Hacked - The Year of Quantum

Episode Date: February 4, 2025

Quantum computing raises big questions—about security, power, and who benefits as the technology advances. To kick off 2025, the International Year of Quantum, I sat down with Joan Etude Arrow—Founder and CEO of the Quantum Ethics Project and a Womanium Quantum Solutions Launchpad Fellow—to talk about what’s real, what’s next, and why it matters. This is our chat with Joan Etude Arrow, here on Hacked. 🔗 More from Joan: quantumjustice.bsky.social Learn more about your ad choices. Visit podcastchoices.com/adchoices

In 1994, American mathematician Peter Shore developed the algorithm that now bears his name. And in doing so, he sparked an arms race that's still being run today. And what Shore does, it has only one application, and that's cyberterrorism. Shor's algorithm is a quantum algorithm. You need a quantum computer of a certain scale, a machine that leverages quantum states, such as superposition and entanglement rather than classical bits, to run that algorithm. But if you could build a sure-capable quantum computer, it could do something that's effectively impossible on classical computers at any useful scale. It could, theoretically,

break RSA. RSA. RSA, one of the oldest and most widely used methods for securing digital data. Which is the foundation upon which all of our digital life on the internet is based. When you send a private bank transfer, when you send private photos or messages, those things would no longer be safe. Yeah, RSA is super common in keys, public and private keys. So it's big time, like if you're a developer, you use SSA, use GitHub, any of those things. Often your identity is authenticated to the Git servers. Any kind of like that digital layer of identity authentication is typically RSA keys, public and private keys. It was developed in the 70s, and as you said, it's how we secure a lot of our digital lives,

like private accounts, messaging, online banking. It is like a cornerstone of modern digital privacy. It's not the only one, but it's a big one. And the whole point is that no matter how powerful traditional computers get, the math to crack RSA is effectively just too hard. But a quantum computer capable of running Shores algorithm, that could theoretically break RSA and similar encryption methods

leaving all of those digital secrets. Decades of stuff we've tried to keep private puts it all up for grabs. Regular computing is cheap, widely available. You can run a lot of programs on a personal laptop. You can run a lot of big programs on a supercomputer. And quantum computers, by comparison, are extremely expensive. And so the only place where a quantum computer

will be desirable to use in place of its classical counterpart, that is what we refer to as quantum advantage. When a quantum computer is able to produce a computational advantage that you don't see classically. When Shore announced his algorithm, even though it required computers that didn't yet and still don't exist, it kicked off this scramble. We rely on that encryption today to safeguard secrets. We're still going to need secret decades from now. So the first one to get there is going to hold an extraordinary amount of power. I like to pitch it as it's the space

where the quantum computer beats the world's biggest supercomputer. The day that that happens is the day that quantum advantage has been achieved and it'll usher in a whole new social era, a whole new era for society. Last year at DefCon, we met Joan Aetude Arrow. She's the founder and CEO of the Quantum Ethics Project and a Womanium Quantum Solutions Launchpad Fellow.

She was at DefCon teaching quantum computing at a table to amongst others beginners like me. 2025 is the International Year of Quantum, celebrating a century since the discovery of quantum mechanics. To mark that occasion, sat down with Joan to talk about quantum computer. Its potential, the ethical challenges that it raises, and how it might shift power dynamics in society. Our conversation was a lot of fun. It was a little bit mind bending and got a lot of insights into how to make sure this technology benefits everyone. Yeah, I remember our conversation, like when we talked to her at at DefCon, I remember I had a good conversation with her about machine learning and how

it's advancing their ability to tune the algorithms for quantum computing. Cool. No, it's a lot of fun. I think folks are going to enjoy this one. Our conversation with Jonah Tudero here on Hacked. When we first met at DefCon, you were sitting at a table in the quantum computing village. You had a little audience of people kind of gathered around and you had a sign in front of you. I want to make sure that we get this right. What did that little sign on the table say? Yeah, it was just, I made it spur of the moment.

It just said, ask me about quantum. I just slapped a couple stickers on it from, you know, DefCon. They have all kinds of different stickers. So I tried to populate it with a couple of stickers, give it that DIY hacked together feel. They're all about the stickers there. Okay, ask me about quantum. I am here to ask you about quantum.

Love it. For anyone that doesn't know, what is quantum computing and how is it different from the classical computers that we are talking on right now? Yeah, so quantum computing, the original idea for it was come up with in the late 80s by Richard Feynman. And the simplest difference is in the hardware that's used to do the computing. So the devices we're talking on now, laptops, computers, cell phones, all those devices, their fundamental piece of technology is the transistor. which can be thought of just as an on or an off switch. A yes or a no, a one or a zero. Those transistors together give us the ability to store and manipulate and send binary.

And that's many of many folks will know, just the basis of everything we do in the digital space. Photos, text messages, audio files are all at their most fundamental level, a bunch of transistors in yes or no position. A quantum computer replaces the transistor with something that can be more than just one or a zero. We're going to dig into it a little bit deeper. Before we do, your focus specifically is on the ethics of quantum computing. What brought you into this? Yeah, so I started off in quantum machine learning. And so thinking about, when I talk about ethics, I'm making reference in many ways to the field of AI ethics,

where folks think about the human or social impact of AI or machine learning. And when I was a graduate student, I was at the Institute for Quantum Computing at the University of Waterloo in Canada, I was looking at all these fascinating conversations being had about what does Chat GPT do for society, the good things, the bad things. And then I turned around and the folks in the quantum spaces that I was primarily working in were talking about enhancing those technologies, enhancing AI, enhancing machine learning with quantum. But we weren't having a conversation about what that would also mean for society. And that's kind of where the idea for the quantum ethics project was born, is to try and understand these possible societal implications. A lot of, we had a bit of a discussion before this, and a lot of those implications seem to sit on the far side of something called quantum advantage.

What does that mean in practical terms? And why is it sort of an important milestone in this field? Yeah, so regular computing is cheap, widely available. You can run a lot of programs on a personal laptop. You can run a lot of big programs on a supercomputer. And quantum computers, by comparison, are extremely expensive. And so the only place where a quantum computer will be desirable to use in place of its classical counterpart is that is what we refer to as quantum advantage. When a quantum computer is able to produce a computational advantage,

that you don't see classically. I like to pitch it as it's the space where the quantum computer beats the world's biggest supercomputer. The day that that happens is the day that quantum advantage has been achieved and it'll usher in a whole new social era, a whole new era for society.

When you say a whole new... Oh, there's like five questions off of that. When you say a whole new era, broadly speaking, what do you mean by that? Yeah, like think of it, think of it in terms of what society looked like pre-computation. Digital computers were invented,

Alan Turing, people see in the imitation game, right, World War II, cracking ciphers. But it wasn't long until computers were then being used to calculate things like missile trajectories, which were done in World War I, totally analog,

like with tables of data, people doing hand calculations to try and calculate these sorts of things. So just kind of of try to conceive of what society looked like before the era when computers existed. There was so much you had to do by hand. There was so much that was just infeasible to do because it took too long because it was too complex.

And when computers came along, they basically brought a cheap way to solve problems that we weren't able to hack with a pen and paper. Think of quantum computers as the next step up on that ladder. So there are problems that exist today, which you can't hack, even with the world's biggest supercomputer. If we had a quantum computer for some of these, it would bring in a whole new set of problems that we're previously out of reach. And so trying to, and I'm trying to give a sense of how big a deal that would be for humanity, for society. if we could solve some of these problems like drug discovery or better optimization routes.

We rely on optimization all day long. If we could do it better, that would have a huge impact. Drug discovery is fascinating. The one that I was curious to talk about, and it makes sense we have sort of a security bent on this show. To call it solving a problem is maybe unintuitive. In my mind, it's encryption that these things. that this could circumvent certain types of encryption.

Is that, do I hold, is that thing I have in my head that this could theoretically undermine a lot of the modern encryption that we're using? Is that accurate? Yeah, well, that's the reason quantum is being invested in by major governments around the world. I mentioned the original idea was late 80s. Richard Feynman, you know, late 80s,

had the idea that a quantum computer might be good for simulating things in nature. But it wasn't until 1995 or 1994 that Peter Shore publishes this Shore algorithm, which is kind of the first one a lot of people here. Maybe people might be familiar with this. And what Shore does, it has only one application, and that's cyberterrorism. It breaks RSA encryption, which is the foundation upon which all of our digital life on the internet is based. when you send a private bank transfer,

when you send private photos or messages, those things would no longer be safe. And even up to the level of major government secrets could be intercepted, decrypted, and an adversary could find those. This is why many governments around the world are investing in this technology, not because necessarily they see the good for society,

although I'm sure that they do, but primarily because they see it as an arms race. And the first government to get the new weapon, the new RSA-breaking shore-capable quantum computer, they'll have a major leg up over every other nation in the world. Yeah, you were telling me a little bit about something called Catch-in Store, which is the idea that there's a great deal of incredibly valuable information out there in the world. It is all currently encrypted, and that's all well and good until we get a Shore-capable quantum computer that can wrench that information out from inside of it.

that. So that's a, that's a very bad milestone. I guess we're sort of watching approach on the horizon. Yeah, absolutely. And there's two things about this that it's important for people to take away. One, catch-in store means that we're not using quantum computers to break encryption today. That's not happening. We need much larger, much larger computers than what we are able to currently build. But the other side of that is that, yes, you're absolutely right. As soon as Shore said one day it may be possible to break RSA. Adversaries see it as a long-term

hacking strategy to just catch these things while they're still encrypted with what RSA, which we know now to be vulnerable, and just save it. Many of them will be added. Many of those messages will be out of date, but some will still contain pertinent

military secrets, critical infrastructure. Those things don't move in 30 years. So those, those may, the catch-in store is betting that at least some proportion of the messages they're grabbing today will still be valuable when a shore capable quantum computer is switched on. This is asking a completely unreasonable question that no one could be expected to answer. But here we go.

How close do you think we are to a sure capable quantum computer? Is this a thing I should be waiting for the news alert for or is this decades down the road? I would put it at decades, but I got to say, anybody who claims they know when it's going to happen is selling you something, and that's not what I'm here to do. The critical number, the number that tells you kind of how far away we are roughly is the number of bits that these computers have. We call them Q-bits or quantum bits. Today, we're building devices on the order of a few hundred bits at best. It's not even clear yet how well those couple hundred bit devices are performing. They're so new.

And to keep in mind, like, the number you would need to achieve sure that ceiling is constantly being brought down. People are finding more clever ways. The feet requires fewer bits. And I've seen estimates anywhere from a couple hundred thousand to millions to the largest I've ever seen is a billion or a couple billion cubits. So somewhere in there, keeping in mind that the devices are getting bigger and the ceiling is coming down. When they meet in the middle is when you'll get that news alert. But I put it as very far away currently. Trying to understand the difficulty of what I'm asking, what a sure-capable quantum computer represents.

Every single one of these cubits, they're built with different architectures. But the one I like to latch on to because I think it's easiest for people to picture in their minds is one cubit, is one atom, just a single atom that you are now using in place of that transistor, we talked about before. So you've swapped out your transistor and replaced it with a single atom. And now I'm telling you you need 100,000 of these atoms at the lower end, all arranged on a table, capable of being controlled at 10 to the negative 6 or so precision, control, meaning you hit hit them with a laser.

So that lasers intensity, its timing, its frequency, all have to be tuned just so for one cubit. And you are choreographing a light show of lasers on 100,000 of these atoms with that level of precision for hours of compute time. And if it goes wrong, if you hit that atom with a slightly hot, hotter laser or for slightly too long or those errors are going to start creeping in and collecting.

And now you can kind of have a sense of what it is we mean when we're saying build a computer that's that big. That is the engineering challenge. I'm about to reveal the incredibly low ceiling on my technical knowledge here. But just to try and make it make sense to me, we've got this light show of lasers pointed at these atoms. A traditional computer, it's a transistor, it's zero or one. How should I understand what is happening inside of these qubits that's different than that zero or one, that on-off state of a traditional transistor? Yeah, it's a tricky thing to get folks to wrap their heads around. The picture that I show to my students, and I recommend you just to try and keep in mind, imagine a transistor, the state space of a single bit. It could be zero, could be one.

Imagine those as two points, one on top of the other. that's a that's all that's that's what a transistor is you put a bunch of transitions together it's the combination of all possible up and down positions of these classical states that's a bit string a cubit is where you still have zero and one you still have access to those classical states but now imagine a globe where zero is the north pole one is the south pole but you can go anywhere on planet cubit that you want Hmm. Those are all the possible states of just one cubit. And as soon as you go from one cubit, you can't even picture this in three, in three dimensional space. You need more. But this is called the block sphere. It's the main way that we've tried to represent the state space of a cubit. The thing I try to tell students is imagine planet cubit. Imagine this sphere as an actual planet in space. And that planet has different types. of mineral resources at different points on the surface.

The two types of material resources are called superposition and phase. I'm not going to just really, we can get into what those two things mean, but just imagine they're like, you know, two different types of ore, and every point on that surface has a different balance of these two types of ore. Those resources are what you're feeding into the quantum computer that we believe are part of the picture of what makes quantum computers special. Because you have access, you don't have access to these with just your North and South Pole. You only have access to these critical quantum resources when you have the full globe.

And these resources, it seems, are helpful in computing problems faster. One of which, we were talking about this earlier and I got on this groundwork laying tangent. you were talking about RSA a little bit. We were talking about the shore capable quantum computer and how far away that is. On the flip side, I'm curious, at that moment when that advantage emerges, RSA becomes vulnerable. And as such, almost everything it feels like that we're doing on a computer right now, all that information becomes theoretically vulnerable. Am I right in that?

Are there other types of encryption that we are using based on traditional computers that would be less vulnerable to this? is it pretty much once that door is open, retroactively, everything is visible? No, so there's hope. And this is the field of post-quantum cryptography. A lot of folks in that area who know a lot more about that subject than I do. This is the kind of the defining challenge of a lot of cryptography research now is, how do you come up with a cryptographic protocol, something beyond RSA, that is still secure to all the known tricks with a classical computer,

and also a quantum computer. And to keep this in mind, the challenge here is we can't even prove RSA can't be broken with a classical computer. That's not a proof that is out there. That's just there's a lot of payout if you could hack it. It's been the standard for decades,

and no one's hacked it. So we believe it to be uncrackable, but we don't know for sure. So every time, I've heard a couple of different, stories of this, several times now that someone thinks they've found a good post-quantum cryptographic scheme. They've done all their research on it, but the ultimate test is put it up, put something valuable behind it, and then let the whole community of hackers attack it and see if

they can crack it. And there's been embarrassments. There have been embarrassments. I don't have a specific, this is back in maybe 2022. But in 2022, I was at a conference where we were talking about quantum policy. And I was hearing that it had only taken a couple of days for the latest candidate to get cracked. Interesting. So it's about, you know, coming up with a good candidate with all the tools of research that you have. But ultimately, the ultimate test will be probably the audience of this podcast. Can they hack it? And if they can, it's no good. Interesting. So it was less secure than RSA, which up until this point has proven pretty much unhackable. Correct.

A quantum computer with advantage would be, I'm going to undersell it, extremely valuable. Yeah. And for a while, probably extremely rare. And as such, I would assume it's going to be controlled by a pretty powerful actor that would get to it first. Digging to the ethics, I'm guessing at kind of the top, top level. What are the ethical concerns? that that type of situation would result in. Yeah, and let's just be clear.

We were just talking about Shore, which needed all these hundreds of thousands of cubits. Quantum advantage, the number of cubits you need changes based on the application. So factoring happens to require a very big computer, but you might be able to achieve advantage in chemistry or finance with the devices that people are building today in the next few years.

This is the area that I specialize in. This is what we call near-term quantum advantage that doesn't require quite as much engineering what we call fault tolerance that Shores algorithm requires. So pre-fault tolerance, pre-all of that, what can we achieve in the next few years? Can we achieve advantage?

That's kind of where my research as a graduate student was focused. And you bring up this point about critical, you know, limited quantum resources, maybe only one company in the world will have this device that can deliver advantage. And kind of a thought experiment that was really my first example from my field.

So I specialize in a type of quantum algorithm that uses machine learning to figure out the correct computation to run before you run it on a quantum computer. Or there's kind of a back and forth. There's a classical optimizer that's trying out a different, tells the quantum computer to try a different,

front circuit. Quantum computer tries it, evaluates the performance, sends that back, you've got an optimization loop. The two prime candidates, the most popular versions of this approach, this variational quantum algorithm approach, have applications either in chemistry, drug discovery, or general purpose optimization, let's just say finance for the purpose of the thought experiment. My concern was that, you know, if I were to name my number one best application for society, and that's really what I mean by most ethical, when I talk about ethics, I mean how do we ensure this technology is the most beneficial for society? The tagline of my organization quantum ethics project is that our mission is to try and ensure quantum benefits everyone and

harms no one. L lofty goal, but that's the mission statement. And so if I were to name my number one benefit everyone application, it would be to accelerate the rate that we can discover new drugs. Right. We could potentially discover, you know, current, you know, new drugs for the supersede preexisting drugs that maybe have fewer side effects, or potentially just discover drugs that cure things we can't cure it right now. I think that's unarguably a very good use case. On the optimization side, a lot of people are interested in the fact that in financial markets, people are already using tools out of the quantum toolkit, like the Monte Carlo method. It's just a type of algorithm for predicting really complicated systems.

So they use this stuff in finance all the time. And the hope is then that, well, if quantum computers are on the scene, maybe they'll be even better at predicting which investments to put in a portfolio. So now pit these two against each other in the free market, where you've got one company controlling the quantum resources, and their goal, you know, the legal fiduciary responsibility of any corporation is maximize investment return for your shareholders. You've just spent decades building this extremely expensive device. you need to start making profit. So you've got a certain number of hours that that computer can be put up on the cloud. Certain number of hours per day

outside of maintenance and retuning and whatever needs to go into keeping that device running. So say you've got eight hours that you can sell to whoever wants to use your device, who wins the bidding for, the hedge fund industry or anyone else? I think the hedge fund industry is something around like $2 trillion. And so I think that there's, you know, my concern was that, you know, I looked around

at all my colleagues, all these brilliant people with so much that they wanted to offer humanity. And I, the basic question was, you know, is this it? Is this what all of our creativity and time and blood and sweat and tears and practical exams and everything is going to go towards building Jeff Bezos or somebody else's some billionaire's newest favorite money printing machine? Well, and that's interesting because the money printing machines are competing against each other. A company goes, obviously pharmaceutical companies

are competing against each other too. Sure. But it's sort of advantageous to all of us if they both happen to crack the new drug that solves the old problem. That's not bad. I'm not mad if two of these companies find the solution at the same time.

Totally. I don't really gain anything if one hedge fund makes more money than the other hedgeman. I guess if I'm invested in it in a very abstract, like I guess there. But you see what I'm getting at? Like, it's absolutely. One is, one is advantageous to all of us. Well, what you're describing is that, and this is one of my things that keeps me up at night the most as someone who worries about this stuff, is that quantum computers, um, they might be society's next biggest contributor to wealth inequality. It's already

the issue that I think a lot of people are really frustrated by the cost of living, inability of my generation to buy a house, these sorts of things. And we are already put up against these tech billionaires who have used machine learning and all the fanciest tools of the digital age to enrich themselves to a huge degree at the expense in many cases of the rest of society. And so when we're thinking about this new gilded age, which has been many ways driven by the technologies of 20 years ago, Internet, digital computing, kind of making that more and more available, the day when quantum enters into that space, we have, it's possible everyone's going to forget that they were being developed to help people. It's cool. I was previously just scared shitless about AI acting as a gasoline-esque exercise. on wealth and equality.

And now I get like another cool second thing I get to be anxious about. So many things. But like that exact piece is what got me started into all of this in the first place. Because the discourse in quantum machine learning is, well, how do we make chat GPT even more effective at what it's doing right now? How do we make these algorithms even faster, even more efficient at... reallocating everyone's money into the hands of like six people. Think about the last time you heard a breach story on this show. It always starts the same way.

Someone somewhere saw something too late. An alert buried, a signal missed, an SOC that just couldn't keep up. Arctic Wolf set out to solve that problem by rebuilding security operations from the ground up for a world where attackers are already using AI. They created the Aurora superintelligence platform with fully agenetic system powered by the swarm of experts. Instead of single-purpose bots or lucky-guess LLMs, this swarm is full of deterministic agents that handle whole entire workflows. Humans stay in the loop and on the loop to validate the critical decisions and keep everything trustworthy. And all of this is just off running on their secure operations graph. A constantly updating intelligence engine fueled by more than 9 trillion telemetry events every week and over a decade of real-world incident response.

The system reasons on real signals and real context not synthetic training data. and the result is the new Aurora Agent SOC. It's the first SOC that is agent led by design. You get agents that coordinate, agents that investigate, agents that respond at machine speed, and hundreds more that automate the repetitive work that normally buries human analysts. Arctic Wolf didn't try and bolt AI onto an old model.

They rebuilt the model entirely. What makes it even more effective is how it works with Arctic Wolf's concierge experience. The team brings customer-specific context directly into the platform so every AI-driven decision reflects your environment instead of generic assumptions. The automation frees your concierge security team to focus on higher value strategy and proactive risk reductions while the agents handle the grind. If you want to see what trustworthy, production-ready AI and security operations actually looks like, go to arcticwolf.com slash hacked.

Ever feel like cyber threats are evolving faster than anyone can keep up? Last year, 2025 was nothing short of a record-breaking, year for major breaches, from sophisticated ransomware operators to AI-enabled attacks that turned defenses on their head. Organizations around the world saw headlines they never expected and cybersecurity teams were tested like never before. But here's the thing. These incidents aren't just news headlines. They're learning opportunities. And that's why Arctic Wolf is hosting a live webinar on February 5th diving to the most impactful breaches of 2025. Their field CTO and security leaders are going to unpack not just what happened, but why these attacks succeeded. And

most importantly, what businesses can do to fortify their defenses for it's too late. You're going to walk away with real insights in how threat actors are evolving, how defenders are responding, and what strategies can help you stay ahead of the next big breach. It's not fear-mongering. It's practical, actionable, intelligence from experts in the trenches. Register now at arcticwolf.com slash hacked. So you've painted a bit of a picture of this moment where one of these things that offers some kind of meaningful near-term advantage to people.

It goes online. It's AWS. We can all just, whoever throws more money at it, gets access to it for more hours of the day and takes on this sort of disproportionate advantage in whatever marketplace they're existing in. If it's a hedge fund,

they're just going to start printing money. If it's pharmaceuticals, maybe they make some cool discoveries and start printing money. What's the longer tale of it? Like, that's the moment it turns on. It's incredibly scarce. if you kind of had to sort of run that simulation out a little bit longer, it becomes more plentiful.

Where does it go from there? Yeah. It's, you know, I don't identify as a futurist. I think even if I try to look a year or two into the future, my vision is murky. But I think something really interesting has the potential to happen. Because walk us back to ARPANET, when the Internet was being created as a way for researchers to communicate. more efficiently with each other. The folks building ARPANET had no way of predicting

Instagram body dysmorphia, the teen mental health as a cause of these tools. So, you know, we're thinking about like a graph with as farther into the future are more and more degrees of removal from what we know, but also greater and greater intersections with technologies that don't exist yet. And quantum computing isn't the only quantum computing

isn't the only quantum in the revolution that's happening right now. It's also quantum sensing technologies, which people want to use for a number of different applications. And the quantum internet is also something that is out there. What I think is both exciting and scary, depending on how you look at it, is how are these technologies going to start coming together in weird ways? How are you going to say a quantum computer isn't the only computer, component in the device, but it's a device with quantum sensors, quantum networks, quantum computers, and stuff I can't even name right now. What are those things going to be doing? I have no clue. But I think that the thing that is what I come back to whenever I think in this space and what gives me a lot of cause for optimism is that like take a second, think back to, you know, you see these. photographs that come out of the 1950s of people smoking on airplanes. And that just seems ridiculous to us now, where we figured out what was bad about smoking, we regulated it as a society,

we decided that certain things weren't going to be possible. We got to choose how we are going to engage with this substance or this technology. I don't think we're there even for the technologies of classical computing and networking. We're still, we're just, now, you know, having these discussions about how these technologies have come together to cause teens to struggle with their mental health or to destabilize, you know, civic discourse. And we have yet to solve those things. The piece of this that is so hopeful about quantum is that it's a lot of it is just a new version of a tech we've seen before. When Arpanet came around, there was no Stone Age internet we could use to infer some of the potential

pitfalls, but we now have all the lessons of the modern era to guide us in the development of quantum technologies. So that mystery machine, which has all these intersecting components that's completely unpredictable, we still have the choice how that future machine will be used because we have the ability today to decide how we want to relate to these types of technologies, how we want to develop them with a certain sense of responsibility, a responsibility to not only ourselves and our neighbors, but our children and future generations, that these technologies should be a light for all humanity.

And they don't have to be these engines of inequality. We talked about some of the potential applications. We talked about just, let's just print money with the fancy new computer. computer. Let's develop some new drugs, which sounds great. What other big, broad category should people be thinking of this as having application for? Well, something that's really interesting that we've been looking into in my organization, which sounds on the surface like a very ethical use of quantum, is quantum for trying to help address climate change

or the, you know, rising temperatures. And I'll also, there's, is there's, is there There's a lot being done here. I'll speak about two specific, or a couple specific applications that we've specifically looked at. So three are coming to mind. One is to try and use quantum computers to find ways to build batteries for electric vehicles that either charge faster or store more energy or both. If you're trying to think about the challenges that face electric vehicles, I mean, one of them is that gas-powered vehicles, you can refill

them in like five minutes. Stop at a gas station, fill it five minutes, drive it. You also have a very wide drive radius, and I know that people are trying to extend that radius as much as possible. It's a very interesting engineering problem. So the battery is at the core of how good your drive radius is and whether you're having to sit at a charging station for hours, which is often infeasible for a lot of folks who aren't on a long road trip. We need to get to work. And so that's one, using quantum to try and find new. batteries. The next would be we've got these interesting nascent technologies like carbon capture, which really don't work. They don't grab carbon nearly fast enough to justify all the hype.

But if you could, you find a better catalyst in your carbon capture process that pulled a lot more carbon out of the air a lot faster than good news. We don't even need to switch from gas, carbon producing devices. We could just pull it out as quickly as we can emit it. We're nowhere near that. So some people think quantum computers could be used to find better catalysts for those processes. The problem is, the problem that I draw with basically every algorithmic solution to the climate crisis that I have studied, is that all of them require, we mentioned a little bit ago with Shore, a shore capable quantum computer, the secret sauce that makes it work, is something called fault tolerance. It's where you've built a system that tries to eliminate those errors, those drifts of,

oh, you had it with too much, your laser was a little too hot there, it lasted a little too long over there, they're going to start drifting. Fault tolerance is a way to correct those drifts as they are happening so that you can still do these large calculations. We talked about how far away fault tolerance is. 30 years, maybe. I don't know. As soon as you put something at 30 years, nuclear fusion has been 30 years away for 60 years.

It doesn't seem to get any closer, even though we hear about certain milestones that have been hit. So if you're telling me here, I think this is a really interesting ethical kind of example, because it sounds like using quantum for climate would be a great thing to do. But you're here trying to tell me

that we should be funneling money into quantum computing for the climate on a solution that won't even be here for another 30 years when the climate crisis is an imminent threat today? So there, it's just kind of a discussion. And, you know, there's folks who've been working on new solutions. So it's not like this is the end of the story. I'm very interested to see whether solutions can be produced that are much nearer term.

If you can give me a near-term advantage in the next year or two, that is a very different conversation than asking me to take, say, a million dollars I could be spending on solar, wind, and energy storage to trap a bunch of atoms that won't do anything valuable for several decades. So those are some examples. I mean, you use the word hype. And I don't really typically think of hype. It has much more of an ethical dimension in the context of this technology than I think of it normally is having. Normally, hype is a great way to separate an investor's money from their pocket.

It's a great way to get a press cycle going. It's a good way to get a headline. But here, there's sort of an opportunity cost. How much is fighting misplaced hype part of the ethical project of quantum computing? So this is actually my focus of specialty. I wouldn't consider myself to be an expert on quantum for the climate. I do quantum algorithms, but my area of specialty is this question of how overhyped is quantum really? How overhyped is quantum machine learning? And the way you try to quantify hype is by

benchmarking. So my organization, we've actually just launched a benchmarking project for those variational algorithms I mentioned previously for the drug discovery and the finance. We are doing this sort of open source, anybody can contribute benchmarking project in partnership with the Unitary Fund, who's a nonprofit, and they have a benchmarking community already kind of in place, where we're trying to kind of give an answer. And the trick has been coming up with benchmarks that regular people can understand. And when I say regular people, I mean like people with a reasonable level of education. They're not experts in quantum, though.

And the challenge has been most of the benchmarks for the quantum field have required quantum expertise to understand. So someone hands you a number, like quantum volume, gate fidelity, T1 and T2 time, that means absolutely nothing to a potential user of a quantum device. It just says, I just want a linear system's set of linear equations to be solved. How well can you do that? And how big a system can you solve? That's what our benchmarking project is about.

It's a big focus. And the reason it's a big focus, and I've actually written columns on this, is that all of those good uses of quantum that I mentioned before, those are going to be delayed by a lot. If we today promise are the capabilities of our technology in the stratosphere, when their real life capabilities

are here in the dirt, which is, in my view, what's been happening up until this point. And there's a historical precedent here. When AI, the term AI, was coined, it was the 1950s. There was this not quite a conference, but a gathering of some of the greatest minds of the time. I think von Neumann was there. Like some of the people who founded information theory and were working in this space. And their goal in the 1950s, what they put on their DARPA,

grants was they were going to build an artificial general intelligence that rivaled a human brain. And they promised that on the backs of vacuum tube computers, which is all they had at the time. And they got a lot of money from DARPA and other places. And they didn't deliver. And then AI became joke science for several decades. It experienced what we call an AI winter, where all the funding got pulled. it was still, we know today, AI was perfectly valid area to push, but they were promising too much, but having too little to show for it. And AI actually went through several of these cycles.

They had another one in the 80s where there was a resurgence in interest, they overpromised, under-delivered, and then lost all of that funding. And critically, they lost the credibility. It takes decades to fix those sorts of problems where if today we have venture capital, who are putting money into quantum, if we continue on this track of over-promising and under-delivering and critically not showing these non-experts a benchmark that they can freaking understand, we are going to hit a point where they just decide we are, as you said, separating people from their money, which is not what I think is happening. I just think we have faced a challenge where quantum sounds so hard so advanced, that I don't think folks in my field even think that a benchmark that regular

people can grok is possible or a reasonable thing to try for. And I think that that's very dangerous. I think we should hold ourselves to a higher standard of science communication and find ways to make it make sense to our legislators and our investors and our parents. Like I would love for someone to be able to like I'd love for my grandma to be able to understand how well a quantum computer runs in terms that she can understand.

And I think that explaining things at that level is perfectly possible. But we're just now, I think some folks are just now starting to figure out how to go about doing that. And we're trying to lead that at the QEP. I'm sure it's a massive project. But what does that grandma

literacy level of a benchmark for this technology, what might that look like? It's the same number we just talked about with Shore. How many qubits? Except it's not how many cubits can you build on an algorithm. It's how many cubits can you, how you solved your problem. How big a problem was it?

And we can quantify that in cubits. So to give you an idea, linear system solving just came up. One of my students completed a benchmark on that for the first phase of the project. If you want to solve a linear system of equations, you'll need north of 100 cubits. Anything less than 100 cubits, and I think you're guaranteed to have nothing of interest to the advantage community. So we knew from the literature that the number one, like the best implementation out there, 10 cubits. Currently, that's the, that's the crem to la crem. 10. We tried to do our own and we did worse than that

Or my student did worse than that And the reason is because these things require a lot of optimization of hyperparameters. But without getting into a lot of the technical details, that number, 10 versus 100, is the number I want people to be able to walk away from, walk away with it. And let's just be clear,

the solution quality is also an important factor. Like you say you solve the problem. problem. How well did you solve it? We, my, this is very hand-wavy, so don't quote me on this, but my touchy-feely vibe for what would be needed in the field of linear system solving is something on the order of 10 to the negative 6 in terms of how far away your solution is from the one they want. So you need the number you're spitting out of your algorithm for the solution to the linear systems to be only different from the true value six points to the right of the decimal. We only tried three points to the right of the decimal and still did not do, yeah, we couldn't do better than 10.

On realistic noise, we couldn't do very well at all. And that poor quality, you know, one's first thought might be, well, are we just bad researchers? And, you know, my student as an undergraduate, I don't have a PhD. I'm not going to claim to be the greatest researcher in the history of the world. But we deliver, we implement to these algorithms as written out of the box. just like this is how the algorithm was supposed to be done. We didn't do any advanced work on this. We didn't throw a team of PhD students at it for several months.

We just had a student implement it to see how well they could get. And I think there's value in those kind of quick and dirty estimates. So that when I feel that with these sorts of benchmarks, we can go to a lawmaker or an investor. like investors often have a specific application in mind. And I can say, and the critical metric of interest, what all this is bidding towards is, it doesn't matter these kind of low-level hardware specs,

like quantum volume gay fidelity in T-1 and T2 time. I haven't even defined those. We don't need to talk about that. The thing that matters is scale. How high can you scale the algorithm that your user wants to run? And then tell them how far their, their algorithm is from how big it needs to get. That's our benchmark.

And we hope that it'll help kind of clarify and avoid the issue of a quantum winter if we're able to as a community come together and be honest about how well these things are performing or not. 10 out of 100. That is one of the more useful, tangible ways of putting – because that's one out of a 10. Yep. You are a 10th of the way there. You are a 10th of the way there.

It's like, are you looking to invest money in something? That's very, that's very interesting. And Mike, my, I have colleagues who have criticized that benchmark with some very valid points. That upper limit of 100 is going to get higher. That, that 10th value is not an objective derived thing. It's a fluid, squishy feel. It's not a, yeah, it's, it's a mood.

It's not rigorous in the mathematical sense, but I do think that it helps to deliver some intuition about where we are in this particular moment. It'll change in the future. And I'm really interested to see, like, plot how it changes. How does this, how do the capabilities of quantum computing compare to the capabilities of our ability to simulate quantum computers with classical methods?

This is the race that really defines my field. And to give you an idea, there's a fun little story of big companies getting embarrassed in a big way. In 2019, Google publishes this, they claimed they had achieved advantage for a device with 53 cubits. And their claim was that they calculated something in three minutes what it would have taken by their estimate, a supercomputer 10,000 years. And in a few months, people had figured out that their estimate of 10,000 years was a really dumb estimate. They were not trying their hardest to bring that number down, and people found ways to do it in a month.

Last year, IBM made a similar claim. They had 127-bit machine, so that's an important measure for everybody. 127 cubits was the record-setting paper for last year, where they could. claimed, look, they didn't claim advantage, they claimed utility, which is slightly different. I'm not even sure. It has some sort of rigorous difference, but they were effectively claiming advantage for all intents and purposes in the community. That's how the community regarded their result. And it took, you want to guess how long it took people to classically simulate what they did?

Ooh, hit me. A week. Or a couple weeks, on the order of a few weeks. Sure. It's salesmanship. The point being when these companies make these claims, they aren't doing, like, if I were to be running a group, I would have my group that worked on the quantum algorithm side competing against an internal adversarial group trying to simulate all the stuff they're doing. Cool.

And have these two compete. They weren't doing that because it seemed that the TensorFlow Network community, the folks who work in classical simulation, were able to very quickly reproduce everything that they were doing. And so that's what brings that number higher, that 10 out of 100. That 100 threshold, it moves higher when people simulate an even bigger system for a particular application. It's the needs of the rigors of science and the needs of a for-profit company that wants to be able to make the big flashing announcement, just being completely at odds with one another.

Well, the big flashy announcement isn't just about flash. It's about, I'm sure, these companies did very well in fundraising shortly after those announcements. I'm sure their stock price went very high. But how many times are they going to do that without producing any calculations of value before the investors start to feel like what you said previously? They're just being separated from their money. You're responsible quantum consulting. you're working with companies,

you're helping them navigate hype and reality and what's actually going on here, what are some of the biggest misconceptions that you see when these companies come through the door? What is the number one thing? You're like, if you're wrong about something, you're probably wrong about this. Yeah, well, I think there's this sense,

and I actually think the best way to frame this is when you met me, the Quantum Village had kind of three main areas. It had where I was, it had an area for talks, and then it had an area for companies. And if you go to the companies, you get a very, very rosy picture

of what Quantum is capable of doing in the near term. They're selling something. The value that I think my organization provides is that we give that kind of hype-free expert perspective on Quantum that helps you kind of cut through

all the salesmanship. And the biggest misconception I have from a lot of people is that quantum machine learning is magic and we'll do on all sorts of different things. And that's exactly why the benchmarking work that my organization is doing in quantum machine learning, I think is so valuable because we're trying to help people understand. Like, look, you're a company, you may have a valuable pathway to incorporate quantum. And we want to help you find that specific to your company's needs. but you are more likely than not to take a very expensive stab into a dead end.

Because it's hard to understand the material, and it's hard to sift salesmanship from truth. And so that's why the QEP, you know, our focus on hype is so important, is that we're really trying to help make sure that companies can understand exactly what, if any, value quantum may have for them. And if so, because we have not just ethicists, but most of the people in my organization are quantum people by training. So we can help folks navigate exactly what quantum looks like to make sense for their business. I think a lot of people do that for cybersecurity and they do it well.

But there's a lot more to quantum than just cyber. And that's where we're hoping to provide the most value. Maybe I'll wrap up here. You used a phrase earlier, talk about the most beneficial. for society as this, I think there's no sign of this stopping right now. I don't think we're maybe another winter of this happens, but right now it feels like there's a lot of attention on this. How do we make sure that this tech is developed and used responsibly moving forward in service

of that goal, the most beneficial for society? Yeah. I think the conversation is going to happen at a number of different levels. The first level, I would say, is how we develop our workforce and I specifically was the former deputy director for workforce development for Q-Sense which is a the quantum center based out of Boulder's the University of Colorado Boulder and so I've worked specifically in this area and I think right now you know I come from a really rural low-income

background and folks like where who come from the places I come from we don't typically see the benefit of technological revolutions even when they work for some. They don't typically work for us. So making sure that our workforce development focuses on communities who are more rural and still, but nevertheless still have a lot to gain is going to be really crucial. I formally managed the Quantum Research Exchange, which tried to provide access to quantum internships and career pathways for community college students. And as a community college student myself,

I thought this was incredibly valuable. So that's the first level. I think we need to have a much more, I think we need to do our workforce development a bit differently, where previously we've really focused kind of top-down, Ivy's first, and it trickles on down from there. And then the second piece, when we're talking to individual researchers

or graduate students, postdocs, the individual level helping folks understand how does my work connect to the ethics conversation? And this is something that we did. We worked with the Center for Quantum Networks in Tucson, Arizona, to develop, to work with three different, we worked with three different PhD students

all working on some level of the quantum internet stack. And we developed a research proposal that would integrate ethics questions into the work that they were already substantively involved with. And the idea here is that ethics and these sorts of, this, really, ethics for me isn't the word I use too much anymore. It's responsible innovation is the core of our mindset here, where everybody who develops the tech should see themselves as Oppenheimer. They should see themselves as working on the Manhattan Project of the 21st century. and having a similarly great responsibility to ensure that the technology that they unleash is not what we saw in World War II. And then finally, on the policy level, I think, you know, we talked about the regulation of smoking in the 50s.

And I know the conversation is trying to happen with, I guess, some of our more elderly senators holding it back around how we are supposed to regulate tech today. But I think trying to work with folks who are making these laws, thinking about ways that we can ensure, for example, the government is about to approve its five-year renewal on the Quantum Act, which was the United States's multi-billion dollar investment in quantum five years ago. It's about to be renewed for the next five years. And that's going to be the lifeblood of many centers, like the Center for Quantum Networks that I worked with previously, as well as a lot of the major companies. that are working in this space rely on federal funding. And I think the conversation that we had about these companies who are letting, this future company that might have a quantum computer that basically tries to make as much money as possible. I think that folks who have paid the taxes that paid for some of those advances should see the benefit. And we should have some sort of requirement that technology that was used for public funds is used for public good. Joan, I clocked it two minutes into sitting at that table at DefCon, you're a wonderful educator of all of this.

So I just appreciate you sitting down and making it make sense to someone like me. Yeah, happy to. I love having the chance to chat with folks, because I think Quantum has stayed too long in the top of a very snooty ivory tower. Some people in my field, I meet them, they're like, I went into Quantum because it's the hardest thing in the world and no one but me will ever understand it. And I think quantum, by definition, should work for everybody. It should work for the little guy. So we'll find ways to make it make sense. And then, you know, people will be able to have a say in how their future is shaped. Appreciate your time, Jim.

Thank you.