Podcast with John Martinis, 2025 Nobel Prize Winner and CTO of Qolab

Overview In this podcast, 2025 Nobel Prize winner John Martinis was recently interviewed at the 2026 APS March Meeting by GQI’s own George Schwartz and Steve Lee from the PR firm San Francisco Agency. John discusses his foundational work in superconducting qubits and his current role as CTO of the startup Qolab. He explains Qolab’s vision to reach a million physical qubits by utilizing wafer-scale semiconductor manufacturing and “deposition and etch” techniques rather than traditional liftoff processes. Martinis highlights the importance of systems engineering and improved wiring solutions, such as moving from coaxial cables to integrated flex circuits, to make quantum scaling more economical. Addressing future security, he suggests the world should prepare for cryptographically relevant quantum computers within five to ten years by adopting quantum-resistant protocols. Finally, he reflects on his transition from academia to business, noting that his career has been defined by a unique ability to “pick apart arguments” and a dedication to clear scientific communication.

Transcript George Schwartz: I’m George Schwartz, I’m with the Quantum Computing Report by GQI. And at GQI what we do is we try to serve as a third-party independent understanding of the entire quantum ecosystem specifically in computing, sensing, and networking. And just kind of getting a sense of all these roadmaps that are being proposed nowadays and understanding where may be particular challenges that you’ve know very well into reaching those scaling and reaching utility-scale quantum computing or cryptographically relevant quantum computing. Steve Lee: So, John, why don’t you tell us what you’re famous for here, we all know that you’ve just recently have won the Nobel Prize for physics this year in 2025. And but maybe you could tell us a little bit more about your history and the basics of how that came about. John Martinis: Well, the Nobel Prize was for essentially my thesis experiment in around 1985 we were doing that. And at the time we were making the first superconducting qubits, although it’s prehistoric qubits, that word really hadn’t been coined at that point. And at the time we were just trying to show that electrical circuits could obey quantum mechanics. And there was a kind of a philosophical reason at the time where no one had really demonstrated that quantum mechanics could happen in electrical circuits, everyone thought this is a physics of the small, how electrons, photons, atoms work, things like that for small, it’s a physics of the small. And then you know we had a chip that was about a centimeter across and then with large wires you could see with your eye. And then that system could obey quantum mechanics of course if you get it into the right regime. So, I was very fortunate to work with John Clark in his lab with Michel Devoret to do that experiment. And then of course that experiment was pretty noteworthy at the time, people paid attention to it, but I would say it’s only what it led to with a whole series of experiments on quantum devices and then over time qubits and the like, now it’s of course a huge industry out there. So, it was kind of one of the foundational experiments which was why I think it’s as important. And I would say at the time people had done some prior experiments but things were murky or whatever, by doing the experiment really well we were able to lay bare what the foundational elements were, which is the fact that it was a microwave device and you had to kind of mix microwave engineering from quantum mechanics and kind of understanding in that way, think about filtering, thinking about the microwave circuits. And because of that as people did more experiments, I think it became easier to kind of put the whole field in proper context as to how to build these things. But of course, many, many people contributed over the years and to make it with the field what it is today. Steve: Well and then after that you were at Google? John: Yeah, so I was at NIST for a while in Boulder, Colorado. Then went to UCSB where we were developing the qubits. And you know working on some of the early algorithms to show that you could do a very non-trivial algorithm even with a few qubits and it worked right. And you know we eventually moved the group to Google. Um, basically I saw the fact that it would be very hard to make a quantum computer in an academic group. There’s a variety of reasons for that. But it may be no surprise and then Google had the resources and could hire people for many years. And we ended up focusing on the quantum supremacy experiment that was 2019. And you know that that was a big thing for the group. And I would say in 2019 after that that’s when people started taking selfies with me in the hall. That that’s kind of leave you know gets to a certain level it’s totally off scale now. But yeah, that was that was kind of nice that people recognized that. Steve: That is really great. And then now you’re with Qolab. So, can you tell us a little bit about Qolab? John: Yeah, so I decided to leave Google in 2020 at the beginning of COVID. And I wanted to do things a little bit differently; Google was changing their management and it’s fine. And then we decided to you know focus on building qubits in a different way. So, what’s really nice when you come here you can see the technology people are building to scale up the qubits. To a thousand qubits, maybe ten thousand or so, and it’s done in a certain way, it’s pretty interesting. We have a vision of how we’re going to scale it up to let’s say a million qubits and do that in a more economical way than what people are doing right now. Because right now it requires a lot of complex components and 3D kind of integration and all these things, it’s very nice, very beautiful. We think we can do it using semiconductor silicon wafer manufacturing, wafer-scale manufacturing and also package it in a more conventional way that can be done more automated. So, in the end I think it’s going to be cheaper qubit systems we’re planning on. And also make the qubits better and more reliable because we’re using this sophisticated semiconductor processing. George: So, on that, that leaves a great segue on Qolab. With some of these players like IBM and also Google and QR for example we see these roadmaps of potential numbers of logical qubits up to thousands potentially of logical qubits. What is your vision for Qolab over the next couple of years of how kind of milestones that you’re looking at? John: So, our vision is I think in the end similar to theirs to have you know million physical qubits, these kinds of logical qubits. You know we all kind of honor on similar roadmaps. Of course, the roadmaps are what people are projecting to do. And then I think there’s still a lot of technical problems that you have to solve to get there. I think our approach is to think well what are we going to need from at five to ten years when you’re at let’s say thousands of qubits, how are you going to get to a million, how are you going to do that in a manufacturable way, how are you going to keep the cost down. Because right now I was just up at a booth right now and they were talking tens of millions of dollars for you know thousands of qubits. And if you go to a million that’s going to be tens of billions of dollars. Now sure people will bring the price down. But we really want to figure out how to make that more compact, how to put that together in a better way. So, we’re looking at a longer-term technology. And you know we’re a small startup, you know we can’t compete right now with what everyone else is doing, but I think we found something that was kind of unique and different that we can do you know to to get what we need in the next few years. Steve: And for clarity, are you talking about getting that all on one chip or are you talking about doing that across multiple? John: So, we’re thinking about a chip something like this for the qubits, about 20,000 qubits, which is just one chip. You know and we have to get the yield up to do that. And then we connect it to a wiring wafer, again 20,000 qubits, and then do that just you know in one module. And then from that you can place them together as modules to get to you know millions of qubits. Steve: And and also for clarification just on the wider level before George gets more technical on me here. Are we talking a specific modality that you are betting on? John: Yeah, it’s superconducting qubits and we’re thinking about the transmon qubits with adjustable couplers, adjustable qubit frequency. This is what I did at UCSB and Google working on that. However, the wiring system and the qubit fabrication capabilities we think could apply to any kind of qubit system that you wanted. Maybe even for spin qubits or other kind of qubits for the wiring solution. So, you know we’re trying to think a little bit more open-ended about what we do. We have to develop that technology in order to get there. George: So, you’re saying you’re like thinking about 20,000 qubits on say a handheld kind of wafer-ish something of that size. So, do you envision the current size of your superconducting transmons to be small enough to hit that 20,000 metric or do you think you’re going to have to probably get a little bit smaller? John: Well, this is based on a millimeter spacing of both the qubits and the couplers. And we have a prototype design where that could fit quite easily. And you also have to have other components on it and the like. So, you know when we a prototype design that seems that you know it could be done. Of course, you know you have to build it and see what all the issues are before you would know the exact number. But that’s the kind of kind of numbers we’re doing. And you know for example with the wiring wafer you have to make sure you can bring out all the wires from that. So, you know we put all that together and it looks pretty reasonable. But you know the real part is getting the qubit fabrication to be improved, figuring out how to do all the wiring wafer, all the subsidiary things you have to do. And that’s kind of what we’re working on now in our Series A for the qubits and Series B is bringing up all the control wiring. And just show that you can do it in a different way. George: Sure, absolutely. On talking about you know the all these wirings and things of that nature that are going to be put onto this wafer ideally potentially cleaning up space, one aspect we think about is the coherence times of your superconducting transmons that T1 time. And if you can keep it at the same size I guess you can be able to maintain your metrics but you might want to be able to push that up a little bit. And you mentioned in your talk that you know there’s maybe some ideas of moving to say what Princeton did with tantalum or alternatively a different way of thinking of how the electric fields integrate with the dielectrics. What is what are your approach on Qolab there? John: Okay so now we’re talking about the quantum chip. And we definitely have some different plans for that. I’m going to say it’s different than what everyone else is doing. Because it’s different it’s riskier. But we feel that we really need to go in this direction to build million-qubit devices. So, I’ll be you know let me get some numbers right now. Right now, people make their junctions and make these things with a process called liftoff where you make a photoresist stencil, evaporate the metal, and then lift off the stencil in acetone to release the metal properly. No one in the semiconductor industry would think about building a big chip with liftoff. When I mentioned this in in Asia from someone who used to work at TSMC he started laughing out loud. I mean it’s just you know you just don’t do things that way. Now in the end that might be the right way to do it, people might think of more clever ways. But you know our particular thesis is we should try to get rid of liftoff because it’s not a you know truly scalable technology. That at least that’s our hypothesis. That’s what we’re doing different. And you know there is some thought to this because we know that when people are building 100 or 200 qubit chips now, they have a problem yielding all those qubits. Now it doesn’t have to be perfect, but you know it has to be pretty good. And if you have a problem yielding 100, 200 whatever, at least that’s what I you know kind of know a little bit, maybe it’s better I don’t know. But if you have problems with that, how are you going to do a million, how are you going to make it reliable. And you know really at a 100 or a thousand you should be looking at basically 100% yield. And that’s what the you know that’s what the semiconductor industry can do. So, we just have a totally different way to build it which we think is deposition and etch. Think it’s going to be more reliable. We’re developing that, we’re getting a lot of good success on it. You know we’re going to be testing some new chips next week that we hope you know looks like we fixed a bunch of problems and it’ll be better. But you know it’s experimental physics; you have to just see and probably have to improve things some. But you know I it’s not like we have to make you know the best T1 devices in the world. In fact, you know the best T1 devices and the T1 devices in a complex thing are kind of different beasts. And you know if we can get something like 50 or 100 microseconds on our you know complex device for a T1, we think that’s quite significant if you can you know make it reliable and whatever. And obviously over time we want to improve upon that. But there’s something that’s a little bit more subtle about this than just T1. And that is these things called two-level states. And what it means is that at certain frequencies that are defined by the random structure of the of the oxides in it, where it resonates with these microscopic states and then your T1 gets really bad. And you get decoherence. And you know what we did when we saw that at UCSB and Google is we just operated away from that. And then you try to avoid them when you build the whole thing, it’s this complex calibration routine to do that. Well, you know my particular thought on that is this is fine even 50, 100, 200 qubits maybe a thousand, but for a million qubits that’s just you know it’s too hard of a calibration problem. So, we want to get rid of that, we want to make it better. Again, it’s another thing that I think we just need to do better with. And you know if you look at the theory of where these TLSs are coming from, we think that in our geometry of the way we make it that since all the electric fields coupling to those states are through the air and not the silicon, then it’s going to be a smaller magnitude. So, they’ll still be there but they’ll maybe be ten times smaller. In which case it’ll be so much easier to operate that you know I think it’ll make the world a difference when people want to scale up. So again, there are these subtle things you have to really engineer around and make the materials better. Steve: Well and you were saying earlier in your previous talk that that maybe there was some role for AI or something like that in the error correction or in being able to figure some of that out. Is that what you’re hoping for with that? John: So that’s a different subject. So, what happens is when you do let’s say something like the surface code and you run this and you get from the output of your quantum computer while you’re running it some data from it. And that data if you decode it properly tells you where the errors were and then you can correct for it. And the nice thing about the surface code there’s something called minimum weight matching where you can find the most probable error in it and you can build the model. And there’s an algorithm for doing this that people have worked on and that’s known. The problem is that’s kind of a CPU model that you know is provably good but you know you have to you know have to figure out how to get that to work on millions of qubits. And of course you have to distribute that computing among all these CPUs. So you know I think there’s an issue of whether you can get that to work. And then what people now are doing is thinking about doing that decoding with AI. Basically, training you know training the data by let’s say putting in a model where you put specific errors in and then you fit to a large language model kind of thing so that it will decode that proper thing. So, it’s just training. George: Now you mentioned the surface code and one thing that we see with as you start getting to very large depths that the resources of the surface code to reach high-quality logical qubits can be very taxing. Do does Qolab thinking about maybe increasing nearest neighbor connectivity beyond say four nearest neighbors? John: So yeah let me talk a little bit about that. What I like so the problem is it’s a systems engineering problem where you have to properly optimize twenty things at once. And when people talk about these different codes, they’re optimizing the physical to logical qubit ratio. Which is great. And then you want to invent new codes and that’s really great. There’s no problem with that. The problem is people write papers, you know this is very promising, but have they done the twenty end system engineering studies to know if you can do it. So, you know one of the popular codes that everyone’s talking about my understanding is they don’t have a definitive decoding mechanism for that. Now maybe some magic AI will get you there. And that would be great. But again, it’s magic. If you don’t actually have an algorithm, that’s okay, and I know in years past people have proposed codes where it’s known that the decoding is you know NP complete or something, that it’s impossible to do efficiently. So again, you know this is great people are doing that but you have to look at the full systems engineering. And of course, those codes require long distance communication which might be hard and you have to figure out that. And you know the more connections you have, the worse fidelity your qubits have. So, it’s a very complicated systems engineering issue. And then of course you have to be able to decode fast. And you’ll make it more complicated than that. And for example, the LDPC codes, the little bit I know about it is you’re encoding lots of qubits in you know one patch. Okay now that might be great for memory but when you start doing logical operations, well how do you kind of disentangle one qubit to the other. In the surface code you tend to have one qubit per patch. And thus, when you do logical operations, you’re kind of moving these patches around and it’s much more clear how the logical interactions work. Again, I’m not an expert on this but these are all the things that need to be worked out. And unfortunately, when you know you hear the press releases you know it’s the best thing. Yeah, they talk about the best thing. And if you know anything about systems engineering it’s the worst thing that limits you. So, you know what’s the worst thing and how do you worry about that. But again, it’s very good people are working on it, you just have to be very careful about understanding if the full end to end will work. George: And on that like long range entanglement operations and you mentioned say you had potentially Qolab’s thinking 20,000 qubits in each module that could be connected. What kind of long-range entanglement operations are you guys thinking about right now? John: So, what happens is that would encode about 20 logical qubits. And the nice thing about let’s say the surface code is to do logical operations, they can move arbitrary length in your architecture in kind of one move. You can move one thing or you can move many things. And there’s a read-out resource trade-off that you can work with. So, the local connection is only needed for the error correction but once you get to logical it doesn’t and then it’s kind of simpler. So that’s what I would say the surface code is it’s hard to get the logical qubits but moving them around is fine. The other thing for example the surface code is your tolerant to about 1% or so, maybe a few percent dropouts. Because you’re never going to make a a perfect system. Whereas you know what I’ve been told for example that’s where you have four connections, if you do the heavy hex which IBM does, that’s on average two and a half connections and you can do error correction there. But there if you have a dropout, it’s kind of catastrophic. But because the surface code is multiply connected you can kind of work around it. Again, I’m not gonna say I understand all that completely but it just points out there are systems engineering trade-offs that you have to understand to be able to, you know, understand exactly what you can build. Steve: You know for some reason I think that maybe you understand you what you don’t understand is more than most people understand. John: So, I have I have this unique ability to pick apart arguments and figure out what’s wrong. Which is not typically what people do. And a lot of people find that very annoying and it’s hurt me in my career. But you know… Steve: Clearly it has not hurt you in your career. John: Well, you know I would say that helps you with the Nobel Prize. But in terms of getting along in a big corporation where you know you’re supposed to be all positive and everything’s okay and we’re on our roadmap, you know then it gets a little bit edgier. So, you know I’m learning how to do that better. But I I’ve always tried to be very open about what I think are the critical issues. And but you know I also understand that whatever I’m worried about, someone may have thought about it and have a solution for it. And that would be great. And if I can encourage people to do that and find those solutions that’s really fantastic. Steve: Well and I also think challenging people with the things that you come up with is also very positive. It’s a very positive outcome. John: Well, I think it is. But I’m just warning you, warning the audience that you know although this is very powerful it can get you into trouble. Trouble isn’t always bad. Yeah, but that you know that’s okay. George: I wanted to go back to the some of those challenges that you pointed out. In your talk you showed this notable chandelier from the Google supremacy 2019. And you mentioned this wiring and you very astutely mentioned that we should really be moving towards what the microprocessor started looking like and so what is Qolab thinking about how to remove all these SMAs, coaxials, into thinking about some kind of vertical connections and how to get all that control? John: Well yeah, so the obvious thing to do is to go from coax to flex. And I actually started working on that at Google in 2017-2018 realizing we eventually had to get there. And you know one of the problems at Google is that there was not a universal acknowledgement that you had to do that. Which was very frustrating for me. Because for me it was absolutely clear. And what I like at this APS meeting is you look at a lot of vendors here and you know they have quite sophisticated solutions, you know kind of things I was thinking about but really cleverly done that you know I’m really really happy. But you know I think those solutions will get you to thousands of qubits, but I’m not sure whether that’ll get you to millions of qubits. So because of that it could you know it may get you there and cost too much money, okay. But anyway, I started thinking about that after I left Google, how to improve upon the flex and we came up with this series of really nice inventions to get around this and to you know build it more as an integrated circuit as much as possible. Which is you know the next thing you want to do after flex. And what we decided is you know to be unique in our company we’re just gonna start working on it now. Cause it’s gonna take you know five years or so before you see the limitations of what people want to do and then we’ll be ready to to do that. Now of course that’s a conjecture and you know we’ll see if that’s true. But you know as I’ve been thinking about this, I think that’s a really good alternative to what people are doing now. George: And is one aspect you guys may be thinking about some form of multiplexing of both control and readout? John: Yes, of course we’re doing that too. And so, there’s a variety of ways to do that. There’s a variety of ways to do the control circuitry. We’re learning that people are developing some new interesting control circuits here that we have to learn more about that might make our system even simpler, which is what I would like. So but yeah, the basic idea is if you look in the GPU industry and obviously there’s tons of money being put in that, they’re going to wafer-scale processes with lots of different chips and chiplets on it and the like. And you know we want to take advantage of that in the next generation. And we’re working with the big companies that know how to do that. And it’s going to have to be made differently than the way they’re doing it, but we hope that the tools and processes that they do will enable us to get that done. Steve: Yeah, it is kind of amazing that you know when I saw your presentation the original one the first one and saw the cryocooler that you guys made you know with its you know just bunch of like copper pipes and whatever strung together with solder. And where things are now that you can sort of buy that stuff off the shelf in a way and stays. And I think that that’s every generation… John: Yeah, and well you know and, in my thesis, I had to make the copper-nickel coax to drive the qubit because you know you couldn’t really get that. So, you know another thing we had to make completely ourselves. Steve: Sure, I mean John Clark was like showing how he made the slug after tea one day someone was complaining and somebody said oh why don’t you just put a big drop of solder in the middle of the thing and oh that might work. You know. John: Yeah. and I think it’s interesting with John because you can see the state of the technology in the early sixties. And then you could see what we were doing in the middle eighties which is you know you can see the microwave engineering there. And then you know you look at what happened with the quantum supremacy experiment and now things are even getting more sophisticated. And but what I see is what we want to do in ten years’ time when we use this wafer scale integration is a whole different scale of this which will be I think very, very interesting. George: With that kind of it’s a nice segue into when we think about what Qolab’s kind of business model you’re thinking about. For example, when we saw the evolution of the transistor there are companies developing end to end processors, in this case we’re thinking maybe potentially full stack quantum computer. But then there’s maybe some people that are just focusing on certain levels of the stack which you guys have fantastic answers to of how you’re thinking about the qubits, the flex connections, but what is Qolab’s approach there of selling this potentially this whole machine? John: So, the business approach is very interesting. You know when the dawn of the computer industry, let’s just take IBM, they were what’s called vertically integrated where they did everything. And built everything and they were very successful at that and you know synonymous with computers although there were other competitors. But then let’s say in the 1970s and certainly into the 1980s, it went horizontal where people were focusing on the certain components. The obvious one is Intel with the CPU but there were people with the hard disk drives and just other components. And that way everything could be manufactured more inexpensively, people can focus and really accelerated the technology. So, what we do see now is there’s companies like you know Google who I worked with and IBM and Rigetti and other you know atom systems and the like that are very much vertically oriented. They’re trying to do everything. Okay and you can see they’re buying things from vendors and the like, but you know pretty much the key technology. And that’s what we saw when we were working at Google is they were trying to figure out how to build everything. And we think that’s the dinosaur model. And you know you look at what happened in the semiconductor industry and if you look at what’s going on today, you see there’s a lot of component manufacturers that are really doing useful things. So, the Qolab model is to collaborate, it’s quantum collaboration, and most importantly we want to collaborate with the semiconductor industry who has all this expertise to build this. And not to fabricate, not to raise money to build a clean room, you know use existing clean rooms, use existing processes and you know build up the technology that way. Steve: I mean finding the best of breed, that gives you the opportunity to you know when people are focusing on stuff like you’re focusing on things you very often get the best of everything or you have the liberty to choose the best of everything. John: And I’m going to say for example when we look at work with applied materials, they really understand how to build the tools, how to optimize the process, you know we have to explain to them what we want to do with the quantum, but they can kind of match that with their tools and we can try things. So, I think that’s a a you know I think it’s a really good model and but it does mean that let’s say you build a quantum computer and the like you have to share you know what you can sell with the quantum computer. And you know our hope is you know it’s so transformative that you know maybe I can just be a billionaire instead of a multi-billionaire. I don’t care! That would that would be… I don’t care you know on that end. I just want to get it to work. And actually, but actually the most fun part is that when you talk with these experts they can explain to you things that you normally don’t know about or how their machines work that’s really, really fascinating. And I enjoy that a lot. Steve: You know it’s really not to not to go a little bit away from the from the technology but it is very interesting to meet somebody who used to be a scientist or you still are a scientist but now has to do business. How has that transition been for you? John: Well, what happens is when you’re young you want to do basic science and the like and as you get older your interests and maybe your creativity changes some, it just happens. But what I like is I get to learn about all these new technologies. And I’ve always enjoyed engineering, I now get to learn how professional fabrication is done, I get to learn about the business, how to run a business, I study a little bit about persuasion skills and I get to learn… I learn how to get podcasts just by doing it a lot. And you know it is a constant learning experience. It’s just I get to learn more and more things. I actually find you know really as interesting as physics. Steve: Have they put you in the accounting room yet? Have you had to deal with balance sheets and all those kinds of things right yet? John: Well, that’s the CEO’s job. That’s why I’m CTO. I didn’t want to be CEO because you know you’re basically you’re raising money, okay. And that’s great and you know I help out, we talk about it. But I’m truly trying to keep my focus on the technology and you know explain for you know explain what Qolab is in terms of its technology. Steve: Right. I you know George, I want you to keep on telling asking these things. One other question about the technology here before I take him off into the into the world of communications and business. George: Maybe a few more questions. I wanted to also ask about potentially the scary option of if you have a quantum computer, a cryptographically relevant quantum computer running Shor’s. If you can start thinking about RSA encryption which is kind of built in in all infrastructure throughout right now, and other math-based encryptions like ECC. It seems that a quantum computer could potentially be able to decipher all that communication. Do you think the world currently is ready for that kind of capabilities and what do you think we need to be doing? John: So, so I can comment on this from you know my point of view. People have known about this for a while and like the US government I don’t know in 2010 early 2010s they started really thinking about this. I’m sure they are thinking about the government encryption what they’re going to do and you know they’re worried about all this data being saved and then being decrypted at some other time. The way I’m looking at it is when you look at the promises of a quantum computer, I think there’s a lot of hype, okay. I think it’s harder than people think. But it might actually happen. So, what I kind of tell the general public is that you should be planning on this to happen maybe in five to ten years. You know kind of kind of that horizon. Sundar Pichai says three to five years, but as the CEO of Google he has to be sooner rather than later or else you know he’ll lose his job, right? So, I would say our numbers are consistent with each other, okay. Again, I think it could be even longer, but that’s a good framework. And that basically means people really have to start thinking about seriously changing it. But the nice thing is that we have quantum-resistant codes, NIST knows about this, I hear there’s hardware. I understand Google has these kinds of protocols in place that you can use in some way, I don’t think it’s standardly used, but people can use that. Yeah, hardware is in place so you know I think if people want to start changing it’s now the time to really think about that. And I think it’s very reasonable to do so. The one thing I’d like to say if people you know get worried about this is it’s very well known in history that all crypto systems have a lifetime. Okay and it’s unrealistic to think RSA would last forever. You know and even, you know, I’ve actually heard from Michel Devoret, we were talking about this last night, is the mathematicians now think that if something can be decrypted with quantum, there might be a classical algorithm that you know can do it. It’s kind of, it’s almost like it’s in the same class. We don’t know. Just may take forty years to do it, right? Yeah, it may take forever. But to you know figure it out, okay, but you never know. People are really smart. Yeah. So, you know the fact that we have to change it is not really surprising. And people just have to do that. Now I haven’t, have talked to someone in one of my trips with a company who worries about this. And what they say interesting enough is that present crypto systems have been designed for a long time and are around. But there’s like different crypto systems for different kind of communication. And that they feel that the change in this means that they’re going to have to have a more unified way to do crypto systems. Which they think will be better in the long run. And also, that the basis of the crypto system they will make it modular enough so that if you have to change again it will be relatively easy to do it. So, you know you should talk to the experts about that, but it may be that it’s time for an upgrade anyway. Right. And try to kind of do that better. So, you know so maybe this is a good thing that people are thinking about. Steve: And I keep on thinking, you know remember how in 2000 when it was changing from 1999 to 2000 and it was like oh my god the world’s going to fall apart, right? John: I was too young at the time unfortunately. Steve: Anyway, so now you’ve become a well-known figure, you got people asking you questions all the time and everything like that these days. Can you tell us a little bit about how your life has maybe changed a little bit with regards to all of that and how you how you deal with all these requests for interviews and opportunities and everything like that these days? John: So, you know obviously yeah, the Nobel Prize changes your life. What was interesting is I think I mentioned this before when we did the quantum supremacy experiment, I started getting the selfies in the hall and whatever. So, I got a a feeling of that which was very kind. It’s very nice. I like practicing giving podcasts and doing podcasts. And you know generally each podcast is a little bit different, I generally learn from the podcast how to express something or maybe a different way I haven’t thought about it. And yeah, I’m now a public figure where people want to you know ask these questions. I hope I can give you know honest answers about where we are and I think that’s good. I think that you know the field in general has a lot of kind of hype and a lot of murky concepts and I try to you know correct that a little bit because I think that’s going to make the field more stable in the end and you know better it’s better for us. Steve: Do you think that you know very scientists have a little bit bad reputation for not being very good communicators. Do you do you have any advice for scientists and engineers and people who are more technical about how they communicate or things that they would learn from you? John: So, I think it’s like a lot of other skills you have to practice. I mean obviously when you start out you know as a graduate student, you’re not a very good communicator. You get better over time. I’m going to say doing podcasts really helps me and do that especially because I have to use vocabulary that the normal person can more understand. Right? And you know and you have to practice that because you have a certain thing. So, it’s kind of interesting is your mind is split where you have the very precise vocabulary when you’re talking to physicists and then you, you know, have something for here and then you have maybe for high school students and then politicians, you know. Right. Making a joke, you know. You’ll appreciate it, I’m making a joke. But you know you have to learn that you have to use different vocabulary and you have to explain it. But in fact, you know if you want to if you think about systems engineering and running a big project, you have to know how to talk very deeply with people, but if you’re talking about the hardware to the software engineers, you have to talk in a different language. So that kind of practice in communication and different ways is something that’s very important. Now for me when I was a young student, I always read the Feynman lecture series on physics where he explains to first and second year students basic physics. But sometimes he goes into quite complicated things, but he has a way to explaining it in a simple way. And I’ve always tried to develop that skill and I generally you know when I give my talks I try to use these simpler concepts to make sure everyone can understand what’s going on. I don’t have to get into the jargon. I don’t have to talk about Hamiltonians to talk about qubits. Right? You’ll have to be careful with your jargon and to explain things. And what I try to do is try to explain what a quantum computer is is using electrical engineering language. Of having an instruction set, having an enhanced instruction set with quantum. And that our qubits are nothing but microwave resonators but they’re non-linear. And you know they have certain properties from that. Again. things that can connect to the average engineer and make sure that they can get a a sense of what’s going on. And you know of course here you know similar kind of things I want to talk about. Steve: In press and media very often, we get this attitude of you know the scientists and engineers are actually afraid to talk to press and media very often. Because just because it’s like oh my gosh, I’m gonna say the wrong thing. John: Well, you can say something that’s an error. Right. Right. And so, I just try not to worry about it. I don’t like looking at my previous podcasts because then I start critiquing myself too much. Okay. I should but that it’s just my psychology. I of course I get the feedback while we’re doing it here. And I think it’s mostly just the practice of doing it. And you know nice with the podcasts is I get one-on-one feedback as opposed to a big lecture. And then I can kind of tell if we’re discussing in a way that you understand. And then you just get better at it by doing it. Steve: Is there a time that you were like completely mortified or something like really went wrong when you were in an interview of some kind? John: I would say beginning as a graduate student it’s hard. And then you always get a little bit nervous when you do that. When I gave the Nobel lecture I wanted to do a really good job. And I think it came off okay, I was really happy and I made connections with all the audience and I try to do those things. And I try to practice that more to be a better speaker. Again, for me it’s just practicing it a lot. Steve: Right. And also, with podcasts I mean now you get a wide variety of different podcast subjects so that expands you know what you can talk about. John: But most of the time these are things I’ve kind of talked about before so you know you can kind of ad lib it okay. But every once in a while, you get something new and then you have to you know you have to stop and think about it. But you never know, people are really smart. Yeah. So, you know the fact that we have to change it is not really surprising. And people just have to do that. Steve: Right. George: I have no more questions, Steve, any last concluding remarks? Steve: No, I just appreciate the time you’ve taken with us. John: Thank you so much, John, we really appreciate it. This has been quite fun. It was really a blast. Really, really liked this, it was fun. Steve: All right, great. Thank you. Dr. George Schwartz is a Quantum Expert with Global Quantum Intelligence (GQI).Steve Lee is Communications Strategy Director at the San Francisco Agency. March 28, 2026