 I'm the senior executive editor for global technology at Bloomberg News and welcome to quantum's black swan a black swan is of course an unanticipated event often with not always but often considered to have negative repercussions quantum computing is of course the ambitious alternative to classical computing which uses bits representing either Zeroes or ones and replaces them with a quantum particle called cubits. I look forward to understanding that a little better. But luckily we have a great panel with me is Anna Paula a cease the chairman Europe Middle East and Africa of the IBM corporation. She leads IBM's AI and hybrid cloud strategy and it's responsible for IBM's relationship with the EU. Anna welcome. Thank you. We also have Jack Hittery from the CEO of Sandbox AQ Sandbox AQ is a SAS company at the nexus of AI and quantum computing Jack welcome. And we have Joel Mesot president ETH Zurich and Joel is a physicist and a researcher in the field of neutron scattering quantum materials and quantum materials. Well Joel I'm after after a couple of days at Davos my neutrons feel a little scattered. I wanted to mention that. But you know there really has been a black swan in computing over the over the last year and it's it's really been a generative AI and that of course has taken up so much of the attention here at Davos so it's you know it's interesting and I think valuable to spend some time on quantum computing and what where the opportunities still are and where it might intersect with the remarkable opportunities and challenges posed by generative AI. So Jack maybe why don't you start out and set our terms. Hopefully I didn't butcher that too badly but you know described to us. We're living right now with the reality of generative AI. Where is quantum computing and how do the two intersect. It's great to be here and great to be with my fellow panelists. It's a very exciting moment in advanced computing. So both AI and quantum really form part of a whole. And that whole is the ability to compute our world in very very new ways. And we look at generative AI and AI in general. There's many more tool sets in AI by the way than just chat GPT is and they're really going to change our world. Those GPT is run on GPUs graphics processing units. I know it has word graphics in it but we realize that we can actually run AI on that. But GPUs are good for something else. We can start to run the equations of quantum on these GPUs even as we're scaling quantum computers themselves and qubit spread that you mentioned. We're also now running the equations of molecules equations of drugs for Alzheimer's Parkinson's on these GPUs. We can actually run them now at scale. That's something new. That was not possible two or three years ago. The GPUs have gotten that good. Nvidia Alphabet Amazon many others making these GPUs now AMD in the game as well. And so what we're seeing is actually a tool set. We have AI in the one hand. We have running quantum equations on GPUs. We know those equations from 100 years ago. It was one of the universities where many of the greats of physics invented quantum physics 100 years ago including Einstein himself was trained there. Heisenberg and Heisenberg. And so we have a lot of history right here in Switzerland. But now also we can come to quantum computing and the cubits. There are seven major ways to build a quantum computer. Each has a pro and a con. Some you have to cool down to cooler than space itself. Some are room temperature. Some scale easier. Some maybe not. But each one has some interesting characteristics. We'll see that interesting race play out over the next five to 10 years. But now we have an array of tool sets. And ultimately what we're going to see is a meshed hybrid computing infrastructure. One where IBM building its quantum computer. Superconducting Cubits as an example there will mesh with GPUs and CPUs. So we'll have CPUs GPUs running AI running quantum equations on GPUs and then QPUs quantum processing units all in a hybrid mesh and code will get smarter and smarter. We know of course about co-pilot from Microsoft. We know about coding now with AI as your companion code itself is going to get smarter and realize wait a second where should I be running. Should I as code be running on a CPU a GPU or maybe a QPU. And there's going to be interchange amongst these different technologies. But I'll just plan to see now for maybe the next round of comments which is that quantum is not just about computing itself. Although that's a major portion of quantum. The black swan moment I believe was actually going to come first in quantum sensing. I'll leave it at that. We'll get back to that. I'm still trying to get my brain around quantum computing. And I'll ask Anna and Joel. You know if as Jack says we have powerful large language models running on GPUs using a classical architecture of zeroes and ones. Why do we need quantum computers. You know why is IBM making the significant investment that it is in this technology. Well I'll give an example of the work that you're doing with Cleveland Clinic. Right. Which is my hometown by the way. Oh great. Which is really simulating and understanding the behavior of some molecules in our body. Right. And how some proteins can really affect the way that our body behaves and identify cure by just understanding how those different components can interact. This is very very difficult to do with a classical computer. It would require a computer probably the size of the entire universe to process that we are reaching the limits of how you can create a small processors. Just to give an idea we are now reaching one nanometer chips in our development at IBM. The atom of a silicon is point two nanometers. So you're starting really to get to the limits of physics of what computers will be able to do. And that's why we need quantum computer to solve the complex problems that you're going to see going How much is IBM investing in the technology. We don't we don't publish how much we invest in terms of numbers but we have never missed one one milestone of our roadmap in the development of quantum and I agree with Jack 100 percent is not just about the Harder. It's about the software. You as you know we have developed a kid that is called kids kid for developing on quantum computers. We have now more than five hundred thousand people registered on the platform. It has been downloaded more than two million times benchmarks say that sixty seven percent of the developers that use quantum computers are using this kid that we have developed. And then of course there is all the ecosystem that you need to create. Because it doesn't it doesn't work if you have the computer the software but you don't understand what types of problems you will you will solve with this with this with this computing capacity. So it's really the combination of all these three elements that we are working on. Joel drug discovery new material sciences cryptography. I mean all these applications sound very promising but do we need quantum computers. Absolutely. Maybe step back because the word quantum is everywhere even with pharmaceutics as we have seen recently. So I see three levels and you mentioned actually all of them were mentioned. So first you have the quantum sensing. OK. It's great to have a huge computer but at the end the signals have to come from somewhere. So we develop sensors and quantum sensors that allow to have a much much higher precision than what we can achieve now with the usual sensors. So for example you can imagine from space to have a monitoring of climate change for example with such an accuracy that you can detect which trees. So this is more spectroscopy sensing and so on. So this is one level. And by the way quantum sensors are already part of our life. We don't just don't realize it. OK. We are using that. Then there is the quantum cryptography. This is a little bit a separate field. It's based on the same physics but it's a very special field. By the way one of the I think the very first company. I think one tick was in Geneva. Right. Correct. University of Geneva. Now it was sold to South Korea. Big company. We were not able to fundraise in Switzerland at that time. And what this company does is quantum cryptic. Really the messages. And if you want to intercept the message then the message is destroyed. So it's extremely safe. There is a side effect to it is that you can use quantum computers to break almost any code. The difference between a normal computer and a quantum computer is that you since how to say you work not only with zero and once but with any kind of combinations. And so the speed at which you can solve a problem is exponentially and not linear. And this is where you say you can scale a universal computer of the size of the universe to something that's sizable. Right. And so why we need that. There are some problems that we just can't solve. Those are problems that are extremely chaotic. Let's take traffic in one of the big cities. Can you solve your endeavors. So that's an easy one because they are just two streets. But let's say in a big city where you have a complex interaction of you know trams metros trains cars trucks. It's almost impossible to solve the problem now in life in real time. And that would be possible. Or the problem you showed with pharmaceutical. It's a same kind of exponentially growing problem that you cannot solve with a normal computer. OK. So those are these three different levels that we are discussing now you mentioned. Yes. The breakthrough is is done last year. I mean it's not new right. We have been in research we have been doing this. I remember using AI for my physics for my big detectors 10 years ago. You know. And but now it's in it's everywhere. And by the way the quantum cryptography from I.D. Quantic is in some mobile phones. I think of Samsung Samsung. Exactly. But unfortunately not. It's not allowed in Europe. I don't know why. So I would love to have this. OK. But I could just build on. Sure. Just pick up on one point that Joel said. And then back to Anna. First on quantum sensors. I'd like to bring out the point that AI is critical to the success of quantum sensors. Why is that the case. Well every one of us here flies. In airplanes and your lives are in the hands of the navigation system. If you're going over an ocean particularly there's no landmarks to look at the pilots and your lives depend on that navigation system. Today we use GPS. But what if GPS is not available over huge swaths of the ocean right now in the Pacific Rim area particularly near Taiwan. There is no GPS. Huge swaths of the Middle East GPS is not only being jam but being spoofed. That is even more dangerous because now the pilots. Think they have GPS and they're actually going in the wrong direction. Four planes went into Iranian airspace in the last four months. Not intentionally. So this is a major issue. We can use quantum sensors to detect the unique magnetic footprint of every square meter on Earth. The way birds and whales navigate. They've been navigating like this for for millions and millions of years. Just just define a quantum sensor. Sure. Great question. So right now we all have sensors. There's a thermostat in this room hopefully somewhere. Regulating the temperature. And those are classical sensors. What Joe was referring to before is classical sensors. Quantum sensors use a quantum principle the principles of superposition and entanglement these fun concepts that we can talk about some other time. But they use very specific principles of quantum mechanics to detect our world in a much more fine way. I'll give you a concrete example. The magnetic field of the earth magnetic field of the heart. Magnetic field of our hearts. All of our hearts right now we know have electricity because you take an EKG and ECG. That's picking up the electrical signal of the heart. But we know from the physicist that every electrical signal has a magnetic signal with it. And that signal from the heart is very very faint. So faint that you need a quantum sensor to pick it up in a good way. Now there's so much noise so much information. If you have an iPhone if you have a watch any of the magnetic signals in this room can confound that sensor. So we have to pass it through a GPU into an AI model a model trained on the data of what is a heartbeat look like and what is a iPhone look like and distinguish the two. And that's why right now the black swan moment of quantum sensing is upon us. We have quantum sensors right now in hospitals around the world today. They're going through their approval process right now. In the next 24 months I predict we'll have those approvals and they only working because we can pass the myriad of information magnetic signals that is picking up and distinguish between the heart and everything else because of AI and GPUs. So this convergence of AI and quantum is what's now happening. Both for heart sensing as an example giving us the kind of read that an EKG cannot. How many of us know a story of a friend who had a wonderful physical exam at a doctor's three months later had a heart attack. What was the problem. Did they develop heart disease in three months. No. Our diagnostic devices are not sufficient today. We need to move into the quantum realm to understand our own bodies force the heart and then of course the brain. Same thing with the brain. The brain has a magnetic signal as well. Very very faint. And for that we need to go to orders of magnitude more from 10 to the minus 12 to 10 to the minus 14 Tesla really really faint. And the same for navigation around our world. Boeing and Airbus have both announced publicly in December. This is all very recent that they're now going to be incorporating quantum navigation navigation boxes that are not connected to any satellite. There is no way to hack them. No way to get into them. They're not connected to the rest of the comm system and they can navigate again the way birds and whales do letting the pilots navigate completely with confidence without GPS. That is flying right now on these tests. Can I add something. One thing I'm passionate about is to explain where all this comes from because there is a lack of understanding that fundamental research has brought this. So you you mentioned this huge difference in orders of magnitude 10 to the minus 12 10 to the minus 15. It means zero point 13 zero and a one at the end. And those are typical sensor that we developed for particle physics experiments. You mentioned the neutron. So one of the measurement of the neutron is the electrical the other part. And you need a cage where we've almost zero magnetic field inside. And therefore at that level and all these technologies were developed over decades in the research labs at CERN at the Palscher Institute at Los Alamos and exactly. And this is where it's coming from. And there is one aspect of quantum entanglement that is very well known to everyone. You all heard about superconductivity. Right. Superconductivity is when electrons starts to play with each other's and make crazy things. For example you can transport electricity without losses over kilometers. So this is a manifestation of quantum entanglement at the largest scale. So this is just marvelous. What's nature can do. OK. Well these sound like very exciting and promising early deployments. But Anna would you agree that quantum computing as we're kind of traditionally describing it still remains very much in its infancy limited commercial deployment. And what technological breakthroughs still need to happen to really get the full fledged quantum computing with all of the opportunities and potential applications. So in the studies that we have done you really need capacity. You really need infrastructure of at least a hundred qubits and the ability to run at least three thousand circuits or gates in a in a in a processor in order for you to really demonstrate that quantum computers are better than classical computers or can substitute classical computers in solving solving those problems. So I think that is that is point number. So that's the goal. That's the goal. Where's IBM. We are very much at it right now. So we just announced Heron which is a hundred and thirty three qubit that is able to run eighteen hundred circuits or actually gates and our goal is to get to three thousand by twenty the end of this year by the end of twenty twenty four. So with that we're going to start really seeing quantum computers able to solve to solve their problems. Now for you to really get to a point that is commercial then the scale is going to be seen. Are they are they economical. They are. I mean as long as we as we have problems that are worth right investing. Of course they're going to there's going to be the return on the investment. But then we are going to be talking about billions of qubits in a quantum system. And the next investment or the investment that you're that we're really focusing on is how you reduce the errors that happen and the noise of those systems. So how you isolate the the multiple components of a quantum sentry computer in a way that those qubits behave and are coherence to deliver the outcome of the operations. And I'd like to just go back to the point that Joe raised on security. Right. One of the big questions is is quantum really going to create a cybersecurity amour get them. All right. Any any it's going to. Right. I mean it can really decipher all the the crypto keys that exist today. But there are already algorithms that we have developed in our labs and in partnership with the with the Institute of Standards of the United States that demonstrate that they are going to be protecting data from a future quantum attack. Right. Or quantum quantum enabled attack. So I think that if we combine all these elements plus create the ecosystem of developers in order to drive the applications that are going to take benefits of those computers then we start to create really an industry around quantum computing. I just want to mention that this point of I explained what it is about. You know when you have a normal computer sent suddenly it doesn't the disk you have or the the memory you have doesn't work anymore. It's because some of the bits have been flipped over. Yes. And this is because there was an interaction maybe a magnetic field passing and a quantum computer since it involves many particles working with each other's are very sensitive through any perturbation even warming a little bit and then it disorders the system. So there is a lot of work to do to get this noise down by a few orders of magnitude. And I just want to mention. Yeah. Because you need to run. Exactly. So we two years ago. Andras Val have one of our leading researchers that you probably know also was able to show that you can have a self correcting system. So if there is an error it will self correct in the system. That's what was quite a break. Jack you have something to say. But I also want you to address the timeline for a commercial scalable quantum computer. Sure. I want to build on what Anna said in terms of building up the quantum computer and also the breaking of cryptography. Every one of us in your phones right now you have secure cryptography with you know it or not. That's how you're communicating with your bank with your telco. That's how you're sending messages when on WhatsApp when it says encrypted end to end that little message. This is what we're talking about. That's all gone. That's going to go away. Now how is that going to go away. Anna correctly mentioned before hardware and software. Let's say I have a river and we want to build a tunnel under the river. We don't just start from one side and they just keep going. We start from both sides and then we meet in the middle. So by analogy if I have the hardware folks IBM is doing an incredible job advancing the superconducting quantum computing methodology as an example. They're digging from one side. The algo people are digging from the other side. And what's happened is in paper after paper we've seen that the number of cubits a number of these special things that we need to actually crack RSA to crack the kind of cryptography is coming down. So the two will meet faster and faster under the river to make this tunnel that breaks that breaks the banking system that breaks the telco system that breaks the energy system that breaks government secret. So this is a major issue. And the goodness is Anna mentioned NIST and other organizations here in Switzerland around Europe have banded together to create a set of protocols that replace today's protocols. Those are called post quantum cryptography. But it takes about seven eight years for a bank for a government to transition to this news to this new protocol. So what's very important right now is that we understand that this does not work in just a linear fashion. People got surprised this year by Gen AI. People got you know those of us who've been in AI for many many years knew that yes we were already building LLMs for quite a few years. And so what's going to happen here is the same thing. People are going to say like wow what a surprise and shock that all cryptography is broken. You heard it here today at the World Economic Forum that this is going to happen. I'm sorry I'm so fixated on timelines. OK. I feel like we have the same conversation at the end of every year. When will we be surprised by sure. So again because the hardware is being developed at a faster pace and because the software the algos that actually crack. People are building on a 1994 paper by Peter Schor at MIT. But that paper just serves as the background. Now they have we have even better algos to use with fewer cubits. And so you know over the next number of years we're talking now if we look at some of the timelines certainly by 2029 2020 30 we're going to see scaled fault tolerant quantum computers. And you might say today I need a certain number of cubits. And my prediction to you today Brad is that the number of cubits is going to come down that you need because the algo people are figuring out more and more interesting ways to crack your cryptography. You can think about the brass prize of being able to to decrypt everything in the world. And so this is a major issue that the governments the world have come around on. Industry is a key partner as well as academia all coming together to address this issue. But we have to act now. Well and there's a lot on their plate. I mean they're dealing with Gen A. I. There's still social networking and the disruption to the democratic process. So Anna and Joel is there. Is there enough urgency around the opportunities and challenges that will be posed by quantum computing. And what kind of international collaboration. You're suggesting it's already taken place. You know is there enough attention focused on it. There is a focus just to give a number. Since we since you opened our quantum computer on the cloud in 2016 there have been executed more than three trillion of circuits in our systems. So the amount of experimentation and driving really you know the usage of how can we really demonstrate the power of the computing in quantum is happening. I agree with the timeline. That's the timeline that have been working under. I mean by the end of this decade we should be really see our quantum computers doing very meaningful and commercial tasks in terms of collaboration. We have established collaboration for example recently with the University of Tokyo and Chicago with the support of academia really to advance in the progress of the development of the technology. So we haven't seen a slowdown actually we announced at the beginning of last year partnership with Iker Baskin Spain in the north of Spain to set up a quantum computer there. IBM is establishing its own quantum computer in Europe in Frankfurt. So we continue to invest and the road map hasn't hasn't changed because of all the things that we are doing now also in in generative AI and artificial intelligence and generative AI is going to be even it's going to bring even an additional wave of innovation because for example we are infusing generative AI in Q-skid to accelerate the development of code for quantum computing. So it's hard to give timelines because things are happening faster than we have originally projected. I think we should realize you said what you have now and at the end you will have millions of qubits. So the facility will not be an iPhone. Right. Yes. It will be a big facility. And one of the worry that we are sharing in the scientific communities that we should not land in the situation that we have now with AI where this is in the hands of a few companies or a few states that we don't know too much what they are doing with. And so I think academia has a role to play to make a system that is open to everyone. And the way you do it now we should proceed with that. And sometimes I speak of a certain of quantum that we should at whatever level the best would be really international. Maybe European can play a special role. Europe can play a special role. Switzerland being I hope soon again associated to Horizon Europe. And so I am discussing these ideas with our colleagues. You have to know Switzerland is excluded now from any quantum program in Europe. Right. And this is a strange because it's political. We have the ambassador negotiating here. He can explain more. This is a we are kind of playable of politics. Sorry. And so so but I think with or without Switzerland we need some entity of that size because it's going to be. I don't know what you don't want to speak about cost but it's going to be a you know very expensive. Right. And it would be really bad if it's only in the hands of a few either states or companies and the rest of the world would be mean isn't that unfortunately inevitable. That's over. It's expensive to build to design. How how do they how do you know won't wealthy countries have have access and not wealthy. This is where I think international organization will have a role to play. And then that's why we involved also there is a Geneva the Geneva science and diplomacy anticipator and we are talking about this with the diplomats and so on. So we need to have this connection as well. Right. If I can just build on Joel's comments. There's two calls to action that came out of the weft in the last four months that I'd like to bring to and there's a new paper published. Yeah exactly. So the first is the quantum divide. This is a key call to action that speaks to this issue. IBM ourselves and others involved in putting the quantum divide call to action together. It's it calls out that there's only 25 countries right now that even have a national quantum blueprint. Even the start of a roadmap of where this is going. 180 countries. Nothing. And so there's a specific set of recommendations about how to get this out there. Not to accept that there will be a quantum divide the way we had a digital divide. Those of us who remember there was a divide now for the last 25 years. Those with access to bandwidth to broadband those with access to mobile phones those without. We're finally closing the digital divide in the next three to four years as 400 million people every quarter get a smartphone for the first time in their lives. India Geo Mukesh Ambani is one example of what he's doing in India. 600 million people will get a mobile phone in the next two years that never had one before. But the quantum divide is just now beginning. And so this call to action that we put out with the weft is around that. But yesterday we put out another call to action again IBM ourselves others at touch and academic partners. The quantum economy. What we have not yet articulated until yesterday is that this is not just a piece of technology. This is not just a bunch of research that's happening behind some walls. This is fundamental stakeholders basic research happening universities core companies established traditional companies startups like ourselves. Many others involved in creating a whole new part of our economy. And it's critical that we get everyone involved in that economy. The majority of the global one one thousand companies do not have a robust roadmap yet as to how they're going to use AI and quantum together to solve core problems. When we look at pharma I'll just look at pharma just for a second. And we look at the fact that it takes 14 years on average three billion euro to make one molecule to become a medicine even with all that time and money 90 percent failure in clinical trial 90 percent failure here in Europe 90 percent failure in the US in Canada and most of the developed world. So the numbers aren't good. AI and quantum working together can change those numbers. This is a time to change. If I had a construction company and said hey Brad I'll build you a new house. There's a 90 percent chance it's going to fall down in the first 30 days. Would you like that house probably not. But that's what we have right now in the pharmaceutical world. We now AI and quantum working together can change that. We can make digital twins of that molecule for Parkinson's for Alzheimer's for brain cancer run in simulation and synthesize new data. 2023 might have been the year of generative AI. One of the things that 2024 25 about is generative data using the tool sets from the companies and universities represented here. We can actually use quantum principles to generate high quality accurate data as to how one molecule will dock with another. How a drug could dock into the brain into a tumor. These AI and quantum tools are now with us and they'll get better and better as the machines get better. So this is a moment and that's why the web put out this call to action called the quantum economy and the quantum divide. Right. And I want to follow up on that. But please start thinking of questions for our panelists. I'm very aware that we're just scratching the surface here. But to follow up on your point and I want to ask you all can generative AI language models running on quantum computers hasten the development of AGI. And in fact do we need quantum computers. Do you think that's a lot of jargon bread. I know maybe deconstruct it for a group. I can kick off and please others join in. There is an emerging field of QML quantum machine learning. And one of the areas that's most interesting about thinking about neural networks these distributed networks inspired by the brain from the 40s and now leading up to gen AI to GPT things like that is to process data that itself comes from a quantum sensor or the quantum world. Right now when you have a quantum sensor and you start analyzing it with a normal computer you have to collapse that data that data now goes back into the classical world. But there is the possibility of taking a quantum sensor sensing something and keeping the richness of that information directly into a quantum computer. A number of papers have come out from a wonderful researchers. It's right now in basic research right now. And about what would happen if you have the full complement of a quantum computer working on quantum data in its purest form. So that's one interesting area. It's not really going to be applicable in this year or next year. But over the next five to seven years I think that's an interesting area of development. Yeah. Well I think that normally one of the distinctions we need to make is that quantum computers are not going to substitute classical computers. Right. I mean that's not they're not going to substitute classical GPU CPUs don't go away. They're not going to go away. Right. Because there are certain types of problems that quantum computers are really not good at solving the terministic outcomes for example are not probably the best usage of a quantum computer probabilistic computer. Quantum computers are for operations that need to simulate how the nature how the nature behaves at the end of the day. So and many times it's not going to be even cost effective to run certain applications on a quantum computer. So it's not a substitute. And that's not the purpose. I think it's really about the combination of the two the two capabilities the classical computers with quantum bringing AI on top of it. I totally agree on the aspect of probably quantum can help us get better data to solve the problems that AI is going to work on. So I think our strategy continues to be very much a hybrid one where we'll have to continue evolving all the all the fields of computing in order to achieve the ability to solve this complex problems. If you want this is nothing new. We don't use a supercomputer for every simple calculation. Exactly. OK. You dispatch the easy part to a normal computer or maybe cloud computing and then you do the hard part and then you recombine. And the idea is to do exactly with the quantum computer. And by the way we are running the supercomputer of Switzerland and we have 10,000 GPUs which put us in a good position because try to get additional GPUs today. It's kind of tough. So we are there. And then we already combined with cloud computing. And now we the idea is to if we get a collaboration at some point we will put also the quantum computer at this site. Let me go back to my sorry. Go on. No I want to come to a different point on the preparation for there is something we should not forget is how to prepare the workforce for what's coming. Right. So we took already a few initiatives at ETH we were the very first to have a master on quantum computing including not only physicists but engineers, electrical engineers, mechanical engineers, mathematics and computer science or it's really transfers. It's highly successful. I mean they learn in the first year to build two cube heads. I mean I would have dreamed to have this right. It was impossible at my time. And so that's one level. But those we need to think about the next level vocational training. So we started to think because Switzerland is very good at that. So we started to at ETH we have 150 vocational students. And so we start to integrate them. And you know this combination of highly skilled vocational trained people and engineers is extremely powerful. But we need to start now to get them ready in five years. Yeah. Let me talk about a little of a few things on this on this area. So we have up to now is killed about 8 million people around the world on programming in quantum through our collaborations with the schools around the world. And there was one very interesting in terms of how two companies come together in that field. Last year we announced with the credit mutual in France to create a quantum academy. So they are preparing their workforce in order to be able to operate in a quantum world. And this is becoming very much a reality where companies come together and get the collaboration of on our side bringing the technology on their side identifying the problems that they are going to to benefit the most from applying quantum computer and making sure that their workforce is prepared for that. And before you before you weigh in Jack please I guess raise your hand if you have a question for for the audience and do we have a microphone. So why don't we'll do that. And yeah why don't you I just want to come back again to this interplay with with AI back to back to the scaling also that Joel and Anna just discussed. The only way to do this at scale to train literally billions of people ultimately that will need to know something about both AI and quantum is to AI based training. So here at Davos we're announcing with a number of partners later today one of the first rags I know another acronym yes that you'll have to know. So LLM's large language models GPT's we know about that. But they're great at speaking and doing language and writing essays things like that college kids are loving it. But when it comes to their knowledge base often they hallucinate. We know about that hallucination. One of the ways to address that is to a retrieval augmented generator a rag and a rag means that it uses the language knowledge it knows by being trained on the web and other kinds of corpus is like that. But when it comes to a knowledge base if for example in today's announced we're talking about health care training training to be a doctor trained to be a nurse. It only looks at valid medical and nursing textbooks that we gave it any kind of training that does it has lesson plans and quizzes and videos. It only looks to this corpus. No hallucination from science fiction. People ask recently how many people live on Mars to one of the GPT's. It said two billion because it read a science fiction novel. That is not acceptable in the medical realm. This. How do you know. And. And so. So here too all of us on this stage companies academics and others have a responsibility to think about again billions of people that need knowledge around this and I think is one of the best tools. But we have to go beyond just the LLM's of today. We have to now go to rags. We have to go to other techniques that take us into validated high quality information as we're doing this training. And I think that's something at scale that can help close this quantum divide that is now just opening. OK rags rags. We got to work out our acronym. Yes. Yes sir. I'm John Chippman the executive chairman of the International Institute for Strategic Studies headquartered in London. And of course we work a lot on the reality that technology is a principal arena of geopolitical competition now. And reference was made to regulation in the AI world. People generally speak or Anna does about three types of regulation. A market driven approach. A rights driven approach and a state driven approach. US EU China. What would be the right regulatory approach to quantum. Who's going to lead it. And what kind of competition will there be in that regulatory space for quantum. Well. Great question. I do believe it's probably going to follow the same pattern that we saw in other in other waves of computing. And I think that European Union is probably going to play a very strong role in coming up with it. I mean President Biden issued an order around security. Right. Some quantum safe methodologies going forward. So I think that will probably be on the places where the technology is developing faster is where we are going to see the framework of how this should be applied in a safe way. Just to build on that. It is true. Not only did the president put out a number of executive orders but Congress actually pushed through and passed a cybersecurity bill specifically around quantum safety. And here's the amazing thing. We were in the room. The Senate was voting on this particular bill. A hundred hands went up saying yes. Have you heard that happen recently. A hundred hands in the Senate. A hundred out of a hundred saying yes. How deep was the understanding. It might be easier than with AI because of one problem I mentioned that it's quite sizable facilities. And so it's easier to regulate because you don't have this wild west of everyone developing codes and stuff like this. So I would be more optimistic that we can regulate this in a better way than as it is with AI. With AI it's developing so fast and so diverse. It's going to be very difficult. But again I think it's important that to bring the point that we shouldn't regulate the technology which we have to regulate how the technology is used. Right. And the risks that it introduces and not the technology itself. I think we have one question behind us. Thank you for an amazing panel. More at Spare Lens partner with Lightspeed Venture Partners. Jack if I understood you correctly you said that we might have initial commercially viable quantum computers as early as 29 30. But at the same time you said that governments will need seven years from today to transition to quantum proof. You did the math. You exactly did. The session is titled Black Swan which has a connotation of kind of like probability or chance but it sounds like we're actually guaranteed screwed. Yeah. This is exactly. It's a very very good point and well put. This is a situation that you see a train wreck happening right in front of you. We know the speed of the train roughly it's getting faster that I can tell you. And we know that's going to hit a wall. And that wall is called the entire global economy. And our economy the reason why it's 100 plus trillion dollar a year economy globally is because one of the key drivers of that is cryptography. I know people often don't talk about the economic classes but hopefully they will soon at Etihad and other places without cryptography. We would not be able to do e-commerce. We couldn't do the Swiss system. We couldn't have bank. We'd have to fly to Seattle say hello to Jeff Bezos and say here's my credit card. Actually he moved to Miami. So now we have to go to Miami. But the point is that we can do this over the internet over the open internet because of cryptography because of this public key cryptography. This is what we must now replace with a better system. And as you pointed out it does take governments and banks and others seven eight nine years to do this. Here's the upside. AI that we've been talking about already is transforming hacking ransomware that we all know about already is a major problem the last few years. We are predicting five extra ransomware this year compared to the last few years. Why is that because hackers know about GPT. They can make spearfishing emails much better now than before. And so when you look at your cryptographic status as a bank as a government as a telco. Yes your future proofing yourself for quantum. But you're also helping yourself today for attacks against ransomware. Right. And so this is a dual positive that's happening right now in the cybersecurity world. Cryptography was never really looked at that carefully. People looked at firewalls. People looked at malware viruses worms. You heard a lot about that. But now it's time to look at cryptography as a core part of cybersecurity both for today's attacks ransomware AI driven ransomware and for the attack that is literally guaranteed to come. Probably have time for one closing short comment. There was a question. I think. Yeah. Yeah. What about what about let's let's do this very quickly. State is state of the art in China. Are they working on it. Of course. Of course they are. That's all I can say. Yeah. OK. Yeah. China. We are not working with them. That's all I can say. Publicly with Panjian Wei is a well regarded physicist. He leads the China effort. More than 2000 engineers under him. It is a very serious and very competent effort along many lines. Quantum computing for different modalities of building a quantum computer there. All public information. They publish in peer review journals. So this is something you can read about. Superconducting neutral Adam. Many other modalities. Quantum sensing very strong program. And one of the areas that China has focused on more than most is quantum cryptography. Quantum communications. A parallel internet to today's internet that can have coherent IE signals of superposition and entanglement. That's an area that China has really focused on as well. OK. We're going to have to leave it there. Please join us back here in 2029 when Jack will present a fully functioning quantum computer. Jack IBM and I'm Anna Joel. I want to see it here. Thank you guys for a great discussion. Thank you.