Quantum computing won’t be an incremental improvement – it will be a step-change in the power of the technology that underpins a huge part of our economy. Exponentially faster computing won’t just help us solve problems more quickly, but it will also allow us to tackle problems that have been impossible to answer and find answers to questions we did not know to ask.
Chad Rigetti, founder and CEO of Rigetti Computing, speaks to Azeem Azhar about just how revolutionary quantum computing will be.
They also discuss:
- How quantum could usher in a new computing revolution.
- Why having the lead in quantum computing is turning into the contemporary equivalent of the “space race.”
- How quantum computing could change how we do science.
@Azeem
@ChadRigetti
Further resources
‘Building A Quantum Computer with Light’, (Exponential View Podcast, 2021)
‘Making Quantum Computers a Commercial Reality’, (Exponential View Podcast, 2021)
AZEEM AZHAR: Welcome to The Exponential View podcast where multidisciplinary conversations about the near future happen every week. Now as an entrepreneur and investor, I’ve been an insider in the technology industry for over 20 years and during that time I have observed that exponentially developing technologies are changing the face of our economies, business models and culture in unexpected ways. I’ve written about those phenomena in my book, The Exponential Age, and I return to it weekly in my newsletter, Exponential View, as well as on this podcast. If this is your first time listening to my podcast, welcome, and if you’re a returning listener, I am so happy to have you back. Tell me how we’re doing by leaving a rating or a review on your app of choice. This also helps other listeners find me. I talk about general purpose technologies, paradigm shifting innovations that change the way we work across many fields, and with good reason. Developments like artificial intelligence and 3-D printing have an outsized effect on human progress and change our world at an exponential rate. Today, I want to dig into one of the most revolutionary general purpose technologies, quantum computing. Quantum computing represents a step change in our ability to solve difficult problems, in the same way that the development of the modern computer did back in the 60s. Modern computing relies on on-off one or zero bits to represent information. Quantum bits or qubits can represent several states at once, using a phenomenon known as superposition. This allows them to work much, much faster than classical computers, and means adding more qubits makes quantum computers exponentially faster. Today’s guest runs one of the leading companies in the space. Chad Rigetti is the founder and CEO of Rigetti Computing, building quantum computers with a stated aim of solving the world’s most intractable problems. The company is set to go public in 2022, becoming just the second pure play quantum computing company to go to the public markets. Chad Rigetti, welcome to Exponential View.
CHAD RIGETTI: Thanks Azeem. It’s really great to be here.
AZEEM AZHAR: Quantum computing is such a remarkable technology and also a remarkably hard technology for people to get their heads around. How would you describe quantum computing in ordinary language?
CHAD RIGETTI: Quantum computing is one of the most exciting and transformative emerging technologies in the world today. It is a disruptive form of computing that is going to bring solutions to problems that are today well out of reach of any ordinary or what we might think of as classical computer. As such it also represents just a phenomenal area of economic opportunity and an opportunity to leverage a fundamental driver of economic development in advanced computing technology. And because of that, it’s obviously seen a real level of investment and excitement around the world and in America, the U.S. specifically, over the past kind of half decade or so and we’re just thrilled to be kind of at the leading edge of all that.
AZEEM AZHAR: We are probably eight or nine years into sort of this current wave of commercialization of quantum computing. It’s quite interesting for me to see how the debate has changed. It seems to have changed from not if to when and you’re starting to see analysts and consultants scale what the opportunity is and it’s sort of $500 to $800 billion sitting there if you believe some and I guess some think it’s probably going to be even higher than that. As someone who’s been on that journey for a decade, I mean how do you feel now compared to perhaps where you felt when you were scratching ideas on the back of a napkin?
CHAD RIGETTI: What we’ve seen in the past few years that you just described in terms of really a broader recognition of the near term possible impact of quantum computing has been part of my kind of founder thesis on quantum computing since I started the company. There’s really fundamentally technical reasons why I think that’s happening and there’s also just general economic forces at play. So we’ve known about the potential for a long time. We’ve known about the possibility of exponential acceleration for high impact problems for a few decades, but around 2010, the first … What you characterize as kind of serious industrial efforts to build quantum computers started to get going. Starting to establish not just the scientific principles but an engineering roadmap and plan to solve the enormous problems that really needed to be solved at that time to bring quantum computers to a state where they could be reliably programmed, deployed, and even tested and benchmarked against kind of classical computers. And in the early parts of the 2010s, what was happening was there was just incredible progress being made on the hardware, on the algorithms, on the error-correcting codes, and all three of these got on a trajectory where it started to appear as if if you’d map out those trajectories over another half decade or decade, that the technology is going to start to be able to solve practical problems and over the ensuing decade, that is what has happened.
AZEEM AZHAR: What’s fascinating here is that a little more than a decade ago, we would have looked at this problem, we would have struggled to understand how we can build the qubits that are the sort of fundamental transistor as it were of the quantum computer, the elementary unit, and turn those into processors. There was going to be some complexity about how you get useful algorithms that can do things and then there was also going to be this issue of correcting the errors because these qubits are … They’re tricky little beasts to wrangle, and I think what I’ve seen of course is there are now four or five different technical approaches to building those qubits, you’ve got the sort of superconducting qubits, trapped ions, photonics, some other approaches and there’s progress being made across all of them. What has surprised you about where in those three vectors there has been remarkable progress or has it been across all of them?
CHAD RIGETTI: I think the progress on the kind of pure hardware side … I don’t know if it’s been surprising but it’s been I would say the most fundamentally promising, and the reason for that is that it all starts with the hardware. That’s the fundamental core ingredient that is needed to leverage the progress in the other ones and so the progress on hardware where you’ve now seen multiple different players demonstrate systems with … For now let’s just say tens of qubits, and going from one to two was a big leap, and that was really the first few years of the 2010s. But now with systems at five to ten to twenty to thirty, forty qubit level, as the systems have scaled in number of qubits, the fidelities have improved as well. A combination of those two things, scalability and scaling being demonstrated, while also driving and enhanced performance on the systems. That’s what we need to see that that’s happening and that has been the most promising thing overall I think from my perspective.
AZEEM AZHAR: Is that the point that … These qubits are noisy and in the olden days, by which I mean three years ago, we would say, “Well the qubits are noisy so you need to have hundreds of physical qubits for every actual precise logical qubit to do the work and if you need 100 logical qubits to do something really useful, that means you need hundreds times hundreds which is tens of thousands or hundreds of thousands of physical qubits and these things are hard to make and they’re hard to wire up and they’ll be hard to control. If the fidelity is going up, that ratio of the number of logical ones you need for each physical one declines and that makes the problem more tractable. Is that what we’ve been seeing?
CHAD RIGETTI: What has happened is the efficiency with which we’re able to define and implement algorithms and protocols on the physical quantum computer have just made tremendous progress. So there’s been improvements at everything from the encodings to the efficiency with which the quantum algorithms can be executed, that is the circuit depth or the number of cubits needed to represent those problems. The way I usually think about this is if you go back to the early days of the 2010s, there was kind of Shor’s algorithm that was out there way over the horizon that needed ballpark a million or more physical cubits to be implemented, and then there was this notion that if you could even build a quantum computer with 30 or 50 physical cubits, you’d have a great deal of difficulty simulating that with any classical computer, simply because of the exponential scaling of the state space of quantum computers. So there’s this impossibility to simulate beyond 30, 50, maybe 100 qubits at the physical level, but then the known algorithms required a million physical qubits, and so there is this what I think of as kind of a redacted region in quantum algorithm space between 100 physical qubits and a million physical qubits, and what has happened over the past decade is that that redacted region has started to be unveiled a little bit, and through extraordinary work from theorists and from hands-on empirical work akin to what’s happened in the AI and machine learning over the past decade, just by developing, testing, benchmarking things, we’ve started to gain a deeper understanding of how to actually leverage and apply that quantum computing resource to practical problems and in some cases, proofs have followed that have said that’s probably not a durable approach to performance enhancements over classical. In other cases, proofs have started to guide the path towards where a durable approach might follow from that empirical work. So this blend of kind of theory at the high end and theory following and leading empirical work has really been what has unfolded in the last five or six years that I think is the kind of accelerant of the overall industry and that’s blended with progress in error correcting codes as well that have found ways to say partially error correct a system or error correct the most critical operations that are needed, and that’s led to a much stronger kind of theoretical basis for what can be done at the few hundred to a few thousand physical qubit level and a pretty rapid rate of progress in clarifying that and so now I’d say a decade ago it was maybe unlikely that we’d find real commercial applications at the 100 to 1,000 qubit level, now I think it’s not certain but much more likely today that we’re going to see that.
AZEEM AZHAR: I would love to talk about applications, but I want to just clarify one thing before we jump to that. So there is this concept that goes around, this idea of quantum supremacy or sometimes people call it quantum advantage. How should we be thinking about the relative technical milestones that start to become useful? I think back to being a kid with a computer in the 1980s and I was just really, really waiting for 8-bit graphics card that would show 256 colors on the screen and it was a very clear target that I could aspire to. Is there something similar that we can think about in quantum computing?
CHAD RIGETTI: Because of the different kind of hardware approaches, even within super-connecting qubits that we work on and then across the different modalities are relatively different today, the kind of 256-bit analog breaks down just a little bit but what we can talk about and focus on is this really exciting concept of quantum advantage. People have used quantum supremacy, quantum advantage. Quantum advantage, and people use these terms in different ways but the core concept around quantum advantage is take a practical problem that someone is solving with a computer today. Can you insert quantum computing into that overall computational workflow or pipeline, and demonstrate some sort of advantage. Some sort of benefit to the end customer. Be it lower cost of generating those solutions on an ongoing basis, faster time to solution, or maybe a higher degree of accuracy, that better, faster, cheaper kind of idea on a practical problem. That’s quantum advantage, and that is the milestone that I believe is really going to lead to a kind of inflection point in the rate of revenue growth within the quantum computing industry because enterprises, governments, organizations are going to have an adoption vector to come into quantum computing and start using it that is purely based on the performance of the technology and not on the kind of second order effects of this is an emerging technology that is going to be critical long-term for them. There’s a real compelling reason to use it now.
AZEEM AZHAR: You have this concept of narrow quantum advantage which comes before broad quantum advantage. So what are the kinds of things that we can do with narrow quantum advantage? Which industries would benefit and what are the problems that you can solve with it?
CHAD RIGETTI: In narrow quantum advantage, you’re demonstrating a benefit for a specific use case, and that might be pricing a certain flavor of derivative for a bank or for a hedge fund slightly more accurately or a little bit faster than they can do without quantum computing in the loop if you will. The very exciting application space within computational finance is going to be one of the first to see narrow quantum advantage. There’s a few reasons for that but one of them is just that there’s such a myriad set of application areas that are already compute-based and where the data exists in a directly computable form, and it’s relatively straightforward to identify … So it’s relatively straightforward in that domain to identify a subroutine within the problem where quantum can potentially provide an enhancement and iterating against that to demonstrate that for the customer, for the partner. So finance is going to be one that we think is going to see near term value.
AZEEM AZHAR: It’s quite interesting what you described there which is the quantum computer taking part of a process in this sort of overall map of [inaudible 00:14:03] which makes me think back in the old days to numerical coprocessors that you would have had in the late 80s, the way you would have sort of pushed off your number crunching to the side of the computer but all of the displays on the screen and all the human input would be done on the main processor. So in that sense, a quantum computer won’t be as generalized as these typical computers we have on our desktop because it will be used to pick off key difficult elements of a process.
CHAD RIGETTI: The trajectory that we believe is going to unfold here is really similar to what’s happened with graphical processing units or GPUs. GPUs originally obviously were graphical processing units and simply handled the graphics on your computer, your workstation. And today they’re obviously used broadly in all sorts of computing including machine learning algorithms and in large scale computation overall and quantum is really going to take a similar trajectory, where it starts out as kind of a coprocessor for handling specific subroutines within a broader computational workflow, and over time, as almost kind of trees in a forest grow and emerge, more and more of those subroutines will quantum accelerable or amenable to the quantum subroutine and therefore the applicability of that quantum coprocessor will grow and grow and grow to consume a larger fraction of the overall computing pipeline or workflow and of the overall computing market if you will.
AZEEM AZHAR: When we talk about the application of quantum computing to particular tasks and say we’ll use classical computing on other tasks, I thought it was related to the kind of issue of computability, that classical computers’ work would be kind of Turing complete so they could do anything and maybe quantum computers weren’t, so they couldn’t do anything. But it sounds from you that it’s more a question of sort of economic viability, that it might be just too expensive to do certain types of mundane tasks on today’s quantum computers but if they get much, much cheaper, we could move every workload onto them.
CHAD RIGETTI: It’s interesting. I do think that cost is going to be one of the kind of maybe underappreciated drivers of adoption of quantum computing. As we start to identify these quantum accelerable workloads in industry, in government, et cetera, the early quantum advantage may well be attained because it’s cheaper to run that subroutine on a quantum processor than on a 300 or 400 million dollar supercomputer operated by a national lab. Or maybe you’ll get more out of your hours of supercomputer time if you address one particular bottleneck with a QPU or quantum processor unit, and that economic incentive I think is going to be a really critical early driver in the narrow and then broad quant advantage space.
AZEEM AZHAR: I understand why quantum computing could show a kind of quick advantage in the financial services industry, they’ve got the data, they’re used to computing in complex areas and they can make a lot of money with really marginal improvements. But I’m quite curious about what the output looks like. Does it just come back to [inaudible 00:17:03] as a number? When you price a derivative using a quantum computer, what comes back?
CHAD RIGETTI: It’s got to look the same as it does for how you’re doing it today. This notion that we have taken at Rigetti of focusing on using quantum computers in tandem with existing classical computing and existing kind of classical algorithms to provide an augmentation also then allows you to integrate into the existing platforms and kind of workflows that those customers are using. Quantum computing sounds incredibly mysterious, but it’s classical traditional in, it’s a quantum processing pipeline, and then it’s classical data out, and that reality actually makes it a lot simpler to view how that gets integrated into workflows like asset pricing or portfolio optimization or derivatives. One example that’s really exciting, at Rigetti, the team recently worked on a problem in weather modeling using this kind of quantum-enhanced machine learning approach. So this is a really fascinating application area because the ability to model and predict the weather obviously has huge implications across all sorts of industries, public sector, private sector, financial trading, and so much of the quantum algorithms and application research that has gone on over the past five or six years has led to extraordinary progress, but it has focused on problems that are building up from the toy problem level to adding incremental complexity but is still relatively distant from large scale industrial applications and practical problem instances. So what the team at Rigetti did was kind of focus on a practical weather modeling problem called synthetic weather radar. We used an existing fully functional machine learning pipeline and added a quantum step in there to carry out a generative modeling process and when doing that were able to show that the overall results are competitive with classical, and in some cases led to better meteorological metrics from the quantum-enhanced pipeline. The reason this is exciting is a) it’s weather modeling, a large scale problem of broad application, and b) it really demonstrates the power of this notion of inserting quantum computing into an existing pipeline and augmenting the capabilities of that existing pipeline.
AZEEM AZHAR: In that example I think you ran that on a 32-qubit computer which is enormous compared to where we were a decade ago but it’s going to seem quite small to where you want to take this over the next decade I guess.
CHAD RIGETTI: That’s right. That’s kind of on a current generation system from Rigetti and those systems, as you add qubits and as you decrease the error rates, both of those lead to exponential improvements and the power of the quantum resource that that represents. And so going from 32 to say 50 or 80 qubits on the next generation systems or having [inaudible 00:20:00] rates of different operations across the processor just leads to additional augmentation and so the fact that the team is able to demonstrate competitive performance today with a 32-qubit processor really holds a lot of promise for where this is going to go over the next few years and the path towards quantum advantage.
AZEEM AZHAR: I’d love to learn more about that. I mean we’ve talked about the different types of quantum computing approaches, how you build your qubits and so on and you’ve given us some examples of some applications, at least one of which the weather is very dear to my heart as a Londoner. Just talk us through your approach and why it’s promising. There are some distinctive things that Rigetti does in the market in terms of how you build things and how you think of things from sort of top to tail. Why does that approach work?
CHAD RIGETTI: We are a full stack quantum computing company and what we mean by that is that we focus on building quantum computers and delivering the power of those machines over the cloud to end customers and also partnering with those major customers to help them apply quantum computing to their applications that they’re pursuing. By doing that, we both facilitate our customers to get more value from systems today and in the future but we also can drive all those insights back into our product development process and into our full stack of technology. We start at the chip level. We design and manufacture our own superconducting quantum chips within our semiconductor style fab here in the Bay Area, and then we integrate those into QPUs or quantum processing units and operate those over a quantum cloud services program for end customers. In these deep tech areas or kind of hard tech innovation companies, this ability to be full stack and to make engineering trade-offs across domains, so we can make trade-offs in our compiler technology to manage imperfections at the chip level or vice versa, and that ability to do these full stack trade-offs becomes such an accelerant when you’re successively solving problem after problem that stands between where you are today and the next level of kind of commercial traction and market opportunity. So that full stack approach is really powerful. It also then gives us an ability to learn the most from these customer engagements. So we can take lessons learned from working say with Astex Pharmaceutical Company, a company we’re partnered with on applying quantum to drug development. Or with Standard Charter in financial markets, and drive that into our chip design. We can drive it into our compiler design, we can drive it into the developer tools that are used to program the machines and that ultimately leads to an ability to kind of go faster, to have a greater degree of capital efficiency and ultimately the ability to kind of scale once you start to reach that commercial market and move on quickly.
AZEEM AZHAR: Yeah. It’s fascinating, the full stack approach because it also has the additional complexity that you have to be able to hire people who can do all of these different things under one roof and perhaps you don’t get access to the economies of scale for example in semiconductors of there being huge fab operators around the world who can build things for you. So there’s a trade-off, I’m quite curious actually about this idea that you manufacture your own chips all the way through to algorithms. So what does a senior team meeting look like on a Monday morning at Rigetti with all these different hats and heads?
CHAD RIGETTI: Well it leads to this fascinating problem/opportunity in building a company, and that is you have the requirement of having world-class expertise across different domains that aren’t usually under the same roof. That comes with communication challenges. People speak different technical language, recruiting from very different domains and building world class teams in multiple different technical areas for the company becomes a challenge. However, if you can get there and you can pull those teams together and establish a common language, a common unifying culture and vision and mission, then those teams, because of the efficiency with which they’re able to cut across those technical boundaries, because they do now have a shared language, can do things that are much harder to do when you’ve got to navigate not just technical boundaries but also organizational and maybe even kind of cultural boundaries in interfacing to say a large semiconductor manufacturer. And one example that I’ll highlight, it’s very kind of mechanical in nature, is the different clock cycles with which quantum computing innovation is happening today, and semiconductor innovation is happening today. So a lot of these global fabs are on every six month or twelve month tape-outs for new chip designs. We do new chip design starts once or twice a week, and we have been doing so for five years. So this ability to kind of rapidly iterate, drive lessons learned into the next generation chip design almost continuously, rather than doing so kind of on an annual basis just creates this profound accumulating advantage over time in our ability to solve those technical challenges which is the name of the game. This is a marathon, it’s not a sprint, and it’s about deep innovation and pioneering innovation, being able to sustain that over time. So those different clock cycles that you’d have to then kind of interface to each other externally becomes actually a strength of this deeply integrated approach.
AZEEM AZHAR: One could understand why that clock cycle differentiation might emerge because semiconductors are a more mature technology so we’ve been through the sort of hard work of the semiconductors which was done in the 50s and the 60s and then you have the sort of exponential takeoff of different approaches, we learn how to miniaturize and scale in particular ways and then the easy gains are all used up. Is where you are in quantum computing at that point where some of the really hard work was done in those years running up to and around 2010 and you’re finding yourself into that steeper part of the S curve where large gains are to be expected because you’ve figured out a lot of the gnarly work and if that’s the case how long do you expect that to run? I mean the whole speed through miniaturization on traditional processors, we call that More’s law, ran for 45 years or something. I mean are we two years into a 50-year vertical trajectory of improvements?
CHAD RIGETTI: At Rigetti, because of this kind of fab driven full-stack approach, what we’ve been able to do is put in place a chip-level architecture and design that truly can scale going forward over the next five, seven, ten years to carry us through that narrow quantum advantage, broad quantum advantage, and then early stages of large scale [inaudible 00:26:38] quantum computing, and that leverage is kind of pure semiconductor chip level innovation and chip level architecture. We use a multi-chip solution, so we’re able to build our quantum processors out of multiple different physical pieces of silicon.
AZEEM AZHAR: Can I call them Lego blocks?
CHAD RIGETTI: You can call them Lego blocks.
AZEEM AZHAR: They’re a bit like Lego blocks in your diagrams I think.
CHAD RIGETTI: They are and it’s a result of … The team at Rigetti worked on this for five or six years in putting the building blocks in place, and in the end, it’s simplexity. It becomes simple once you solve the problem on the other side but this ability to assemble larger and larger quantum processors through chip level technology allows you to assemble them from individual building blocks is extremely powerful and what we believe is going to happen, that’s going to carry us through a wave of scaling and innovation that gets us through the broad commercial adoption and then down the road, as there has been in the semi industry, if you look over 10, 15, 20 years, there’s always kind of additional river crossings that need to be made with the technology as you come up against new challenges and that’s surely going to happen in quantum computing. Our scalable chip solution we believe can carry us through broad commercialization the next five, ten years of this and getting towards early [inaudible 00:27:52] machines that will truly start to tap into that broad commercial market.
AZEEM AZHAR: Chad, I want to hold something up for you. This is a podcast, we’ll have to describe what you’re seeing. So here is a Sinclair ZX81. It’s got a Zilog Z80 processor in it from 1980. That processor had 8,000 transistors in it and I’m now holding up my state of the art iPhone which 40 years later has 15 billion transistors in its main processor. So that’s a two millionfold improvement in 40 years. I mean is that the kind of improvement we could see in qubit count for sake of argument over the next four decades?
CHAD RIGETTI: I absolutely believe that’s the kind of computing power enhancements that are going to be delivered and in fact as More’s Law has really kind of slowed or even ground to a halt over the past decade, quantum computing has emerged as potentially the successor to that global driver of economic development, and with quantum computing, it’s not just about miniaturization. It’s not just about packing more and more qubits onto the same size of silicon. But there’s actually multiple different ways of unlocking greater and greater commercial value from that compute resource as we’re talking about through better algorithms, better error correcting codes, lower error rates on your physical quantum processor. So ultimately it would be reasonable to predict that there may even be an acceleration in the rate of advance of computing power overall as quantum computers truly come online, not just that it kind of starts to carry the torch from the traditional classical More’s Law that has prevailed over the past 50 or 60 years.
AZEEM AZHAR: I love to hear that because that’s the thesis of my new book, The Exponential Age is a thesis of greater acceleration from technologies like this. These are the longterm ideas and we talked a little bit about your very proximate milestones but what are the medium-term, tangible goals? The point at which you might be able to say, “This is a big shift and here is why.”
CHAD RIGETTI: I think narrow quantum advantage is really the near-term goal in delivering practical benefit to end users on practical problems and that’s really where we are laser-focused as an organization, and I believe just kind of needs to be the broad focus of the quantum computing industry right now. Ensuing from that, broad quantum advantage is where you’re solving problems not with an incremental performance improvement and a factor of two or five or ten, but you start to solve problems that today are intractable, and at that point Azeem what you’re able to do is really start to push computing into domains where computing isn’t necessarily used as part of the workflow today. So maybe you’re disrupting a wet chemistry-based process for new materials discovery or for new drug development. Or perhaps a portfolio manager can actually leverage market data to inform decisions, decisions in managing their portfolio. Whereas today the degree of approximations that need to be made to solve those problems of classical computing makes the classical solution much lower value or not really applied in practice. So getting to that broad point advantage and broad point advantage almost certainly is going to require a meaningful degree of error correction and potentially even [inaudible 00:31:07] tolerance within quantum computing systems, and as we kind of get to that milestone, I think that’s where you’ll really start to see a second inflection point in the rate of adoption because as you start to solve problems that are intractable today, the ease with which those applications can be transferred from one domain to another starts to increase. It won’t require such a deep effort to get to quantum advantage in first place because you’re delivering such a substantial improvement.
AZEEM AZHAR: Right. So what would you say is the most out there example of an application of quantum computing and what it could do that we could wrap our heads around today?
CHAD RIGETTI: Well first of all I think quantum computing is going to have a really profound impact on areas where there’s already a statistical model of computing being used and that is in financial derivatives. It’s in machine learning and artificial intelligence as well, but aside from that, one of the most exciting and kind of tantalizing I think is this … What quantum computing is going to unlock in terms of the ability to truly simulate nature at the lowest levels, and you can pick practical examples or you can pick just really exciting examples from astrophysics and things like that. But Azeem, one that is dear to my heart is this notion of … There’s questions about the intersection of gravity and quantum mechanics, and how those two theories will ultimately be reconciled, and it turns out that the extreme conditions needed to really probe the intersection of these two, they only exist out there in the universe, they don’t exist on Earth today. And so this is an arena where we’re actually never going to be able to do the experiment. It’s not about getting to higher energy scales in a particle accelerator, things like that. It’s not about chemistry. We’re not going to be able to build a black hole in the lab and study what happens at the event horizon. And so quantum computing actually promises these insights that we’re not going to be able to gain in any other way, and from a societal impact and from a scientific perspective, that’s extremely exciting. From an industrial perspective, the ability to do what you’d think of as simulation driven design of new molecules or new materials is tantalizing, what that’s going to unlock for humanity over the next few decades. If you think about the combinatorics of the periodic table, building up molecules of hundreds or thousands of atoms and all that is possible, life on Earth samples a very microscopic proportion of what molecules and materials are possible to be constructed, and exploring that simply cannot be done through synthesis and wet chemistry. But the ability to explore that through computation, through a blend of theoretical tools, through machine learning, and quantum computing to provide the exact properties of those hypothetical molecules or materials that one could construct I think is going to lead ultimately to this kind of quantum age of identifying those new materials and molecules and going about building them and synthesizing them and using them to make life on Earth better, and that really I think is one of the profound longterm impacts that quantum computing is going to have.
AZEEM AZHAR: It’s incredibly powerful, and if we bring those ideas back to Earth, that also brings with it competition and it brings with it competition not just within companies but within nations. To what extent do you think there is an important race emerging in the field of quantum computing perhaps between countries across the globe?
CHAD RIGETTI: There is absolutely an important race unfolding or a race being set today. There’s multiple reasons why that’s happening, and why it’s rational for nations to compete along with this quantum technologies as a key component of that competition. First is the applications. Obviously there are applications that quantum computers will unlock that relate to national security, intelligence and defense. But maybe even more fundamental than that, computing technology has always been a fundamental driver of economic development. More’s Law has been one of the drivers of the global economy over the past five or six decades, and what feels like an inevitable march of evermore compute power, whether it’s delivered through better hardware, better algorithms, better accessibility through the cloud or the internet just fundamentally creates value and prosperity around the world. Nations are competing not just for the impact of quantum computing in terms of what it will compute, they’re competing to be the home of that kind of quantum powered economy going forward. That is a very rational competition to be taking place and it’s leading to a set of very healthy things around the world. Governments are making substantial investments and catalyzing the quantum economy and bringing the development of the technology on shore, and also ensuring that the laws and regulations are best suited to kind of support the development of healthy and flourishing enterprises and kind of having the global leaders based in the U.S. or in the U.K. Ultimately those things are just so important globally that it does lead to this geopolitical competition. Because of the nature of the application space, the market is likely going to end up being slightly geopolitically fragmented. There’s going to be tight alliances between the U.S., the U.K., and close allies and potentially on other sides of that some of the global adversaries like China and Russia and other places.
AZEEM AZHAR: So what would lagging in quantum computing look like? What systems are compromised? What would it mean to come second or third in this particular journey?
CHAD RIGETTI: So as a nation, I think it is much less important to worry about being first than to worry about getting there at all, and as this quantum economy really begins to take roots and those roots are already firmly planted in some places, they are firmly planted here in the United States in Silicon Valley and on the East Coast, they’re starting to grow and truly develop in the United Kingdom and in Australia, and they’re very strong in continental Europe, and also in China. What is most critical is that countries and governments get on a path to ensuring that they are participating in the benefits to their people and to their society of quantum technologies and of the quantum economy at both the development of the technology, the jobs that are going to be created, the social benefits that are going to be created through the technology and the application of it into their existing industries.
AZEEM AZHAR: I mean there is this idea that whoever perhaps gets to a certain milestone first is able to unlock the keys to the modern economy in the sense of quantum algorithms being very good at factoring large numbers, right? And unlocking the cryptography that protects credit card transactions and our stock portfolios and so on. As an observer of technology, it feels to me that that looks a bit like the Y2K problem, right? It’s easy to describe, it’s very kind of clinical, and it’s going to happen and we got to do something about it, but of course Y2K was not like that at all, it didn’t play out in that way. This idea of a Q Day where the ATM machines don’t work anymore because it’s all been compromised by quantum cryptography, is that really a risk that could play out in that way?
CHAD RIGETTI: I think one of the most fortuitous things that has happened for the emerging quantum computing industry is that these applications to code breaking, the RSA-based applications, actually require really what you could think of as the largest imaginable quantum computer, and there is an entire generation of commercial industrial applications that is likely to be accessible long before systems are able to solve that particular problem. If it had been the other way around, if Shore’s algorithm was the first application that quantum computers are going to get to, the way in which the technology is being developed would be vastly different than it is today and it would be much more closed and much more secretive I would imagine. But as we start to get there, there’s already adjustments that are being made in the kind of global communication technology stack if you will. Ultimately, you can fight fire with fire and use quantum mechanics to protect communication against quantum computing attacks. That is a technology that will likely kind of come online at around the same time that quantum computers are getting there, and then in the interim, there’s post quantum cryptography, which are just additional kind of classical cryptography approaches that should be able to protect things for a decade or two while true broad scale kind of quantum cryptography is brought online. So I don’t know if there’s going to be really this kind of profound moment when quantum computers all of a sudden allow us to spy on all historic communication of which we would have a record. That’s a real scenario. In the end, that is such a catastrophic outcome for many nations and many entities that I believe there’s kind of a form of collective intelligence taking hold now.
AZEEM AZHAR: Right. Better now.
CHAD RIGETTI: There’s a collective set of actions that are being taken that make the impact of that slightly less, that mitigate it through many different ways and approaches, and then also, what we are seeing obviously is a much more precise view of the timeline by which that’s going to happen and as we’re doing that, these kind of post-quantum cryptography approaches are starting to get some traction. So overall it’s certainly going to lead to applications in that domain. The question is who’s going to benefit from that, how profound the advantage will be and what ever advantages will be conferred in the meantime.
AZEEM AZHAR: It’s a profoundly dual use technology as all computing is as you described early, and I know that within Rigetti and other quantum computing companies, there are advisors who have experience in intelligence and security and armed forces so on for good reason. In some of the geopolitical competition of course, there’s a much deeper alignment between the research private enterprise and the state direction. That would be the case in China. For quantum entrepreneurs in Europe and in the U.S., what should they be looking for from their governments or nations to provide a sort of supportive and catalytic environment in these developments?
CHAD RIGETTI: I think first we have to recognize that quantum computing is a very rapidly emerging technology. The promise for the future in this technology hinges on our ability to continue to do deep innovation, research and development, and that requires a degree of investment in fundamental research and development at the university level to generate the workforce and train the workforce over time that’s going to continue to drive that innovation and so there is a really big people element of unfolding the next 10 years of the quantum economy and it’s a really exciting opportunity to have at the end of a physics PhD or an engineering PhD to go to work in a high impact industry like this. So that’s one condition that I think countries really need to focus on is investing in that quantum workforce and quantum capabilities through the university system. IP protections are going to be incredibly important and I think we want to be careful not to kind of try to take a closing things down and closing things off approach to early in the development of the technology. I think it’s really important that nations figure out who their partners are that they’re really going to lock arms with in the development of the quantum computing technology and establish these deep programs to catalyze the development of quantum technologies in the private sector, the supply chain, the workforce, and by doing so, really allowing the public sector and government to have their finger on the pulse of where the technology actually is. What it can do, what it cannot do, what the challenges are going forward. So really that kind of … That partnership approach, that’s what we have taken at Rigetti, it is what has served us very well and is what we believe is really in the best interest of the nations that are really developing these quantum economies and seeking to benefit from it.
AZEEM AZHAR: Is there enough of a flow of talent coming through and is it getting more accessible as you abstract away more of the sort of … The gnarlier parts of [inaudible 00:43:09]?
CHAD RIGETTI: There’s incredible people working in quantum computing today. More would be better for sure. At a high level, one of the fascinating things about quantum computing is I don’t know that there’s ever been an industry that has been founded by 100 or 200 physics PhDs collectively, and what that really means about the future, I don’t know. But it’s an interesting observation. Quantum computing, to truly be commercially successful, needs to tap into well beyond physics talent though, and that is fundamental and we need to continue to make investments there as nations and as companies, ensuring that there is more and more training programs, more and more education, and what I have seen I think is very, very promising, both in terms of the rate of adoption and awareness of quantum computing. Whether it is through opensource software programs or education programs or expanding quantum engineering programs at major universities, all those things have been really, really promising, but I think the critical ingredient ultimately is some day in the future, when Rigetti is a 10,000 or 40,000 person organization, we’re not going to be 40,000 physicists. It’s going to be predominantly engineers and other functional domains and what I think is important is continuing to build the economy around quantum and the companies themselves around this blend of the physics talent, the quantum computing talent, and then pulling in expertise from nearby industries, whether it’s the semiconductor industry or the AI research community or computational chemistry or quantitative finance for example. Pulling those people into more of the quantum ecosystem and making it easier and easier for them to kind of getting up to the cutting edge and that comes with better abstractions, greater reliability of the technology that makes it easier to kind of gets hand on immediately, and all that I think is on a really good trajectory. We need to continue to work on it for sure.
AZEEM AZHAR: We’re at this amazing moment, I think back to 1981 when I got my first computer and there was one low power computer in the house and it was remarkable and today 40 years on we’ve probably got dozens in the house, many of which just sit in desk drawers, unused. You’re on a journey that sort of some of us observed in a different industry decades earlier, you work on the edge of science and engineering and commercialization at a level that’s very hard for us to comprehend. What surprised you most about that journey so far and what do you think might surprise you in the years to come?
CHAD RIGETTI: That’s a really good question. What has surprised me I think has been just the rate at which the overall scope of effort in making quantum computing happen has evolved and broadened so profoundly over the past eight years to the point where there’s now hundreds of startups around the world that are working on different aspects of quantum computing technology and capability. Just the incredible overall momentum in quantum computing has been … I think has picked up factor than we might have anticipated. And then there’s been the emergence of a set of players, kind of the top end, that have promise to become some of the big dominant kind of global companies in the space longterm, and that whole journey has happened over basically half a decade. It’s really evolved incredibly fast and I think that’s the thing that has really surprised me. Rigetti is going to be going public later this year or early next year. There’s been other companies come into the public markets. I think the rate at which the broader kind of market and audience comes to understand quantum computing. Maybe we’ll have a positive surprise there, and maybe the kind of market adoption and interest in quantum computing, undergrads deciding that they’re going to do quantum computing for a career, maybe all that will happen faster than we expect and that would just be wonderful.
AZEEM AZHAR: The maybes are part of life working in the quantum space. Chad Rigetti, thank you so much for taking your time today.
CHAD RIGETTI: Thank you so much for having me.
AZEEM AZHAR: I hope you enjoyed my conversation with Chad Rigetti. I really loved talking to him and digging through some of the wider implications of what the next few years might look like. Earlier in 2021, I spoke with Jeremy O’Brien, the CEO of PsiQuantum, a Palo Alto-based company building quantum computers using photons. Now Chad and I didn’t get into that approach, helpfully called photonics much in this podcast, but Jeremy and I covered it in depth. The conversation was fascinating, and if you’re interested in quantum computing, I urge you to listen to it. To become a premium subscriber of my weekly newsletter, go to exponentialview.co/listener, where you’ll get a 20% discount. To stay in touch, you can follow me on Twitter. I am @azeem, A-Z-E-E-M. This podcast was produced by Mischa Frankl-Duval, Fred Casella and Marija Gavrilov. Bojan Sabioncello is our sound editor.