material science breakthrough that quantum could unlock
📍The football field-sized machines nobody is talking about
📍What every CEO should do in the next 12 months
📍The story behind the first entanglement of two quantum computers across 100km of fiber

No PhD required. Just honest conversation.

🎙️ Guest: Bill Vass | CTO, Booz Allen Hamilton
🎧 Host: Eva | Blond & Quantum
 
If you enjoyed this conversation, don’t forget to like, subscribe, and share — more episodes with global leaders in quantum are coming soon.
 
#quantumcomputing  #boozallen  #deeptech  #quantumsensing  #BlondAndQuantum #BillVass #innovation #quantumtechnology

Transcript

SPEAKER_02 0:00

Hey, my name is Eva, and this is London Quantum, the podcast that breaks down quantum technology into real-world data. Here we make the complex simple. No equation, no everything, just insight, innovation, and a bit of a QR. So don't worry, you don't need a feature of the physics of the file. In each episode, I talk to funders, scientists, and investors about how quantum is shaping the industry today, but not in the sound distance of the future. Oh, and if you hear a black cat app pouring in the background, that's my coffee. A very alive cat joining the conversation. Her name is Mel. Let's get started. Hello everyone. Welcome to the next episode of Belong and Quantum. My cat let um stay at home, but today with me is a special guest, Bill Vass, a CTO of Booz Allen. Hello, Bill. Thank you so much for being uh here today afternoon with us.

SPEAKER_01 1:03

Yeah, thanks for having me.

SPEAKER_02 1:05

Bill, let's start from introduction about Booz Allen Hamilton. Please introduce the company before we go deeper into the Yeah.

SPEAKER_01 1:12

So Booz Allen is a high-tech company. We have about 22,000 engineers. We work on quantum computing, we run the GPS satellites, we work on hypersonics, we do a lot of cybersecurity work. We do, you know, a pretty broad range of things, everything from, say, a weapon system through an ERP system, mostly software engineers and some hardware engineers as well, and a lot of data scientists. Uh, we're the top provider of AI to the federal government. Um, and uh we're branching out and regrowing our commercial business, which we sold off a while ago. A lot of people know us for business consulting, but we sold that uh in 2008, and we haven't done that in a long time. And I uh came here after about 10 years at Amazon Web Services, where I was vice president of engineering and worked on quantum there. And of course, my team at Booz Allen does quantum uh for a lot of customers, and we do a lot of quantum operations and uh hardware design and and uh integrated systems and firmware.

SPEAKER_02 2:10

Fantastic. So you are having a perfectly a perfect background to be here with us and explain us a little bit more about what's going on in quantum today and also in the future. Bill, if if you think because as a Booz Allen, you sit between government, deep tech, and national security on this intersection. So if you think about quantum today, where would you point we are right now? Are we like, you know, is it still hype or is it sort of like pre-value, or maybe it's sort of the values are there and we are in pre-industrial?

SPEAKER_01 2:40

Yeah, well, it really depends upon which area in quantum you're talking about. So uh, for example, post-quantum encryption is something we're actively deploying uh for good reason because in the future, potentially with enough error-corrected qubits, you could break uh standard cryptography. And if you record network traffic now, secrets last a long time. We still have secrets for World War II, believe it or not. So that that's an important thing to maintain those secrets, right? And so uh if you're using post-quantum encryption, that will help to protect you in your transport layer. So we're recommending everyone do that. Now, a lot of bankers are doing it already. When I was at Amazon, we made it available in 2022. So it's been available for a long time, but not everyone's implemented it. So post-quantum encryption is really not quantum, it's math.

SPEAKER_02 3:23

Waiting for it to be solved. And actually, it's very interesting that you say about this because I wanted to ask you there's a lot of this hype that we talk about hey, uh, let's collect this data today and harvest first, right, and decrypt later. So is it a real threat, or do you think this is only like the narrative that you use to get the budget for RD?

SPEAKER_01 3:43

I think it's a real threat. I think it's it's it there's a lot of man-in-the-middle collect that occurs across the networks. And so uh, I mean, just think about it how easy it would be to collect cell phone traffic. You just turn on any collector at all, and you can you can collect information. It's all encrypted today, but later you could decrypt it. And so that that's the reason you want to start using it. But post-quantum is again, it's not really quantum, it's math that defends against quantum. The other areas of quantum we work in is quantum sensing, and that's very real today. Um, so the uh GPS satellites use an atomic clock, which is a quantum sensing equivalent, and we have to deal with actually relativity there and make changes to those clocks to keep them in sync. Quantum sensing and radio frequency, quantum gravitometry, quantum magnetics are all really advanced sensors that we're building and deploying and using. And they're they're all based on quantum. Some pretty amazing and exquisite quantum radio frequency operations and things like that that our team is working on that I'm I'm pretty amazed in. And so that's very real today, right? On quantum computing, of course, is where things are advancing quickly. Um that's where where a lot of this discussion, I'm sure, will be focused. And so, and you see a number of advancements occurring in sort of the four different areas of different types of quantum computers. There's uh electron-based quantum computers, which is the most common. That's what, say, I was building on my team was building at Amazon, Google, IBM, electromagnetic cryogenic machines. There's uh optical or photon-based quantum computers, there's neutral atom and and uh basically uh charged atoms or ion-based quantum computers. And those are the primary fork hides. There are others on the fringe, but those are the ones that are making the biggest advancements right now.

SPEAKER_02 5:27

And in terms of exploring the hardware, as you said, on different different angles, is Booz Alan uh actively exploring all of those angles?

SPEAKER_01 5:36

Yeah, so so we build all of those different types of qubits and prototype. We don't we're not trying to build a quantum computer. Um we are partnered with a company called Seek, and uh we're building the firmware for their hardware. So the biggest challenge in quantum computing is around error correction. Um and if you think of it on your iPhone or your laptop, you have an ECC because alpha particles do flip the bits on those. People don't think about that, but that's a reality of life. And so we use error correction code. But if you take a look at the amount of compute in a in an iPhone or a laptop or whatever, compared to the error correction, is very small. The error correction is very small compared to the amount of compute on a quantum computer because it's so affected that you're dealing these atomic particles are so affected by the environment that the amount of compute needed for for air correction is large, relatively speaking. And so what Seek does is they're working on building a hardware that runs inside the dilution fridge and inside the machines that does the air correction in real time. And that has to be very low power, especially in these cryogenic things, because you can't generate any heat, right? So they operate right inside the dilution fridge, for example, as opposed to a lot of systems, the wires come out, if you like, or that which are not really wires, a lot of them are uh microwave wave guides come out. That's how you control those and do the air correction outside of the dilution fridge. That makes the machines very, very large. And a lot of people don't think about it. Some of the first machines that'll have enough error corrected qubits to do operations that are of economic value at speed will be about the size of a football field.

SPEAKER_02 7:09

Wow. So you I thought we're going the opposite way to miniaturize the stuff.

SPEAKER_01 7:14

No, no, not for a while. You think about like the early computers like the ENIAC were very big, right? And now, you know, your iPhone is so much more powerful than in the U.S.

SPEAKER_02 7:23

No, that's true, but I have been a witness in some of the guests before in the podcast when they're showing me quantum computers in the size of small wardrobe. Yeah.

SPEAKER_01 7:32

So yeah, but not with enough air no they haven't gotten the air correction in there and all the other things. You look at some of these machines, they're gonna have to have seven million physical qubits to get a hundred air corrected qubits. Remember, because of that differentiation. But it depends on the machine, too. You have other machines that are more coherent, uh like the neutral atom or the ion tram machines, uh, but their challenge, of course, is speed of operation. Yeah, but you'd have to keep them in operation for very long periods of time to get an economic value of value of them. Right. I think you'll probably see some of the first uh economic value coming out of long-running operations on neutral atom machines, for example. Um the the if you take a look at the photonics machines or the cryogenic machines, they're gonna be quite large. In fact, that's what Seek is trying to avoid. Yeah.

SPEAKER_02 8:17

Right. It's trying to integrating that integrating into the internet.

SPEAKER_01 8:20

Okay.

SPEAKER_02 8:20

So, Bill, let me ask you through partnerships with Seek or other startups and companies in the quantum environment, do you I see do you move more into the role of being an integrator on um system control? Is this something intentional? Do you want to own the middle layer? Um, yeah.

SPEAKER_01 8:37

So so one of the things like we're we're we're writing the firmware for the embedded components in Seek as a product. So it's just, you know, we would generate our revenue as part of what Seek generates, for example. So we're we're doing more and more product. We just launched a whole series of cybersecurity products, for example. Um and so we're doing more and more product work. We'll still do system integration work as well, but uh more and more. And I I think like one of the areas that uh we're trying to work on is the development of a uh a generalized deployment of a Hamiltonian for material science for a quantum computer. Because to use a quantum computer, you have to understand how to basically create the circuits on the machine. Please explain us for for the benefits of the Well, I think you know the ever ever every type of computing in history has to have different uh advantages and disadvantages. I I think it's a it's an easy way to kind of think of how if you look at the history of computing, you know, we started with with our fingers, right? That's why everything is base 10, right? And you know what? Our fingers haven't gone away as a form of computing. You probably still go one, two, three, four, five, or count up the number, how many hours between this or whatever, right? So so in each advancement of computing, the other advancements didn't go away. And so we went from this to writing. Writing was a big advancement in computing. And then there was a giant technology leap called an abacus. You probably remember that. I mean, if you've seen uh uh people still use abacus today, it can be quite fast on an abacus. And then uh we moved to sort of gears and slides. You know, we with we went to the moon on a slide rule the first time.

SPEAKER_00 10:08

That's right. Right. All the computing was.

SPEAKER_01 10:10

Yeah, we had people called computers, that's where the name came from. And they used slide rules. And you know, the early uh um, you know, cryptographic braking machines were all gear-based, gear and motor-based. So and then we went from that to tubes we learned, and that's where the INEAC came in. And again, these are huge machines when you started to scale them up in the switching space, and then transistors and then integrated circuits, and now we're doing atomic particles. So it's it's kind of a natural progression. But each of them has advantages and disadvantage. A quantum computer is not going to replace your classical computer. You're not gonna run a website on a quantum computer. Well, well, but a quantum computer wouldn't be very good at it either. It's good at certain things that those are not, like a Hamiltonian of a large molecule or something like that, that you just physically can't run on these other machines.

SPEAKER_02 10:58

I think it would be waste of energy and everything, right? It's like shifting from arm rate to as well.

SPEAKER_01 11:02

Yeah, yeah, for that. So I think that that's gonna be some of those transitions that occur. But you know, the the I I think you'll use all of these things together, right? You don't like I said, you didn't stop counting on your fingers or writing because we have computers. And so uh you're gonna have these classical computers, and I think the quantum computers will be very much be co-processors for the classical computers. And that was one of the things we did when I was at Amazon with Brackett is we I remember uh Yeah, exactly.

SPEAKER_02 11:31

Let's talk about it because the bracket itself, it was the first time when access to quantum computers get popularized, right? On the huge scale. And you were part of it. So please, once you started talking about it, I would love to a little bit of your insight of the history of of of Bracket.

SPEAKER_01 11:48

Yeah, so the idea behind Bracket and and and uh uh Simone Severini, our uh PhD, was running our quantum on our team, uh really brilliant guy, and he came up with the name Brackett, which is Dirac's no case notation for quantum, right? It's not bracket misspelled. My my my spell checker, my wife, always try to correct it when I was reviewing stuff, but anyway. Um and it's uh we wanted to find a name without a cue in it. All the cue names are yeah, anyway. So the idea behind Bracket was to give, make it so anybody with just a credit card could write a quantum circuit and run it on a real quantum computer. And as these machines continue to advance, as I mentioned, a lot of them have advantages and disadvantages. Everybody, of course, is going to think the best about their machine, which they should, right? Yeah. Uh but what's real, you don't know until you actually try it. So that was one piece of Brackett was to make on the cloud, anybody could use a quantum computer and and write a quantum circuit using entanglement and superposition, which is the unique pieces. Remember in that uh that uh progression of computing, each one had a unique advantage. And so the unique advantage with a quantum computer is superposition and an entanglement, which is very hard for people to wrap their heads around because we don't our our human senses don't sense those in our environments.

SPEAKER_02 13:10

So what you're basically saying is that we may in the future solve different problems on different parts.

SPEAKER_01 13:15

Yeah, for for sure. And I think I think one of my exciting times was the first time we used two quantum computers with a classical computer together to solve an initial problem.

SPEAKER_00 13:26

What problem was that?

SPEAKER_01 13:27

It was it it was basically uh Hamiltonian of a water molecule. It was not anything you can't do on your iPhone, but it was it was fun to kind of do it across the different machines. And I remember texting my kids, hey, today was the first time two quantum computers were used with a classical computer, and they were like, Please memorize the day. Yeah, remember the day. They they won't and then and then there was another time with uh our we we built uh quantum networking, which is something I haven't mentioned. We spun out a company called LightSync that was on our team, and we were using diamond voids to store and repeat quantum state, which allows you to entangle across two quantum computers. And so that was another big day when we we we did the first entanglement of two quantum computers that were they were in different building blocks away, but we ran them through two uh hundred kilometer coils of fiber to show that that they could be entangled that far away from each other. And that was another day I texted my kids that they probably won't remember. But you know, it's like it's you know, it's I I think as these things continue to advance, it's it's very fun to see uh um the the improvements there. But I think there's still a lot of challenges around error correction. Of course. Um, at as you you get to these um these machines of all different types, just because they're so, you know, you think about how much the environment is affected to atomic particles today. I mean, you electromagnetic waves, cell phones, Wi-Fi, vibration, truck drives by. We had a one machine we were working with uh one of our partners, and we were getting these sporadic errors all the time from the machine. And it turned out it was the motion sensor in the room where the machine is where people would walk in and the light would come up. Really? And the photons would interact with the lasers and cause errors.

SPEAKER_02 15:08

Because of people walking in the room.

SPEAKER_01 15:10

So we just disconnected the motion sensor light and that fixed it. But it are those are the that level of sensitivity is kind of what you're dealing with. Yeah.

SPEAKER_02 15:18

Yeah. And actually, you you you spoke previously. Um, I was about to point to you on this the quantum sensing and how this industry is growing um fast. I also see it from the perspective of quantum. We get more and more those startups popping up. And many people experts like like yourself are basically saying that the last two, three years were very crucial for quantum uh ecosystem to develop and technology itself. Um, I wanted to ask about so what actually did happen in the last two, three years that would make quantum more real than before?

SPEAKER_01 15:50

So a number of things. So so the two ways that you interact with most of these machines is with lasers and with microwaves. So the electromagnetic cryogenic machines, you're using microwaves to sense and control and set the qubit state. With the neutral atom ion and and photonics machines, you're using lasers and photonics to do that. Right. Again, these operate at very, very low temperatures, as close as we can get to absolute zero, really, which is kind of amazing. That amazes me as much as much as the machines in some ways. You can get uh the cryostacks we would use it for the cryogenic machines, get down to microkelvin. I mean, it it's really, really low, but you can actually get colder with laser cooling, which is what the other machines use as well. Have a big cryostack. They have a little bit of cryo because they they operate in a vacuum and they want to cool that vacuum down so that because you can never get an absolutely perfect vacuum, but they cool the vacuum down, so there's that. But then there's the and then with the optical machines, you you've got a cryostack on the optical sensor. So there's still cryo involved, but not nearly as much. So communications, of course, uses microwaves and lasers. And so there's been an advancement in microwaves and lasers that uh have that quantum computers have been able to take advantage of. There's also been some advancements in fabrication and material sciences that quantum computers have been able to take advantage of. So I think there's a symmetry also occurring where there's new things being found in the quantum, say laser control systems that go back into comms and make comms better. Because you know, all of our fiber that we do are all laser-based. You're also starting to see more and more like the latest Varerubin chip from NVIDIA has actually got quantum photonic, I mean not quantum, it's got photonics in the chip. Right. So there's embedding photonics in the silicon, right? So to do the linkage for the clustering and things like that. So I think you're gonna see more and more silicon photonics. That helps quantum computers.

SPEAKER_02 17:49

Yeah, and that's all happened in the last three years. Do you think since now on, the acceleration of development in terms of quantum computers and quantum technologies as communication and sensing and network and so on will accelerate much faster?

SPEAKER_01 18:03

I think it will continue to accelerate. I I think there's there's still some big challenges. There isn't really a quantum RAM. There's there's a ways to store some quantum state in a diamond void or a piece of electric oscillator or electric oscillator, but you really need quantum RAM. That's gonna be something uh quantum storage. You know, there's uh electromagnetic tunneling for quantum storage, but again, hasn't really advanced. And so there's not a lot as much research going on there that needs to be. But I think error correction is going to continue to be the hundred things to overcome. And um, I think each of the machines that you talk about out there uh, you know, have their advantages and disadvantages. You know, the electromagnetic cryogenic machines are very fast but noisy, so you have to have a lot of air correction. The uh photonic machines have the advantage of, you know, you're not cooling down photons. You can't cool a photon, right? Uh, but their disadvantage is that they're um, you know, the entanglement isn't deterministic. So you have to to do it many, many times to to get enough entangled cubics. The uh neutral atoms and the uh uh charged atoms or ion-based machines are more coherent but have challenges in speed. So I think there's advantages and disadvantages to all of them. And I think some of them, I I think like I said, I think you'll see some of the neutral atom machines have the largest number of error-corrected qubits, logical qubits first. But that that there'll be challenges because of speed, you'll have to run them to do something economic. You'll have to run them for months potentially with and and they're kind of unstable, right? It's uh you know, there's they're still kind of lab cody type machines right now. But you know, I I think what you always start at like this, right?

SPEAKER_02 19:42

So if yeah, that's that's a very good so but if you can talk about it, uh what are your predictions? How many more years do we need to, you know, get to the level of calling quantum industry?

SPEAKER_01 19:53

Yeah. So so like I said, I I think the quantum sensing is quite advanced and mature and is going to continue to to advance and mature. So I think that's that's well, well launched. I think the quantum networking, I think there's an initial foundational stuff for that. But you don't really need the quantum networking until you have the quantum computers to be.

SPEAKER_00 20:11

Of course. Right.

SPEAKER_01 20:12

I mean, you can use it for quantum key distribution um and other things like that, and you might see see that continue to move forward, certainly post-quantum, which is a math problem, not really a quantum problem. You're seeing that be deployed right now. The biggest challenge is uh is what I will call economically viable, differentiated quantum algorithms on a quantum computer that has enough error-corrected qubits.

SPEAKER_02 20:36

Yeah. Okay, but in terms of the years, what do you think? It would be five, ten, fifteen, twenty, how many more years we need? What's your guess? And I'm asking, don't feel intimidated. I'm asking everyone about that question.

SPEAKER_01 20:47

So my best guess is that you'll start to see the first economically valuable algorithms run on a quantum computer in the next five years. Five years.

SPEAKER_00 20:59

Fast. I guess it is fast.

SPEAKER_01 21:01

I mean, that you won't get to where we want to get to, which is, you know, we we'd love to get to like a thousand error corrected qubits. That that that changes the world as we know it, right? A hundred error corrected qubits starts to become pretty economically valuable. And the first things you're gonna do with those is material science. Because basically, a quantum computer works like a molecule. You're kind of building the circuit that you build in the machine's memory is kind of like building a molecule, if you like. And so imagine if you could reverse engineer material. So think about it. Today we just ran like somebody cooled down a ceramic and discovered it was conductive, who'd have thought, right? But that's a random discovery, right? The the people making the sticky notepads were trying to make a really strong glue and made a crappy glue and go, what could we use this for? Well, this is great, right? They have some.

SPEAKER_02 21:49

So kind of discover it by accident.

SPEAKER_01 21:51

We discover it by accident. So so like like one of the things we were working on at Amazon was you know, we always start with a working backwards document. So we start with like what would The end state. So the end state we we were working on was to do a Hamiltonian for ammonia. Now, why ammonia? So ammonia is the most produced and transported chemical that humans make.

SPEAKER_00 22:13

Really?

SPEAKER_01 22:14

We've been doing it for 120 years. It's the base for fertilizer, it's a base for petrochemical operations. In addition, ammonia is a way to transport hydrogen without carbon. So it could replace fuel. That would be zero carbon. Wow. Right. So that'd be very big. The challenge we have with ammonia is that it takes more energy to produce it than you get out of it if you do that. But it also is just a very energy-expensive system to produce for fertilizer and all these other things we use ammonia for. We know that if you take a piece of bread and throw it outside, that bacteria will grow in the bread and they'll produce ammonia to break the bread down with almost no energy. So we know that it can be done.

SPEAKER_02 22:57

Again, once again.

SPEAKER_01 22:59

But we don't know how to do it. And so if with a Hamiltonian on ammonia and a thousand aeric qubits, we could reverse engineer that process and that would change the world. But think about it another way. What if we could reverse engineer a high temperature superconductor? That would change the world. There's a lot of these kinds of things that we'd be able to do.

SPEAKER_02 23:19

It's actually very interesting. It just reminded me my father always was saying that he will never be impressed by technology until we create a machine when from one side you put the grass and from the other you get the milk.

SPEAKER_01 23:31

So I think these uh that that's where I get excited about quantum computing is getting to that point where you can do these Hamiltonians and they'll be very, very valuable as you do it. Uh I think the the other uses of quantum computers. So at Amazon, for example, we ship a lot of stuff around. There's the traveling salesman problem. That's an NVRD problem that a quantum computer can solve, but you're going to need about a 4,000 error corrected qubits for that. Around 2,500 air corrected qubits for what we know of to do Shore's algorithm, which is the worry around cryptography.

SPEAKER_02 24:05

Yeah.

SPEAKER_01 24:05

Right. So we have a couple of more years. I think that's about 2040-ish time frame, right? For for those kinds of things. So, you know, but I think, like I said, there's still big challenges. You know, that this race is ongoing. You know, you'll see these like these first photonic machines and these first cryogenic machines really being like the size of a football field. And I don't think people really realize that.

SPEAKER_02 24:27

I didn't even think about it. Would we miniaturize then after that? Sure, the same way how we did with classical.

SPEAKER_01 24:34

Yeah, for sure, for sure. I mean, that's what SEC's trying to do, right? If SEEK SEEK can get to that point where we can put enough error correction in the dilution fridge, the machine gets much, much, much smaller, right, and cheaper to produce. So I think all of those things, I think the if you take a look at the photonic systems, the cryogenic photonic sensors right now, they're about as big as your fist. And if you take a look, there's a lot of wasted space, right? We could we could make that the you know the size of your fingertip, right? And that that makes it smaller. So I I think again, the neutral atom machines and the ionic trap machines are not as big, right? Potentially, but um, you know, scale will hit them too, and speed will hit them. So there's challenges there as well. So so I don't, you know, you know, there's there's there's it's it's interesting. It's it's thing there's things I love about all of the designs and things I don't like about all the designs, but um, it's it's sort of inherent to those different technologies, right?

SPEAKER_02 25:29

Speaking about Seek, another startup that you cooperate with. I have a question from a different angle. So let's say Bill, if I'm a startup, um, let's say developing some quantum algorithms, maybe software, maybe error correction. How do I how do I see both Allen? Is it like more like a do I see you as a partner or do I see you competitors? Because you cooperate with the different partners, but you also develop your own product, right?

SPEAKER_01 25:55

So how do I approach So we're not building a quantum computer, we're building experiments. So if you're a quantum computer developer, uh you're a partner for sure. I think we we strive to be partners with companies like Seek and others, and we strive to be very partner-centric, right? So so I I think we're a good, you know, if you're interested in in advancing your technology for big businesses or for uh the government, Booz Allen is a great partner for you to work with. Um, you know, Amazon's a great partner to work with as well. You can get your algorithm on the marketplace, you can stick it on BRAC. People can try it, and when they try it, it goes on their AWS bill, and then we AWS would just send you money, right? So that's a nice, a nice easy channel to do it. And so where we do a lot of work with partners is where we have a very unique mission or business understanding on how to implement it. So you might have a a company that really understands, say, a quantum radio frequency sensor, a rubidium, you know, and a laser system that they're doing on top of that. And we can go and say, okay, based on our FX, our radio frequency experience in classical radio frequency operations, we can enhance this and we can implement it in these ways, and we can put it in these locations, and we can make them valuable for people to use them, right? And that's where we would pull that through. So we wouldn't be building the sensor, we would be accelerating the sensor. I think, you know, the and then we often we also have an investment fund where we invest in quantum companies.

SPEAKER_02 27:31

That's amazing. Tell us, please, please tell us more about that.

SPEAKER_01 27:34

So in fact, for example, we invested in Seek. We have uh uh a partnership with uh Andreessen on A16Z, where we have an investment fund there as well. Um and then we have a portfolio of companies we've invested in.

SPEAKER_02 27:48

And that's for the all deep tech or specifically for quantum? And what kind of size of tickets are we talking and which round? Are you starting on the earliest seed or we're we we're usually around B.

SPEAKER_01 28:02

So more developed companies. And we're more of a strategic investor. There's a few places where we've done more than strategic investing and larger in specific areas that are really important. Yeah, yeah. Our core technologies are things like autonomy. We do a lot of work on autonomy, autonomous systems, physical AI, we do a lot of work there, digital twins and those kinds of things, OT environments, IoT, and things like that. We do investments in defense tech and areas like that. We do investments in space, we do a lot of space work. Um so we just invested in a company that's got a really unique, uh, super powerful uh solar-powered uh thruster, for example, that's really exciting. We've invested in like uh starfish that does intersect and can grab a satellite and move a satellite or ultra-low or earth orbit satellites and things like that.

SPEAKER_02 28:55

It seems like you have a kind of a portfolio. How many how many startups do you have in the portfolio?

SPEAKER_01 29:01

Approximately don't quote me. In that range. It's a couple of pages for when I go through it each time. Yeah, yeah. And so we're always adding to it. And then we have our big partnership with A16Z, which is like G1600 companies, something like that. So and then we also do big partnerships with like NVIDIA and Amazon and and others like that as well.

SPEAKER_02 29:22

So and what's the best way for the startup? Let's say we we have a bunch of startup um in our audience, if they want to contact you for investment, how do they do that? Who's the best person to talk to? Or did it go on the website?

SPEAKER_01 29:35

Yeah, you can't just go to our website and uh there's there's a place where you can connect with us and talk to the investment team. Yeah, portfolio team. And so I do uh personally, I review all the investments. Our our team goes over it. We have an investment group that sits down and looks at each company and uh does a background on them. We do a lot of forward scouting, uh, like there'll be uh system solutions that uh that we're trying to build, and I'll look at gaps that we need to fill and say, here's investments, and they'll go the team will go out and look for a specific subject. Yeah, a bunch of companies that could potentially meet that. And we do uh so we have a whole quantum team that looks specifically at companies for quantum investments. We also have like energy teams that are looking at things like atomic batteries and and and and other areas like that.

SPEAKER_02 30:22

And may I ask you so because obviously you work so close with US government, are you interested only in domestic startups or are you also open for international corporations?

SPEAKER_01 30:32

Um we're open for some international corp corporation, but mostly it's gonna be uh companies that can work with the US government. Yeah, so we have to have a clearance or or have the ability to to to be a supplier, right? Right. So for example, we just did uh an investment in a simulation company in the UK. But they're approved for use with the US government.

SPEAKER_02 30:53

So as long as they're approved for use or can be a specific market, then we know which the of the market are not. Right, right. Yeah, okay. Ah, that's good. Actually, speaking about this, because this is very interesting, and you mentioned previously about the quantum race between the countries and so on. So I have a tricky question for you.

SPEAKER_00 31:08

Yeah.

SPEAKER_02 31:08

Um from national US security point of view, what keep you up at nine to more? Option A is that China is developing quantum advantage first. Option two, that the US is failing in implementing quantum effectively. I worry about it.

SPEAKER_01 31:27

Yeah, yeah. I I worry more about another country getting a quantum advantage before, you know. Uh well, because it imagine another country gets a high temperature superconductor and that changes everything, right? It really changes it. It changes the whole energy equation, it changes transport, it changes storage of power. I mean, you know, in in the end, everything is about power, right? I mean, you you know, you you can't run computers without power, you can't run drones without power, you can't run cars without power, you can't run inf cities and infrastructures without power. Energy density is a really big deal. And so that would be transformational. Same thing with uh the ability to run shores, for example. Hopefully everyone listening to this has already implemented post-quantum encryption and it won't be an issue. But if those I can tell you there are people who are not, and so so I think that's to be another area of of of worry.

SPEAKER_02 32:20

I think it's wishful thinking at all.

SPEAKER_01 32:22

I think it would be economically, you know, uh damaging to the United States if we were not at least at peer on quantum at that time.

SPEAKER_02 32:34

My previous guest, Andre Kunig, um told me recently that basically most of the five h um five uh fortune uh five hundred fortune companies are already having quantum team. It doesn't mean that they're getting prepared or implementing something, but they already trying start trying to do it um something. Bill, a question for you if I am a CEO of a 500 fortune company, what should I exactly do in the next, let's say, 12 months to protect my company or to getting ready from the so I mean the first basic thing is you should if you haven't implemented post quantum encryption, go do it.

SPEAKER_01 33:08

It's we can help you with that. We definitely can engage and help you with that. We have the ability to deploy post quantum encryption sense and deploy for you. If you need help with that, you can call us and we can do that right away. I think the the you know, you know, the all of the cloud providers offer postquantum encryption and the transport layer. So it's it's really not that hard to do. And so, but it's pervasive. It's like a Y2K problem. It's the you know, making the change from six digits to eight digit data is not that hard, but it's everywhere. And so taking the time, that would be the first thing I would do. The second thing I would do is I would take my quantum team, and I don't want to, you know, uh uh pump my firm my former employer too much, but um bracket is right there, go use it. Learn what as as each machine comes out, learn, learn what it can and can't do yet, right? And and be aware aware, right? Because again, you don't want to wake up one day and be Ubered, if you like, by another, by you know, you don't want to be a taxi cab company that got Ubered, right? You want to wake up and you want to know where these machines are and how well they work, right?

SPEAKER_00 34:13

How fast in your domain.

SPEAKER_01 34:15

I mean, you if you think about it, like like I said, the first big transformations will be around material science. So if your company makes things and you should be up to date on that, right. Especially, you know, if you're in um, you know, avionics or um, you know, you're manufacturing, you know, raw materials or you're doing advanced electronics or you're doing petrochemical or health care, you better be on top of where the latest number of error-corrected qubits are and what kind of a Hamiltonian that can actually build. And I think, you know, like I said, today you'll go out and you'll experiment with these machines and you'll say they're interesting but not quite ready. Every next week, they'll still be interesting and not ready, but they'll be more interesting and they'll keep getting more and more interesting. And so I think engaging and staying on top of it and having a team would be important. But the practical thing you can do now is post-quantum encryption, right? And I think other areas, of course, is if you're in the business of sensing, building sensors or using sensors in your environments, quantum sensing is quite real today. And there's things you can be deploying on that that could be transformational for your business too. So just be aware of that and what's going on there.

SPEAKER_02 35:30

Amazing. Okay. And from government perspective, what do you think could be like a first application that can be indispensable for the government without those big headlines and you know, PR and everything?

SPEAKER_01 35:45

You mean not you mean you mean out outside of quantum sensing that they're already using, right? Yeah. I think that's I mean the the government is a big user of quantum sensing and atomic clocks and other things like that already. I think you know, still material science they will care a lot about. So I I think that's gonna be the next big thing that they're they're gonna be focused on.

SPEAKER_02 36:03

And obviously, so can you give us uh one example of material science use case for for government? If you can, I don't know.

SPEAKER_01 36:11

A high temperature superconductor, right? There's a lot of research going on across the government with that. Other areas, of course, is high strength materials, right? So so um there's a lot of research going on with you know uh carbon nanotubes and carbon nanotubes and and things like that. Um uh bullets. Where do I use components and things like that? There's uh uh um you know ways to transmit electricity through it. There, I mean, there's a whole bunch of th uses for all these different types of things. So a lot of the body armor today is not not carbon nanotubes, but is uh ceramics. And so there's a lot of material science around ceramics. Uh there's a lot of material science around radar absorption for stealth, for example. So I mean there's there's just a a tremendous uh number of things in there. So I think the material sciences will be be pretty important for the government. And then of course, you know, obviously cryptography, right, will be another area that they worry about. But I think it's um, you know, it's just it's just one of those things we need to stay on top of and make sure that that we're first to it or or at least at peer. Other areas I think the you know that we need to be aware of is the quantum uh fabrication capabilities. And so not having supply chains for quantum fabrication that rely on other countries. So we need independent, you know, the we're we're working to do that with the CHIPS Act and other things on fabrication for chips, even in even with our silicon fabrication, still there's prefabs that occur before you fab, and those prefabs are often foreign sourced. And so trying to get same thing with quantum computing is the the materials and prefab in the United States. The supply chain domestically delivered, right? So everything from you know uh you know critical minerals, right?

SPEAKER_02 37:56

What's the situation looks right now in the US in terms of all of this?

SPEAKER_01 38:00

Right now, I think we're still very dependent on other countries for the the supply chain for quantum. So that's something that we need to address, just like we're trying to address it for chips and silicon, right? So that's another area that I think. Yeah, the yeah, the the um, you know, a lot of the materials used, uh some of the you know, uh, meta materials used in quantum computers and other things like that um are kind of exotic, you know, piece of electric oscillators or like the it really depends uh upon what you're making the substrate out of. The the fabs for a quantum computer are not as as dense necessarily. I mean, your Joseph injunctions have to be very clean, other things like that on a transmon. But the you're not looking at doing a two-nanometer fab for a quantum computer. You can do a 22 nanometer fab for a quantum computer, things like that. But I think there's still the fab materials coming in are not all readily available. We had uh a lot of challenges in supply chain for fab material when I was at Amazon during COVID, for example. And it put us behind on trying to do that. I think, you know, so so you're gonna see, you know, I think kind of understanding that full that full the people should be looking at the full supply chain for silicon and the full supply chain for quantum fabrication.

SPEAKER_02 39:15

So wait second, so you said this COVID literally pulled behind cloud for um for a period of it.

SPEAKER_01 39:22

Put us behind on our our development of our quantum hardware. We couldn't get some of the PC electric uh material we needed for our oscillators.

SPEAKER_02 39:30

Okay, I didn't know that either. Okay, obviously everyone was aware about during COVID about the supply chain problems, but I didn't I didn't even imagine it could affect the cloud. I mean, thinking cloud, I think digital. Okay. So speaking of the timelines, we we talk a little bit about more, and you said that five years is probably the the time that we need to, you know, develop certain technology to the level when we can really have economical value, uh value on this. Um but in terms of if we if we talk about a full commercial timeline, where basically almost every company will use quantum for different purposes. Do you think we are closer to I don't know, pre-cloud times or maybe pre-nuclear fusion?

SPEAKER_01 40:16

Yeah. So so you know, obviously these numbers are changing. So there are a lot of companies who are talking about 2027 as having the first uh commercially viable quantum.

SPEAKER_02 40:26

And some countries even put it in strategic visions and policy and stuff.

SPEAKER_01 40:30

And I think you know, people are still trying to reach those dates. Maybe I'll be surprised. Uh yeah, it's it's probably more like 2032 to 2035. And then the the the larger machines, larger number of error corrected cubes is probably around 2042. That's but you know, so hard to predict these things. I I think the positive side of a lot of this is on the error correction side, we know what we need to do. It's it's really more of an engineering problem than just a science. You before it was a science project. When I first started working with quantum computers in geez 1994, it was it was science. Right. It was like make make one qubit, right? You know, that that that was that was sort of the the the prevalence of the early days. And now it's it's we know how to make lots of qubits, so we can do that. We know how they work, we understand them. Um it's really the noise and error correction and try how to how to do that. And and that that's an engineering problem, and it's a big engineering problem. That's what I mean. These like I said, some of these machines are gonna have seven million physical qubits. And you know, I know the neutral atom and the ion trap machines are trying to make so their machines run run faster. So there's there's all of this is ongoing, right? And you know, there's challenges in like to build big enough dilution systems and and you know, uh, cryogenic systems. There's challenges in all these areas. So but they're engineering challenges a lot in a lot of cases.

SPEAKER_02 41:56

And we can potentially up obstacle them. Yeah. Um as we did in the different industry.

SPEAKER_01 42:02

Yeah, I think the the other side that I worry about is the skill sets, though. There isn't um enough quantum talking. Enough enough people who have quantum talent, right? I think that's gonna be the biggest problem that we run into as these become real, I agree, is being able to consume them. It's a it's an analog machine, it's not a digital machine. So that's a big difference. The the the quantum computers you're dealing with entanglement, which, you know, if anyone tells you they know how it actually works, they're lying to you. They they know what it does, we know how to use it, but we don't really understand how it works. Superposition is very hard for people to understand too.

SPEAKER_02 42:43

So okay, so they're all different topics and and very hard for potential people. But why we struggle with the talent? And if we know that we are already in a shortage of the quantum talent, why the government is not doing anything about it.

SPEAKER_01 42:55

Well, I think the the I mean the government is trying to do things about it, but I I think it's still part of the problem is, as we said, there there is not a quantum computer yet that has a practical implementation for business. There's a lot of science, there's a lot of I know I just said engineering, but there's a lot of like made-up equations to try to show a quantum advantage, right? Right. But what you really are looking for is like a Hamiltonian for ammonia or something, you know, it's something that that's really commercially going to be commercially viable, right? And so if I train you on a quantum computer today, you'll go on bracket, you'll build a few circuits, you'll go, wow, I understand it. And then you'll go work on AI or other things like that. And then you'll get rusty, right? You'll get rusty on it. And so I think there's a certain base training. I I'm hoping that in every quantum computing courses that that they teach basics around quantum computing, just like you know, I had to learn RS flip flops at NAND and NAND gates, right? You know, when when I was even though I don't, you know, I'm not an electrical engineer, you had to learn it to understand how a computer worked. Uh I think And I was like, you know, to really understand it, I need to program it. Show me how to program it. Because I had I've talked about them a lot. I worked on them a lot, but not not actually programmed it. And I said, he says, Well, what do you want to do? I said, Well, how about we just sort an array? And he looks at me and he's in his Italian accent and he says, That would be a very stupid use for a quantum computer because it wouldn't do that well, right? And then he thought very, but this is a good example, he said in his Italian accent. So we sat and we were on our whiteboard and wrote up how to do it. And it it was transformational for me to understand how it worked.

SPEAKER_00 44:38

In practice.

SPEAKER_01 44:39

Yeah, in practice, right. And so I think that kind of that that light needs to go on. But I think other things that need to happen, if you think about like like NVIDIA has been amazing. Uh, I remember when when Jensen was pushing CUDA. And a lot of people were like, why are they building software? It doesn't make any sense. They should just build good GPUs. What are they doing? And the reality was, as we used to say, that parallel programming causes drain brammage. It's very hard. It's very hard to parallel program. It's uh you have to be very skilled in it. So very high-level programmers, but CUDA makes it easy, right? It abstracts from the complexity of it, right? And so what we need is like a CUDA for quantum computers. Yeah, yeah. And they've they've talked about quantum CUDA, but uh we need something like that that makes it easier.

SPEAKER_02 45:25

Idea for startups.

SPEAKER_01 45:27

Yeah, yeah. Well, or or even just uh like so so as I as I mentioned, that's one of the things we're trying to work on is how would you do uh a Hamiltonian simulator that makes it so a chemist can use it and not know anything about a quantum computer. So a chemist can use a classical computer today and not understand an and gate, right? And and not uh even understand binary, right? They can use a quantum.

SPEAKER_02 45:52

And we use a lot of tools and you know, and don't understand it.

SPEAKER_01 45:56

And so if we right now you can't use a quantum computer without actually understanding the circuits in the future, if we can abstract that. So I and I think the reason that that doesn't exist is the quantum computer isn't there to, you know, if there's a chicken and egg problem.

SPEAKER_02 46:12

But I'm still surprised, um, Bill, you know, because I think so. First of all, last year I was doing the executive program on on Stanford, and I was very surprised, there's nothing about quantum. Like nothing in the program embedded, but also there's no courses. I was looking for some courses, and the last one that they had, it was from 2017. So I I bet a little things change, right? And you know, uh for me it wasn't that crucial, but I was really, really surprised that we knew not educate the you know, corporate enterprise level executive on that problem.

SPEAKER_01 46:42

Yeah, I think I mean there's there's you know, you have to be a curious person. I, you know, and I I've somewhat my just a funny story. My wife was having a a party uh for her business, and so I was sitting at a table, and one of the people at the table says 20 people at each table in our at our house, and one of the people asked me a question about quantum computing. So I started to explain quantum computing, and very slowly everyone at the table except him got up and moved to the other table.

SPEAKER_02 47:08

As the king, if you want to kill the party.

SPEAKER_01 47:10

Right, right, right. So like like with your husband, we I sit around and talk about quantum all day long. But I think because he's fascinated and interested in it, but you know, uh a lot of people are not, right? And so so you you have to, you know, but I I think you need to kind of foster that uh uh curiosity in children, and you need to foster that curiosity amongst engineers um to continue to build this out, and it's gonna be very, very important. I mean, the engineering problems are are significant to still let's hope it will change shortly, right?

SPEAKER_02 47:40

Move the AI also. There was a only bunch of geeks previously right uh interested about that before it's exploded and now everyone's talking and it's buzzwords, yeah. So yeah, we need a couple of more years, I guess. But this is why I'm doing what I'm doing as well. And hopefully the university also will will follow up on this from both sides from the physics. Right. We need more quantum courses. And speaking of, I was actually looking for um postgraduate uh diploma in some quantum aspect, and there's only maybe five courses around the world.

SPEAKER_01 48:14

I know, I know.

SPEAKER_02 48:15

It's very sad, and they're not very flexible, most of them are not online, and you know the professional people usually don't have like a year or two just to move somewhere to you know attend the university. Um there's still a lot to do. Yeah.

SPEAKER_01 48:28

I mean we teach a quantum course at Booz Allen. I I think we've probably trained around 1,500 people for external people or only on your your employees. Employees, employees because we're trying we're trying to continue to build that skill set.

SPEAKER_02 48:40

And that's a good example and hope the other enterprise companies will implement something as well, like internal trainings. Okay. I think with this aspect we probably should finish looking in the time. Bill, thank you so much for being here with us. Let's also pass some greetings for our friend um Dr. Simone Severini, who was mentioned a couple of times today. I'm trying to actually convince him to come join us as a guest as well.

SPEAKER_01 49:04

He should do that. Yes.

SPEAKER_02 49:05

Please talk to him if if you see him next time. Thank you so much.

SPEAKER_01 49:08

Yeah, thanks. It was fun.

SPEAKER_02 49:11

That was Blonde and Quantum. Thank you for joining me on this journey through the quantum business frontier. If you like the episode, please review another Spotify or Apple Podcast and help more people discover the quantum world without needing to untangle the theoretical physics. See you next time.