WISeKey International Holding AG (WIHN) Earnings Call Transcript & Summary

Ladies and gentlemen, good afternoon. Hello. If you can just focus on us few minutes. Ladies and gentlemen, we're going to start because we have 3 roundtables, and we have a lot of speakers, and it is a pleasure once again to host you at the WISeKey SEAL Davos Event. So this is the 19th year that we do this in Davos. So WISeKey, being a Swiss cybersecurity company -- thank you very much -- being a Swiss cybersecurity company has been very much focused on bringing cybersecurity as the main discussion of topic in Davos. We started very early with a small breakfast. We got bigger into big events and now in Europe hotel, which we have been doing for nearly 10 years now. This event is in cooperation with the Tech Accord. We have speakers that represent this amazing group of Tech Accord, which are companies -- leading companies of the world, and they are working together towards cybersecurity issues and defending critical assets around the world. So today is going to be -- the focus is going to be artificial intelligence, but obviously, cybersecurity has a major component on what is going to happen in artificial intelligence in the future, as AI has been reported by the WEF to be one of the risks in the next year. So is that a risk? Or is that an opportunity? Maybe both. We're going to debate that today. So I will -- the problem we have is that we have a very short time, and we have a lot of panelists. So we are requesting, and I really apologize for that. But I think it's great, too, because we have a diversity in the panel from different backgrounds and different industries, which makes very interesting. We also have 2 writers. There are 2 books. One done by Mark Minevich, Our Planet Powered by AI, which is a bestseller in the United States. And also by our friend, David Shrier, basic AI. This book, by the way, I think, is available, right? You're -- yes. Great. So you have 2 amazing writers with us as well, which are going to comment. So in order to start, I will ask the panel to introduce themselves directly, I think it's the easiest way, and then we start with the Q&A. So the microphones are available. Yes, Anastasia will give you the microphone. So maybe let's start with you, Navroop. Yes.

Thank you, Carlos. Yes. Okay. So now it's on. All right. Thank you very much. First of all, I just want to thank Carlos for -- you're the real veteran here. So thank you for hosting this event with such great people. My name is Navroop Sahdev, Founder and CEO of The Digital Economist. Many of you also know The Africa Coalition. We are also the [indiscernible] of that. So welcome to the community as well. So what we do is we are [indiscernible], U.K., based in D.C. We work on all things [indiscernible] governance and build up knowledge [indiscernible] economy. When we talk about [indiscernible] AI and emerging technology [indiscernible]...

If you don't mind, please keep silent, so we can -- excuse me, if you don't mind to keep silent so the speakers can really speak. I hear a lot of noise. Please, if you need to speak, maybe you go out or you go in the back, please. Thank you.

Yes. I'll pause here. Hand it over to you.

Thank you very much. Pleasure to have you.

Thanks, Carlos. Danil Kerimi. I'm advising a number of organizations. But here, I also very much honored to be as a Board Member of SEALSQ. And I just want to -- but those of you who don't know, Carlos adds one speaker per year. So the rest of us that are just sitting here all year waiting for the new person to arrive.

Thank you very much, Carlos. It's a pleasure to be here. Thanks for all your leadership that you do in the IoT, cyber around the world. It's great to be here in Davos. And I'm a book author, as Carlos mentioned, but I'm also a strategic thinker on AI, how it relates to the national AI agenda. I'm also the leading enterprises worldwide, leveraging some of our capabilities. I'm also an investor in a number of companies in the AI space. And I'm also a columnist at Forbes, discussing AI. So I am [indiscernible]. I'm excited to learn about what we are going to do as a group regarding security, and AI is a big topic, and I'd like to discuss it, representing the United States. Thank you.

Thank you very much, Mark. Teresa?

Hi, I'm Teresa Hutson. I work at Microsoft. I don't think I need to explain what that is.

Especially now.

I'm Mark Hughes. I'm from DXC Technology. We provide solutions to [indiscernible] clients from all those threats out there, the ever increasing vulnerability that we see through the complexity of the state, means that we have been really focused on making sure that we can reduce and respond, reduce the time it takes to respond and to resolve very quickly. And AI offers us both opportunities and threats. Looking forward to talking more about that later on, Carlos. Thank you.

Thank you very much, Mark.

Hi, I'm Cristina Dolan, and I am a cybersecurity executive, and I've been an executive of a number of different cybersecurity companies, in addition to an AI company before it was called to be an AI company. Also an MIT Alum, so we have some MIT connections here, but thank you so much for having me.

Thank you.

I'm David Shrier, a professor -- this one is better, they say. I'm David Shrier, Professor of Practice at Imperial College London, where I lead the trusted AI Alliance, which is a global initiative of over 3,000 AI researchers trying to help build responsible and trustworthy AI at the country and global scale, and I also run a venture studio that has 28 portfolio companies in AI and deep tech.

Thank you very much.

Hi, Carlos, and thanks for having me again. I'm Alissa Starzak. I'm Head of Public Policy at Cloudflare. If you don't know us, we run a big global network. We protect against a lot of threats and have a lot of data that AI is very helpful for. So looking forward to the conversation.

Hi, everyone. My name is Hossein Rahnama. I'm Professor of Computer Science at Toronto Metropolitan University and a visiting professor at MIT Media Lab. I also spun off an AI company called Flybits, just reached 25 million active users. So thanks for having me.

Thank you very much.

Congrats. Thanks, Carlos, for having me. Peter Selfridge, I have just one job, unlike everyone else here, government and public affairs for SAP, the small German startup.

So a lot of fuss there. We -- really the China dedication here with 20 people, most of they are in Davos. And I have been in Davos for fifth time. And with Carlos, we are very familiar. And I'm the founder of [indiscernible] Group, a very -- a pretty old blockchain company. We have 11 years' experience in Web3 areas and 4 years' experience in Metaverse, and we are trying to combine AI with Web3 and Metaverse. We've invented a new word called AWM. So this is super metaverse because it is created by super intelligence. Yes, that's all. Thank you.

Thank you very much. So maybe before we start the debate, and as this roundtable is co-hosted by the Tech Accord, maybe, Mark, do you want to explain what the Tech Accord is and give the audience some idea on...

Yes. So the Tech Accord is a group of mainly Fortune 500 businesses, who come together to look at in the cyber context of recent cyber events and initiatives to really come together to share information and to really help formulate and shape policy with policymakers around the world. So many different initiatives. We pick 3 or 4 themes every year that we really focus on, AI and the use of AI with cyber is clearly one of them, and how we can really take benefit from and also protect against where threat actors and are making use of AI, but there's a number of other initiatives. We've got one at the moment running, where we're looking very closely at how we're going to deal better, more effectively with the law enforcement globally in terms of apprehending cyber criminals. So it's a fantastic group to be part of, not least of which. It gives me the opportunity from DXC with our other members to be able to share information very rapidly against emerging threats, which is extremely useful.

Thank you very much. WISeKey is a proud member of the Tech Accord for many years. So we have been following the development worldwide.

I will -- I just had some questions, but we would like also obviously, to be free to add whatever we want. But I'm going to start on the A, safety and ethics. This is a very big subject here in Davos, actually just need to work around the street, and people that have been in Davos for many years will realize that before a few years ago, we have e-commerce, everybody was in e-commerce, then everybody was in NFTs, and everybody was on crypto. And this year, everybody is on AI. So it is amazing how Davos adapts to the technology evolution, on the Fourth Industrial Revolution [indiscernible]. I was in the UN for many years, I was the Chief Security Officer of the UN. And obviously, before we called that AI, we knew that this was reaching a point where ethics and the way we're going to treat AI is going to be a concern. It's both a risk as an assessment of the WEF, but it's also a big opportunity. So maybe, Teresa, at Microsoft, you are a major player on AI. What is your vision on that, on ethics and security?

Well, I think it's -- you have to build it in. You need to think it well, you need to think about...

Excuse me, gentlemen, again, if you don't mind not to speak inside the room. Carlos, please, if you don't mind [indiscernible] there. Yes, please. Thank you very much. Sorry, Teresa.

You can do it? Yes, okay. So as we're thinking about, I think, any technology, it can be used -- my boss wrote a book called Tools and Weapons. And I think that, that actually really does frame how we think about building technology. You have to really think about how you operationalize building in those checkpoints to say, are you thinking through the impact on people? Who -- what are the unintended consequences? What can we learn from other technology that has been out in the world and has a good impact on people? Social media gives people a chance, for example, to participate in democracy, but it can also be used for harm, and we've seen spread of miss and disinformation. So what can we learn from that? How do we operationalize that, both through principles, but then really through engineering process to say, pause and think about this, I think it needs to come from the leadership of the company and luckily it does.

Absolutely. Thank you very much, Teresa. So maybe staying a bit on the policy aspect, Mark, you are advising many governments, including the U.S. government on AI policy. What is your advice to them? What are you telling them to do and how to prevent eventually that AI becomes a future liability on a problem versus an opportunity?

I think President Biden hit it correctly. President Biden's Act is enacted balances innovation. It balances the view that you need talent more than ever globally and in the United States. At the same time, you need to put heavy guardrails on issues related to our national security, the issues related to our identity, to data leaks. And I think it is very important that you balance everything. You don't take a very heavy regulations view because you're going to stifle competition and you're going to stifle innovation. What we see in Europe is innovation, especially in AI, maybe uncomfortable for others to hear, but it has been depressed because heavy, heavy regulations. You have to know that the most important thing, I think, as my colleague mentioned from Microsoft, it's about values. It's about values that you put in, in AI models and systems. And if you develop it correctly the first time around, you structure it, you organize it correctly, you will have systems which will follow certain human-centric approach. And now we're not only in a human-centric approach, Carlos, we're in so-called planetary approach. This is what UNESCO calls, and this is why my book is about, because we have billions of IoT devices. We are using AI to solve the most complex challenges in the world, from health care to safety, issues related to our climate. And we have to make sure that we put proper values and education, and we have diversified teams. This is what I'm saying, not only in U.S., but across the world, diversification, value creation and balancing, never stop innovation process, but we got to do it correctly.

Thank you very much [indiscernible]. Maybe if we stay on the government and AI use. Peter, SAP, what is -- I mean, you are, as a private company versus being a United Nation Organization or other, what is your approach you talk to governments about the use AI? It is being a European company, an asset or not? What is the trust companies required in order to be able to provide AI solutions to government? Maybe some of your insights on that, please?

Yes. Going back to GDPR, we turned that into an asset because being a European company, we are on the front lines as that was being developed. I know many of the companies here probably lobbied along with us to make that a little bit more favorable to business. But ultimately, being in Europe and having those regulations rolled out in Europe, we are ready. And so when the rest of the world started incorporating their own versions of GDPR, we are ready, our products were ready. We look at the AI act in the same way. We actually applaud a lot of it, not all of it. Obviously, it's not finished. We haven't seen the text yet. But we think it will give some companies a leg up if the final version doesn't go too far in their 2-tiered approach and overly prescriptive regulation on large foundation models. So I'm assuming a lot there, but if that goes in a similar direction as GDPR, we use that as an education tool. So sometimes being a European company amongst a sort of a heavy regulated market comes into our advantage.

Thank you very much. Again, I want to repeat again, if you don't mind, especially in the back room, if you don't mind, do not speak, please, because we hear you, and you're supposed to hear us, not the way around. So that would be great. Thank you very much. So let's just stay there a bit on geopolitics and AI. We have China as a representative on this roundtable. And China is investing heavily on AI. First of all, how do you see the global cooperation of AI from the Chinese perspective? And what is China doing in order to make AI inclusive versus being AI separated from the rest of the world?

Yes. So I think our people also -- usually, we compare the AI with the nuclear. So we think the AI has more put a lot of potential, and it's very dangerous. So we can compare it to nuclear weapons. So between China and the U.S., we have a lot of talking about the nuclear weapons. And also I think about the AIs too because AI is far more potential and far more dangerous than the nuclear itself. So I think on this side, this is one topic. China and the U.S., we can sit down and talk together. And another thing is that China has to ensure it's a national safety. So we have to -- the NVIDIA, we cannot buy the high-quality car. So we have to buy Huawei. And we have to invent our own system, our own license, our own whole ecosystem for the AI. So I think for some kind of -- it's good because it's a backup. If we have something wrong in the U.S., maybe China can be a backup. So this is something I can suggest. Also in China, we want to make the AI to be faster -- developed fast, but I think that there will be several issues we should think about. How can we make the computing power, computation to be democracy? So it cannot be concentrated to a few companies. This will be a key problem. The second is about the data democracy. How can we protect people's privacy, data privacy. This is also very important for the China government and also the China industry. And the third one is about the big LLM and all kind of AI models. So we are creating a Web3 [indiscernible] phase. So this will be provided to opportunity for more developers to develop their own projects and use distributed to compute -- computation. This is very important.

Thank you very much. Let's stay on the countries. We also have Africa, right, which is the continent that we expect to be one of the leading continents in the future. Navroop, what is your work in Africa on AI? And how you are explaining the AI evangelization in Africa?

Great question, Carlos. By the way, just a disclaimer, I have a Indian passport. But I do think Africa is the next super power growth engine, if you may, after China and India, which is -- with current engines are. In Africa, the work on AI is specifically talking about the cultural context, for example, of AI. That's a debate that's missing in the U.S. It's a debate that's missing in Europe. We talk about biases in gender, racial bias, so on and so forth. Actually, a lot of people don't know, but I think a lot of the models are being tagged by Kenyan nationals, and they actually don't have the ability to -- or the operational capacity to build their own models. So how do we kind of have this kind of situation similar to where like the Indian engineers are building components and large parts of American companies? So I think we're sort of starting to see that play out with through the African continent as well. But I think it's an opportunity to leapfrog, right? And China did that successfully for many technologies, I mean, years ahead of even U.S. as well. And so the question is, is this finally the opportunity? And I just want to emphasize that it's not just AI, it's digital. It's broadly digital technologies that is providing the basis for all of this. But now you have -- you can tap into the knowledge of the world with AI. I think that's a very powerful thing. I mean I think just this room being full, it speaks volumes about that. The fact that we have a demographic dividend that Nigeria is going to be kind of coming up with after India, which is the youngest population right now. So you will see the I think the influence of the continent across the world as well if it's done correctly. And that's up to Africans to do really with the help of the world.

Thank you very much. Maybe just continue there with Danil. Danil, you actually brought the WEF to China, right, when you were -- used to work in the [indiscernible]. Yes, you help a lot, but I know that you help a lot in [ Dalian and Tianjin ]. So you have a lot of experience to enter into a new territory with technology. What is your views on entering into new territories with AI? What is the key things that you think need to happen in order Latin America or Africa or Middle East, we get at the same time because otherwise, we're going to create -- I remember the digital divide, we're going to also create the AI divide, right? And we need to be careful about that. So what is your views on this?

I would like to actually go back to the book that Teresa has mentioned, Tools and Weapons. And I'm sure that, at least Brad's idea will come from [ Geneva ] where he started, which is where we are based...

He was with us 2 years ago, actually. He was here.

Yes, sir. So divide is one. But let me tell you just a very small story, and then I will be very quiet. If you stay here in Davos till Saturday, you must know that the security here is provided by the Swiss Federal government. If you stay here till Saturday, you actually will see soldiers coming down from the mountains, and they would meet their girlfriends here in the local bars, and there will be iconic scenes like Times Square kisses and all that stuff. But one thing you probably don't know is that Swiss Army loves us. And the reason why they love us is because they get to exercise once a year, and this is their full-fledged exercise. Swiss Air Force loves us too because they get all the fuel they will need to fly all the jets. And once a year, they get any budget request approved just like that. You know who doesn't love us, Swiss cyber forces. They don't need us to exercise. They are exercising every day. And for -- AI is a complete game changer for them. It's like they just went from bow and arrow to gunpowder. And what did gunpowder do to Latin America and Africa and others? Colonization, redefining of boundaries, new dynasties. This is the game-changing moment for all, and it's happening much faster than anything we have seen before. It's happening at the rate that is faster than the Moore's law. And AI is disrupting not only the governments, but also the boardrooms, right? Cloud, it disrupted CFOs, other technologies disrupted COOs, but AI will disrupt the CEO. So I would take it very, very seriously, be it in the boardroom or any governmental building. Thanks, Carlos.

Thank you very much. Very good point. If we stay there, is that education a way of achieving that goal? Maybe we have 2 MIT professors here. Maybe start with you, Hossein. What would your view with it? And what else do you think we need to aggregate it here?

I think classically, if you think about it, AI has always been taught in a computer science or an engineering department, at least until about 8, 9 years ago. And what is changing now is that all schools, most secondary institutions around the world are looking at AI as an interdisciplinary phenomenon. So we now have programs that we are bringing students from design, business, engineering, fashion, performance, to come together to really understand the philosophy and the principles of AI without really warring them about, "Oh, you have to know statistical modeling and calculus before you can use AI." And I think that's important because as [ Danil ] mentioned, I don't think AI will no longer be a CIO or a CDO mandate. It's a CEO mandate, it's a Board mandate. And now if you want to define OKRs, talent will be a #1 priority. How do we bring talent to our organization that can understand AI, biases, generalization, ambiguity, privacy, portability, explainability of data, but also can understand the implication of AI in health care, in retail, in finance, and we do not have any program in the world at the moment, which is mature enough to be able to develop that talent at scale. So some of the work that we do at MIT and in Toronto is to create these AI super courses that are very interdisciplinary, and we build tools, no code tools, easy-to-use tools so that even that journalism student can actually use these tools to build capabilities for the future. And I think that's very well needed. And the last thing I say is, this is not just needed in postsecondary institutions. This is also very important in K-12. Kids should understand the precautions of sharing data. What does it mean? What does it share data to train a model? What does it do to my profile when I grow up? And that's basically what we also need to do in K-12 education.

Thank you very much. Alissa, maybe with you, on cooperation on AI between both companies and government, is that something that is happening? I mean Tech Accord obviously, is a good example as we did on cybersecurity. But cybersecurity maybe is easier to cooperate than AI. I don't know. It's maybe more well defined what the cooperation should be. Maybe your views on that and what you are doing in your cooperation on it?

Sure. So I think there's been a lot of collaboration around AI because I think a lot of companies recognize that these issues are interdisciplinary. They involve the industry, they involve government, they involve civil society. And so trying to figure out what to do and what the right parameters are is actually really important. I think that the tech -- thinking about that on the cybersecurity side, even thinking about what goes into a model, how you make sure that there's security both of the data going in, what happens inside the model, what the output is, all of those things are factors that have to be considered. So having conversations with governments about how to think about those, what is doable and what is not, what is sort of a rational act, I think is -- rational sort of regulation look like? I think it's a very new space. And so unless you have industry involved, it's not going to work. And so you have to have those conversations because government is looking to regulate for good reason, but you want the regulation to be something that is good, actually does what it's supposed to do, put controls in place and -- but also is something that allows AI to develop. So it's a balance.

Absolutely. Thank you very much. Again, sorry, gentlemen, if you don't mind, speak outside, not inside. The room is very small, and we have an echo. So if you don't mind. Thank you. Okay. So in that line, Mark -- David, sorry, David. MIT, you also created your own company. You have been involved in this for so many years. Maybe we can go a bit more into the technology itself on AI because we talk about policy, we talk about ethics, but what is actually the technology? Maybe -- and also for the purpose of the audience, because some people don't necessarily are familiar what is behind AI. Maybe you can give us some insights on that.

Sure. So I'll talk about a few of the different technologies, and I encourage those of you who are trying to learn, you can pick up a copy of my book at the back corner. It's free. So we go through what the different technologies are. So expert systems have been around for a long time. Those are rules-based systems where you code in manually, if then, if I ask, is it daytime, computer will check the time and then say yes or no. Some of the more exciting technologies start to introduce probability. And the one that's got everyone's attention in the last 18 months are large language models and transformers. And what those are doing are -- what we can think of as next word prediction. So if you walk down the street and you see someone you know and you say, hi to them, what do you think they're going to say back to you? Hi. You just did next word prediction. That's what ChatGPT is doing. They have a giant database. They've poured every document they had find into it. And then they're saying, all right, if you have a sentence that begins, the most important thing about AI is its ability to, then it tries to be what's the next word, okay? And that's all it did. It's building sentences, words, paragraphs, et cetera. It's not just one technology, it's actually several technologies. There are a number of different AI systems working in concert. It's almost like an orchestra, in order for ChatGPT to be able to talk to you and help solve problems. What's powerful about it is that it turns out that a lot of the things that we do at work, can be done by a system that's a glorified version of next word prediction. So I spent my summer working with Evercore ISI, the investment bank, and we kind of broke down 800 million jobs into tasks and activities, what people do every day in those jobs. And then we [ mapped ] it against 5 dozen AI technologies. And what we found is that about 1/3 of work can effectively be replaced by AI that exists today, not things that will land tomorrow or 5 or 10 years from now, but today. The most vulnerable professions are economists and software engineers. The -- sorry, Navroop, I know you're a digital economist. And so the least vulnerable professions are ballet dancers. So while I don't suggest that we all become ballet dancers, I do suggest that with no code systems coming, as Hossein was referring to, and basically, the ability for anyone here in this room to simply describe a problem to an AI and it will respond to the answer, to the solution, we need to lean in what we do best as humans. And then we can become better partners for those AIs. So creativity, critical thinking, emotional intelligence, teamwork, these are all things that AI do poorly. If we partner, if we bring that to the AI, we can get the most out of these systems.

Thank you very much, David. Maybe -- apologies for being the last one. Cristina, you are an expert on ESG, and I know that the interrelation between the sustainability, AI, cybersecurity, how you combine all that, how do you equate all that.

So I have a little bit of a contrarian view with respect to AI because I actually think that when you look at every industrial revolution, it's caused a tremendous amount of chaos for people in specific roles, but it's created huge amounts of opportunity. And if you look at this new industrial revolution, there's a lot of forces at play, they're creating a lot of new opportunities. But one area that I'm specifically interested in is in health care, right? So UN sustainability goal is about, what, health care for everyone. What just so happens that there aren't that many people who want to be doctors. So please tell me how you're going to have health care for everyone if nobody wants to be a doctor, right? But if you're able to get data and use a responsible data set and responsible AI to be able to generate tools to help maybe people who are not necessarily [Technical Difficulty] your chances of having great health care on a global basis are much greater. So I think that there are some great opportunities. And I hope that in trying to make AI safer, we don't sort of curtail its abilities and maybe look at what are the data -- what's the data that's being used, make sure that's responsible, understand the models versus basically saying, well, let's just stop AI as a whole because I think you can do a lot of great things. And as we've seen with every industrial revolution, it creates a lot of chaos, but those who understand it and look are going to create great opportunities.

Thank you very much. Maybe just a bit more in the technology sale, Mark, especially on the cybersecurity component of it. Some ideas of it.

Yes. So let's get down and really just talk about what's actually happening today. So the phishing e-mails. Who's received a badly worded fishing e-mail recently? So a few of you. Well, there's a few out there. WormGPT, we probably heard about it. Phishing e-mails almost that first line of defensive users now receiving e-mails that looked rubbish. Sort of almost gone away. So we're already seeing threat actors using now the types of tools that are being talked about the AI tools already. And that is really doing 2 things in the world of cyber. One is it's increasing the pace at which threat actors can do the stuff that they've always done. Social Engineering is another good example. The use of deep fakes, my password is my voice. Those sorts of things are now becoming vulnerable in a way which they were not vulnerable before the advent of some of the application of the AI tools that we see. Reverse engineering code is another good example. However, they are getting a quicker and so their pace is getting better at what they're able to do. The barrier to entry is also coming down. So the reality is today, AI is already being used out there by threat actors effectively. The good news is that what it actually means for organizations is that they have to reassess and if they reassess well what they are doing. It's not just about doing the same things that they have been doing faster. It's about reexamining what really is the approach that they take to cybersecurity. I'm actually quite optimistic because I think that creates a change in the thinking. And of course, when you get beyond the thinking, and at DXC, we have a number of tools that we put in place to help customers really assess that to deal with that, to keep them one step ahead. What that then does it then takes you into a position where you can then actually apply AI defensively because, of course, that advantage exists for us defensively as much as it exists for the attackers. So to give you an example, and I'll quote, Teresa, we're using -- 1 of 6 organizations using Microsoft co-pilot for security, that is reducing some of our productivity times by about 80% plus. First line, second line, the traditional model of going around and doing security is now managing into one. So that's -- that distinction has disappeared almost overnight. So really excited, the use of natural language, questions which we talked about before, means that you don't have to do complicated code strings to be able to identify and then get to that key point to be able to remediate quickly. So I think there's good and bad news, but overall, I'm encouraged, but right down at the [ core ] phase, if you want what's actually going on, Carlos, AI is being used against us by threat actors and it's causing us right now to have to change our thinking quite dramatically and implement AI tools defensively.

Thank you very much. Maybe we can move a bit on the protection of the individual. As you know, WISeKey for 25 years, this is actually our 25th anniversary. When I was in the U.N., after being in the U.N. my focus has been protecting individuals by providing digital identities that they control themselves. This is becoming in some countries, major projects. Also, the United Nation is moving into that. To the point that digital identity might become a human right in the future because everything is going to be around your digital identity, including what are you willing or not to share with AI algorithm. But what are other fundamental rights there, Teresa, you think we need to aggregate to all this?

Well, I mean, what are the other fundamental rights we need to aggregate in thinking about AI development? Well, all of them. No. So I think it starts actually with connectivity, which is kind of where you were saying, this isn't just an AI divide, it's a digital divide. We need to -- and it's -- this is digital. So we need to start thinking about who has even connected because then, one, they can't take advantage of it, but the models aren't taking them into account as their deciding what the next word is in the example earlier. But also, as we're building, there can be really, really, really amazing use cases for these technologies. Voice is a really interesting one. We can use custom neural voice so that people who have ALS can keep their voices once they've lost them to disease. But there are lots and lots of bad uses of somebody with voice. And so I think this is a place, again, where it comes back to what you said about values. The company who builds the technology needs to build in a value-based way and to have that value-based work, aware of the fact that technology can be used for harm. So another very common example is facial recognition technology being used against minority populations in policing, the extent to which we use AI for predictive decision-making, I think, is probably not wise without humans in the loop or intervening. And of course, everyone is worried about missing disinformation, deep fakes and participation in democracy, especially in this year, that is the biggest election year in the history of the world. And so all of that, it's sort of like you have to be able to hold all of those many, many things in your mind as you're developing and have enough steps in your process that you're thinking through that and that you're seeking input from communities who will be impacted by the technology to help build it in ways that respect people's fundamental rights.

Thank you very much. And maybe staying there, the European Union just a few weeks ago, decided in its related to the regulations related to AI. You've been a European company. Is that an obstacle for AI growth for European companies to have that type of regulation? Are we doing the same thing Chinese are doing or regulating AI? Is that maybe the future of everybody moving into regulation? Or maybe your views as SAP. Excuse me, gentlemen,once again, if you don't mind to do not -- thank you. Please. Well, it's like an AI algorithm. I have to repeat the thing every 10 minutes.

Yes. Yes. Yes. First, the final text of the AI Act isn't yet out. So it's hard to tell how it will impact. And the implementation is several years out to -- and one of the greatest fears is by the time it's implemented, what's it regulating. So there's a lot of unknowns even though a lot of the tech and the negotiations for the trilogue are sort of well known. I think as Alissa stated, regulation is coming, and I think it's important for all of us here to get involved in sort of educating those regulators and the lawmakers, the policymakers. I think there's a huge gap, technology always grows faster than sort of policymaking and lawmaking. It's -- that gap is growing. And next year, when we're all sitting here talking about Quantum, that's just going to go straight up. So I think industry and society needs to get more involved. This has been said already by a number of panelists. It's an all-in thing. Industry needs to talk to academia, it needs to talk to policymakers and it needs to go around in a circle of if we're going to have smart regulation so that the technology can build trust and take off.

Maybe a point on Quantum, is Quantum going to accelerate AI to AGI? Is that something that you see happen?

Yes. Yes. Yes.

I don't know -- there's guys...any other reaction to that?

Yes. Yes. So I was riding on funicular with Steve Suarez, who used to be with HSBC, and he said, Quantum will give us faster garbage. So -- but think of it's garbage in, garbage out. So if we don't understand the principles that Hossein is trying to instill through his AI super courses that we give bad instructions of these things. They're going to give us mindless realms of garbage, the clutter that clouds our inbox is going to get worse. So Quantum couldn't be very exciting to solve a lot of problems, but also could create a mass headache for you.

Actually, one of the applications we see happening in Quantum, which is applied to CLSQ is post-quantum capabilities at the microchip level. So we are programming the chips to be able to defend a quantum attack, without even knowing what that quantum attack will be. So it's very theoretically, but we are working on that. And I think this is maybe something very concrete. Any other reaction on the Quantum aspect?

I think as David mentioned, I think at the moment, they're thinking around generative AIs that we add more parameters and then we're going to have better GPUs and TPUs and then quantum. I am generally optimistic about where AI is going, but what makes me really worried is that very few people talk about the data around this AI. Some of this data belongs to us. Some of it belongs to corporations, some of it are public. If we really don't understand the data ecology and the contracts and the privacy of this data, yes, we can compute faster, but can I actually be comfortable about building a GPT like interface to talk to the metadata of my personal photos and say, show me a data that I was very happy walking in nature with my mom. AI can do that from a technology point of view, how many of you trust foundational closed models with that data. Very few of you, should not trust that. What company trust building predictive service saying, hey, show me where I'm going to run out of money. You're not going to expose that. But would you want to have that on an encrypted chipset that you can only access it through a token that is where truly quantum and that will come together to show the next generation of AI.

Very good. And maybe back to China, the fact that maybe I'm wrong, the accessibility of data is easier in China will accelerate AI from your point of view towards other countries or you will have the same problems that we have?

I think China is as we have some different culture from the Western world. So we respect the elders, we expect to the authorities. So we put some, how to say that. So in some kind of filter for the data coming in and the data comes out. So this is what in China we mostly focus on. We do not want to AI to speak anything which will make a government on hedge. So this is -- China got, I think it's a special for China. And another thing is that in China, we also share the same problem for [indiscernible]. So like you know we have invented clone technology, we can clone real human, not before. But no country allow you to clone the real human. So I think it's something we can share with all the world. So there is some -- should be some forbidding area for AI, not to touch. So I'm very curious why the military AI usage will be permitted by open AI. For such a company, why it should support the military usage. This is the same to the nuclear war. So if the AI says -- China government says, okay, then we start to develop military AI. So you will see what will happen. So I think for all the people in Davos, we should make a rig line for AI. Some area to that project, just like clone real human. That's my point.

That's a very good advice, especially coming from China. I think this is something that everybody should pay attention to it. But how you do it? Is that a proliferation treaty like we have in nuclear weapons, is there a way to do it? How do you do it?

In China, we have a lot of government officials. They were doing very -- a lot of details. So they make a lot of regulations, and we have real resources to execute it. So if you want to do commercial AI in China, like you needed to get a license. That commercial license is -- without that, you are not -- you are illegal. You cannot provide there. But maybe this will have some problem for the China's AI development. So -- but I think it's in a specific area like a medicine, you don't want to illegal doctor to serve you. You know that.

Maybe we can discuss on the technical point of view, this comment of AGI. Are we really on an AGI or this is a bluff? Any comments for any of you on AGI, artificial general intelligence, when one AI can create another AI. Any you want to take it? Yes, Mark, maybe we can have different opinions on that.

A couple of comments before I answer the AGI. I think in 2024, which we're in, doesn't this microphone needs AGI. So we are experiencing now the age of algorithmic economy, and we're experiencing in 2024 active use of artificial intelligence agents in every part of our work in every part of our life. Some of them are sophisticated. Some of them are predictive in terms of amount of data that they have gathered. But in this type of economy globally -- and also, by the way, we're in a very fragmented world. You could see huge cultivation of countries around the world that are gathering in terms of this bricks momentum. It's been discussed here as well. So with this in mind, there's a whole new approach saying, okay, can we leverage some of the sophisticated cognitive agents that are being used in everyday life, for example, travel experiences, hospitals dealing with cancer patients, can we develop some sort of an AGI. I think personally, we're years away and Generative AI is not AGI. It is not AGI. Many people mentioned, okay, it's predictive, it's predicting the next word, the next sentence, but it doesn't have reasoning and it doesn't have logic that you need to combine deep neural networks and generative AI to be able to do this. So [ I feel we are not there ] yet, but we're in the age of absolutely impressive cognitive agents this time. It's beyond Alexa. It's beyond SIRI. We're going to see something that every part of our life, every part of our work. And this is why the direction was going into autonomous enterprises. We are entering the age of autonomous enterprises, we're going to have intelligence built in, in many of those companies. So I think there is a process -- there is progress in some of those things. It's not full AGI, but we got to be careful. If we're moving in a direction where companies are going to be autonomous, what is the role of everybody in those companies. But -- so it's an interesting world right now, a world moving into this whole age of cognitive agents, and we just have to be mindful that we have some way to protect ourselves because every part of our life and aspect is being basically implemented by those cognitive agents.

Thank you very much, Mark. Any other comments on that? If not, we're going to open the -- sorry, you have -- you wanted to say something? Yes, please. you too? sorry. I didn't see you.

I want to say I just wanted to quickly add. I mean it's fascinating, I think, sort of the futuristic type of future casting. But as an economist, who doesn't spend most of the time doing predictions, I think the reason why we're seeing so much industry focus on AI right now, it's because of the economic value, it's going to add to the global economy. It's really simple. I mean the model of open AI, ChatGPT is a good old subscription. So it's -- there's nothing groundbreaking happening there. So I think it's also just important to remember like here, of course, we have visionaries and thinkers, but the sort of capital that you're seeing moving in is the economic utility that it provides, the fact that this technology is going to transform the global economy, how we do work, the future work and the future of business, how we relate productivity, all of those things, first and foremost, because that's the first thing big corporations are dealing with. And I think some of the sort of the more fantastical use cases come after the fact. So I just want to kind of perhaps ground that a little bit, so [ $0.02 ] there.

Thank you very much. Do you have a comment?

Yes. So in is a very fundamental thing. We can think about is how we should make the difference between the AI and the real human start from here. So it's as early we can put some kind of recognition. We know the open AI, they have issues the [indiscernible]. So they use [indiscernible] to recognize the real human. But I think that we need more -- like I know the [indiscernible], you are developing a lot of DID system. For the DIDs right now, it's the most important thing. Recall all the behavior of real human and anonymous. I think it's very important to differentiate the AI from the real people as early as possible so in the future, we can create real human society, another 4 of the whole kind of AI bot.

Yes. Actually, there is a test in ChatGPT, if you go to ChatGPT and you say, do you know who I am. They will ask you all kind of things without telling you who you are, which actually goes back to the Internet. There was this dog in a computer saying and the other side, the Internet as now you're a dog. So we are still unable to identify a dog in a human. And if you ask the same question about any other AI algorithms, you will see the same answer. The AI doesn't know yet what the human is, so which is a scary. Any other -- so we have some questions from the audience. So you have a microphone there, please? If you don't mind to introduce yourself, that will be great.

Thank you to this distinguished panel. My name is Jamar Montgomery. Hello, Teresa. We -- AI weapons are already here. Let's be honest, we have already been utilizing algorithms for our weapon systems and to think that AI is not being utilized in warfare does is all the service. So if we're going to have AI being utilized for good, my question is to industry, it has to start with children. If we are not educating our children on AI, then the most usage that we're going to get from AI is through warfare. Why does open AI have partnerships with the United States government because the United States government and the United States military has the largest budget, profit-making enterprises. So my question to industry is how are you going to make sure that our children are properly educated so that they can be the new creators of AI and the new tools that we're going to use for society.

So I think it actually goes back to this multidisciplinary approach in some ways, which is -- we actually need the humanities in the development of the technology to keep ethics front and center, history, literature, philosophy, maybe religion, I don't know. But I think we need to -- I think there's a real value in bringing humanities into the development of technology, and we actually kind of do educate our kids on that. I have 15- and 17-year-old daughters. So we spend a lot of time talking about that. I do think we need to think about computer science or whatever the next version of technology education for sure. But my sense is you're not going to need as much of that in the future. I'm not sure if I'm right. I should turn to one of the professors who thinks about education.

No, I'm not a professor. I'll get into the reactors, but I'll just say real quick, I think it goes back to an all-in approach. And a little bit of what you said and probably what he's going to say. But keep in mind, we have -- and I'm sure every corporate here has the same thing. We have a CSR program, we have education programs. Even at our highest scale, we educate maybe 1 million people worldwide over a year or 2-year period, something like that. There are something like 5 billion children in the world. So that's -- we're a drop in the ocean. And if you add the [ $10 ] million that Microsoft does, we're still just a drop in the ocean. So it's got to be an all-in approach, but I'm interested to hear the question to...

I mentioned some of it. I mean -- but I think AI education and data education and STEM education in K-12 from K all the way to the North American system and equivalent in Europe, I think it's really important. I think the physics of data and AI is as such that it will completely gray out disciplines and how we know industries. If you look at industries, you have the finance sector, you have the insurance sector, you have the health care sector. Data is as such that all of these are going to come together. Do not be surprised if in 5 years, you're going to see an athletic wear company that is also in the insurance space because they can give you dynamic policies. Now, if you think about a university or a school structure, architecturally, they are terrible, you're full of walls. You're like, oh, you want to be in humanities, you want to take your math course, you want to take your music course. You go to that room and you go to that room. I think if we want to bring that education to kids, this is what it means to your data. This is how you should share data. This is how trains are being modeled. An example of that is a project that came out of MIT called scratch that kids are now very interested to learn coding without really thinking about how compilers work or how JVMs work. We should do the same thing for AI, but we should not just think about the curriculum, we should think about architecture, we should think of productivity, interconnection of students to the workforce from kindergarten all the way to post second reeducation, and that may address some of the correct observations that you shared with us.

I do want to add to it because I think that brings up the point on the other side, too. I think there's a reality of what the next generation will look like. I think AI will get incorporated into their education from the beginning, and they'll have to learn how to deal with it in a lot of different ways. So I think practically making sure that we have education systems that know that they can't ignore AI. They can't ban AI. They actually have to figure out how they make sure kids know how to work with it and a lot of different disciplines. That's the other side of that same equation. So it's not just making sure that they learn some of the underlying technology, but that they understand what it then means for whatever discipline they might be in, what they can do with it, what they can't. And you can have really practical examples, right? So going into the workforce for the first time, is to write an e-mail through ChatGPT and doesn't look at, it might lose their job when they write an e-mail that doesn't -- isn't actually what was -- is appropriate, right? So trying to make sure that's a component of education on both sides of that, I think, is important.

I'm just going to say very quickly. Let's not forget David said that it's the data itself in the models. That is we haven't -- no one really talks about it that much. Is there an application perhaps for block chain to actually be much more about a real use case of block chain, actually looking at the veracity of data, and do we know what's in those models and how can we track it and create that almost maybe that is a ready-made use case of block chain that we haven't really thought about yet. That's, I think, we've got to think a lot more about the data and how we manage the data. As David said, rubbish in, rubbish out. We've seen that for many years. And the bias as well in -- that we've talked a bit about today, let's not forget, machine learning, there's been bias around for years. I think about mortgage applications and stuff. We've seen that. So we can draw on some of the experience we have. But let's not forget the data is absolutely key. And how do we really make sure that we focus our attention on getting the right data and ensuring that it's not manipulated for nefarious purposes.

Just very quickly on that point, I think, in terms of education, our education in the United States is very generic. It is not focused on individual students. And with AI, you would now have a chance to actually individualize hyper-personalized education. So I think we need to focus education on individuals, individual lessons, not generic, pretty much most of the at least K to 12 grade generic education. It doesn't do people any good. They're not going to be competitive in this digital AI age. We have to hyper-personalize to every, every student to understand their vulnerabilities, their abilities, their capabilities, and this is our future. If we do this, we have great success with AI and capabilities are only going to grow over the years.

Thank you very much. And yes, I think maybe our kids or future generation will don't even call it AI anymore because we'll be totally integrated in everything they do. So I would like to -- before we close this panel, by the way, I would like to thank the panel, it's amazing discussion, we're going to here 4 hours. We have 2 other panels, then we'll start in a few minutes. We have -- talking of appreciation before you leave, which is going to be introduced to you by David Fergusson.

Thank you, Carlos. I've enjoyed the opportunity this week to host several sessions with CNBC on what subject you think, most likely? AI, exactly. Exactly. And there was a common thread running through all of those conversations. Of course, we talked about education, reskilling, upskilling. We talked about the future of work. We certainly talked a tremendous amount of the applications across all industry sectors, had a fascinating dialogue earlier this morning on the applications for climate change and sustainability management. But the common threat for me throughout all of that and which was certainly reinforced here is the power of collaboration and the need for collaboration. The ease with which we do it, I think, is still to be defined. But most importantly, this is what I'm hearing from all of you. I'm very honored. And WangWei, will you come and join me?

Okay.

I'd like to introduce to you a very special guest. I'm an investment banker by day. And WangWei and I have had the opportunity of collaborating together between the United States, China and across Asia for almost 20 years now. Five years ago, WangWei arrived here at Davos. I should probably note as well that WangWei is considered the father of modern M&A in China. So WangWei brought the Western style of mergers and acquisitions and foreign direct investment to the country and has been a tremendous philanthropic leader throughout all of that time. Together with WISeKey, we have been thinking about how can we possibly recognize and, of course, selfishly encourage more innovation from those of you who have already proven the desire, willingness and capability of having an impact. And so we have created the Davos Innovation Award. This will be the inaugural presentation of this award, and we'd like to begin by presenting this award to each of you this evening for being here.

Just a few seconds. You can find the medal on the back with some Chinese characters. So you can interpret other way, you feel good. So like an entrepreneur or fortune, whatever. So this is a good sign for all of you. Thank you.

Great. Thank you for all you've done and all yet to do.

Thank you, everybody. Just a big hands to the panel, please. I think they did a great job and it's a great panel. And you have not now in the collection to remember that important day on where we all agree and we have to work together to where AI and health of future generation. Thank you very much for your attention. It's always a pleasure to have you here and hope to see you again next year. Thank you very much. And thanks for the audience, very interactive audience. It's impossible to give you a target.