What if generative AI is like electricity? And that the chatbots that we're talking a lot about these days are merely the light bulbs. Don't get me wrong, light bulbs made a lot of things better. Made it easy to read at night, it made us able to operate with better precision, and it made it less scary to go to sleep. But are we losing sight of the bigger picture here? Because if AI is actually more like electricity, it's not only a tool that will help us do what we're already doing in a better way. It's not just the light bulb, it's not just the chatbot. Maybe it's a technology that can fundamentally transform what and how we work as organizations, as societies. And if that's true, are we having the right conversations about it? Are we being ambitious about it? Are we being ambitious enough with what we want to achieve? Are we being mindful and intentional enough with the problems we're trying to solve with it? And we want to set the direction and stare toward meaningful and sustainable development, right? The technology doesn't have a plan or an agenda. Not yet, at least. And it's up to us to harvest and direct its potential in a direction that's good. And while this is in no way an attack on wanting to pick the low-hanging fruits of optimization off of the AI tree, or keeping a cautious eye on the rules and regulations we need to have in place to ensure that everything is safe, we will be spending the next hour more in the spirit of widening our horizons, on taking a step back, and daring to imagine what future scenarios could be made possible through this technology. So, if AI isn't a light bulb, it's more like electricity. Then what is the electric engine of AI? Who's building it? And what do we want it to be steering towards? Good afternoon. My name is Gunvor. I'm one of your hosts for the next hour, where we explore the futures of AI, together here from our implement offices in Copenhagen. While we've been preparing this event, we kept coming back to the comparison of this being a little like Alice in Wonderland. And for those of you who might not know or remember that story particularly well, short version is, it's about a young girl called Alice, who's very curious by nature. And when she sees a white rabbit with a waistcoat, she falls it down its rabbit hole. From here, she tumbles into a fantastical world, where none of the rules are. And what's the rules that she knows that she knows that she knows from her Victorian everyday apply. There's a lot of magic. There's a lot of wonder. But there's also a lot of tension and chaos there. People driven by greed and frustration. So, throughout the lines of her adventures in Wonderland, she meets characters that challenge her, help her along the way. And towards the end of the story, she sort of grows into her own. She grows confident and courageous and standing up for what she believes in and speaking out. It's a great book. It's a very weird book. It's a pretty old book. It's from the 1860s. But the reason that we keep coming back to it is because it actually reminds us a lot about the experience that we have trying to navigate this space of generative AI. Like Alice, we're constantly trying to think on our feet, picking which rabbits to fall in. Trying to follow, trying to figure out which experts to actually listen to. Because the experts don't really look like they used to do. They might not be smoking caterpillars or grinning cats, but still, they're not the people we're used to looking towards. The rules keep changing all around us, and we need to keep adapting and navigating to that. Our imaginations are constantly being stretched and challenged by what is theoretically possible. We're also being inspired by weird experiments going on everywhere. Some things are amazing. Some things are ridiculous. Some things are ridiculous. It's quite difficult to see which is which sometimes. And last but not least, we're continuously reminded and reminding each other that we need to balance our sense of adventure and our curiosity with standing firm in what is actually important in our organizations and what we're trying to achieve. Luckily, just like in Alice's Wonderland, there are people who have already fully emerged themselves and spent a lot of their time in the world. in the space of generative AI. And they can help challenge and guide us as we learn to navigate it and try to figure out what our roles in it could be. One of those people is my good colleague and co-host, Adam. Welcome, Adam. Happy to help. As you've seen from the agenda, over the span of the next hour, we'll hear from a bunch of really, really interesting people. Me and Gunnar will be here to help you try to summarize and make sense of all the different things. It is going to get a little complicated because it is a complicated topic, but we'll do our best to guide you. We will. And keep in mind, none of the segments you'll see today will be longer than eight minutes. Most of them are more like five. So, if you get lost along the way, don't worry. Conversation will reboot in just a minute. And again, this being the internet, of course, there's a chat. So, please use it. If you have any questions or ideas or comments, just shoot them away. We'll be in there to monitor and help. Yeah. And last thing to stay aware of, there won't be a formal break here during this hour, but by the end of the session, you'll get a link with a recording so that you can re -watch anything you want or share it with anyone else. So, with all of that out of the way, let's hear from Benedict Evans, an independent researcher that I had the great unfortunate chance to speak to for a long time. Benedict is, to my mind, one of the sharpest minds in figuring out how AI and business merges together. So, we had a really interesting conversation that we would like to share with you. When you look at generative AI's impact on the business world, when it's become this sense of gravity right now, what are some of the big impacts you're seeing in different industries at this point? So, I think it's important to remember that none of this worked two years ago. And it takes a while to build enterprise software, and it takes a while to work out how to change workflows or build things inside big companies. And so, we shouldn't expect to see the entire world changing right now. It takes a while. And secondly, I think this is a really profoundly important and interesting technology, but it kind of takes a while and a little bit of thought to work out quite what to do with it, quite how to use it, just as would have been the case with databases or SQL or the cloud or any other major technology shift. So, there's a couple of industries where, or a couple of kind of fields where this is very obviously immediately useful without much thought. And so, people doing software development are already seeing sort of 20 to 30% efficiency gains. People in marketing and advertising are adopting this very quickly because it's very immediately obviously useful without much changing. And then there's a lot of interest in customer support. And then beyond that, lots of sort of early adopters, lots of pilots, lots of experiments. Everybody sort of picking it up and thinking, well, what's the right way of using this? I want to catch on to something you mentioned because I think that's a fascinating part of the way that we work here and our role as consultants in the industry. I'm a technology nerd and I work with technology nerds and we know a lot about what these LLMs and generative AIs can actually do. But we don't have the imagination to see many of the business problems. Yeah, I mean, you know, technologies get diffused in different ways in different places. So, the incumbents typically always try and make the new thing a feature. And often it is. So, we know mobile search is just search for Google. There's not like a whole different search engine that's just for mobile. And today you can see Microsoft and Google kind of spraying LLM features all over Word and Docs and Gmail and so on. And there won't be like a new spreadsheet but that has an LLM included. It's just a feature. On the other hand, sometimes you can pull the whole thing inside out and solve the problem in a completely new way. And sometimes as a company outside of the tech industry, sometimes this new thing is just something that you can start using. I mean, one of the striking things about ChatGPT is that it went from nothing to having, I think, we just heard that they have about 350 million monthly users. And you look at survey data that says, you know, something between sort of two-fifths and three-fifths of the population in most developed countries have certainly heard of this and something like two-thirds to a half have tried it. And that's a sort of unprecedented adoption of a new technology. To me, I think the fascinating bit now is to figure out to where does it fit and where does it not fit. And I think we seem to be aligned on figuring out where does that line go and how far along is the technology in being able to help different areas of business? Yeah. I mean, I think one of the ways I look at this is to talk about spreadsheets. So, in the late 1970s, a guy called Dan Bricklin is having an MBA class. And they're drawing a spreadsheet on the board in chalk. And he thinks, hey, we could do that in software. And so he makes a spreadsheet in software on Apple II. And he shows this to accountants. And they look at it and they think, wait a minute, this thing can do about two weeks of work in about 10 minutes. I can change the interest rate here. And all the numbers change all the way across the columns for the next five years. And today you look at that and you think, yes. But if you had been filling in those 500 cells one by one with a pencil and paper and maybe a calculator, seeing that happen on a computer screen changed your entire life. And you could do a month's work in a day. And he has a lot of stories of people who are literally doing like a month's work in a day. And so if you have that use case, this is amazing. If you were a lawyer and you looked at that, you would say, well, that looks very clever and my accountant should see this. But I don't calculate five-year P&L forecasts. That's not what lawyers do. I do something else. And I might use this for maybe doing my invoicing, perhaps. Or maybe my bookkeeper will. But that's not what I do all day. If we were speaking to business leaders, let's imagine there are business leaders listening to this at this point. They're keen on generative AI. They sort of believe in the technology, but they might not be in any of the immediately obvious industries. Do you have any ideas about what could you do as a business leader if you want to foster a culture of using this technology? Well, obviously what you need to do is you need to hire some strategy consultants who can help you build a plan for how you would do this. Look, one of the ways I think about this is that all generative AI questions have one of two answers. The answer is either no one knows. Like, how big will the models get? Will they keep scaling? What will happen to hallucination rates? Will we run out of data? Do we get model collapse? What happens to model scaling? All these sort of basic science questions. Do agents work? Et cetera, et cetera. And then the other kind of question is, how did every other platform shift work? So how did you do mobile? How did you do cloud? How did you do SaaS? How did you think about moving to the web? How did you think about buying PCs if you're a company old enough to remember doing that? How did you deploy mainframes? How did mainframes change your business as a supermarket, as a bank, as an insurance company, as a shipping company? What did it mean when you had SQL and now you could do arbitrary queries on your data? Well, okay, that sounds great, but what kind of arbitrary queries? What would we ask? Like, you're a supermarket. It's 1978. What do we do with our database now that we can ask questions really easily? Long pause. Well, that sounds good. Let's go and think about that for a while. I'm really interested to hear your perspective on the companies who are successfully adapting these technologies. Again, if you are a company who wants to get into this, what do you expect to be the hallmarks of companies who are successful in adopting AI and generative AI technologies? Yeah, I mean, you know, kind of repeat an observation I made earlier that none of this stuff worked at all two years ago. And, you know, the typical enterprise sales cycle is, software sales cycle is like 18 months. So it takes a while to adopt. And right now there's very little, like, entrepreneurs haven't had much time to use ChatGPT to solve some fundamental problem. You know, we're coming up to all the developer conferences and all the big events, the second half of the last quarter of this year. And, you know, we're going to have hundreds and hundreds of new companies and new announcements of new products. And Salesforce have got their story. And Oracle will have their story. Everyone will have their story. And so there will be this wave of companies saying to big, of software vendors saying to big companies, like, this is how this is going to help you. This is our story. This is what we're saying to our, you know, obviously some of that's about them boosting their share price and reassuring everybody. We've got an AI story too. We've got AI too. And so you'll have, you know, in a year, you'll have probably an order of magnitude more people proposing to you, this is how you can use generative AI to solve this. Has this big company deployed generative AI? It's kind of like, you know, it's like being in 2008 and saying, have they deployed smartphones? Well, the CEO's got an iPhone. What do you mean, have deployed? And everybody has now done a bunch of pilots. And everybody has probably put something into production somewhere. So the software development team is almost certainly using Copilot. And if they're not, why not? That's a problem. Marketing team is using Midjourney and ChatGPT. And then it kind of percolates out through the rest of the company. And you have to think about how you enable innovation, while also thinking about how you understand what this is. And I totally love that. And I think it's one of the biggest privileges that we have in our time right now to be able to sit down and work on that innovation and to see the pace of change that is actually happening and sit down and imagine the possibilities there could be for companies for solving their own problems or engaging with vendors or sometimes help companies design new products and features to do these different kind of things. I think that will be it for the first segment. Okay. Thank you. So that was a really exciting conversation with Benedict Evans. Was there anything that stood out to you, Gunmar? A lot of things stood out to me. I think one of the things that confused me a little bit was the concept of time that kept coming back. And then I felt sometimes you were talking about it as if we were moving too slow and we need to get with the program. And then other times it sounded more like you were talking about us maybe jumping the gun or forcing something that we know takes time. I think there is multiple things happening there. And I think for one, we touched a bit upon it, that the technology is moving super fast. So none of this existed two years ago. But at the same time, there's a lot of typical business and societal processes that works at the pace they do. For instance, Benedict mentioned the business-to-business sales cycle that is typically 18 months. And I think that's what really shifts some of these things. So the companies who are able to work with us right now likely were already thinking about the technology even before it really arrived onto the scene. And that's why we'll see a lot more interesting things happening in the very near future because people just got it into their hands. And the people who didn't think about it until it was obvious are just thinking about it now. So I think there are mixed timelines that gives us this sort of different waves that will come throughout the things. And I think there are mixed timelines in different ways, right? Because there's a timeline that has to do with how quickly did you get on board? How ready and able are you to adapt to the things? And then there's also just the distinction. Are you looking into innovations where you can exploit something that's already proven? Or are you out there exploring how could this fundamentally change your business model or the way that you're working? And I could imagine that also from a leader point of view, that's something that you need to set different paces for. Exactly. And I think some of that exploitation is getting more and more obvious. I think he mentioned something like GitHub Copilot and that's something we've seen as well that just slots into existing processes for code generation and software developers really, really neatly. So that's pretty straightforward and quickly becoming something you just kind of have to do. But the top line innovations in new products and services, that's really about enabling innovation in organizations in a different way than we've seen before. And I think that's pretty appropriate. So that's pretty profound. Agreed. And it's also, I think, where a lot of us have to figure out how do we even rig ourselves for that, right? How will we be ready to drive that kind of innovation knowing that you have no idea where it's going to end up, right? You don't know where the technology is ending up and you're not quite sure how it's going to influence your business. But great conversation. And we'll come back to you and Benedict a little bit later. But now I'll hand the mic off to you so that you can tell us a little bit more about how this works from your point of view. So welcome to AI Wonderland. While I call myself an AI expert, that is a humbling term in these days. This technology and this world is changing super rapidly and reasonable experts often disagree on where things are going and what potentials we can expect. But I think it's important. But I think it's important that as business leaders and entrepreneurs out in the world, that we're able to make decisions and informed decisions based on the best available knowledge. So what I'll share with you today is my best take on what we know today. And that can change quickly. So don't put too much into my words. The most important new rules we have in this world are to understand how processes will be changing. I think it's important and helpful to think that a lot of the changes we see today are about assisting human processes. So we see people adapting tools into their workflows. We see co-pilots. We see people use off-the-shelf, chat GPT-like tools to copy-paste information back and forth. They're fundamentally doing the same thing they always have done. But now they're doing it faster and more intelligent using AI. That's a common first step when we see these kind of innovations. What we're starting to see in some places, and what we'll see going forward in more and more places, are when we start to augment processes. We start to change processes because as the AI enables certain things to move faster, we start to change the process to fit the technology. The common saying goes that initially you fit the technology into the process, and the second step, you fit the process to the technology. You have to adapt the organization to what it looks like. And then finally, I think the third stage that we'll start to see, and we are seeing it in the most straightforward and easy places, but it's a lot more rare, is true automation, where the AI are using system integrations and higher levels of intelligence to actually go in and fully replace tasks that humans previously had, and allowing humans to start focusing on entirely different things in the organization. These are some of the new rules of how the world are changing. Here in the wonderland of AI, we will first look at process assistance, then process augmentation, and finally, full process automation. It's not all that AI will do, but it's a helpful way to start thinking about the opportunities that AI present to businesses, and what that enables when you can automate things at scale, or what kind of changes you will actually do to processes, and what that will enable you to do. When you venture into wonderland, it's also important to bring some tools along. What are the way to think about what are the way to do to do with what are going to work in this world. The traditional toolbox of what you used to do business might have changed quite a bit. I'm about to say that I bring you the starter kit of AI. We advise most businesses to lean into three core things, and we sort of cheekily called it the three T's of AI. So, number one, we recommend you to present the tools of AI to all employees. Fundamentally, if your employees don't have access to a powerful frontier large language model, you can think of it like a basic clone or similar things to chat GPT, where they can start working with the technology in their own processes, you really need to do that. This is a really democratic technology. It's really easy to approach, but you need to allow people in the organization to actually use the technology. So, number one recommendation for the starter kit is to actually make the tools and technology available to employees. The second thing is about training, because the technology is weird, and while it is democratic, and while it is democratic, and while it is democratic, and while it is to some degree, it is intuitive, you quickly get into the weeds, and most people who have played around with the technology also find that there are strange limitations, and one moment it works, and the second it doesn't. What are hallucinations? How do we deal with biases? What are the security ramifications of these things? We need to be able to work with these technologies in an intelligent way, and that fundamentally requires training. And training can be very different for different groups of people. But that's the second important point that we want people to work with, is to give the tools available, and then give people the basic training, to be able to work with them, so we can start imagining what's actually possible with this. And the final T is a little task force. It can differ a lot, depending on your industry, how impacted you are, your ambition, and overall organizational size. But it's really important to have some people sit down and think about what AI means for your business in a structured way. And here we really want you guys to go beyond what are sort of the immediate business pains that you could focus on, but actually ask, like, what role do we play towards our customers? What kind of things does this enable us to do? How do we become a better partner to our suppliers? What are our competitors doing? But sit down and have that structured moment of thought about what does this actually mean for us? You might have to revisit that as the rules change, but just doing everyday business is pretty risky at this point. The final thing that I think is really important to consider is the role of risk in all of this. Because we meet a lot of organizations who say, this is cool, this is innovation, but we need it to be risk-free. We might be willing to invest a little bit in it, but because it is this new and fancy thing, and we're doing it on the side, but we have a core business to run at the same time, it really needs to not interfere too much with things. And that's okay. Maybe your business isn't hugely affected, and maybe you have to take a zero-risk approach. It's just a fair warning from our perspective that you won't get really far with that approach. Because AI is, at its fundamental nature, a new thing, and it enables and it forces organizations to do and think differently. Your processes will change. Your ways of working will change. And potentially, your relationship to competitors and customers will change. And that carries inherent risk. And if you aren't able to accept or willing to embrace that, then you will miss out on a lot of these things. And eventually, that becomes a risk in and of itself. The world is changing, and this is hugely disruptive. And disruption is risky, and disruption can be frustrating and annoying. But leaning in and accepting those kind of risks is really important. But it also means that senior leadership who carries that kind of risk needs to go into the organization, needs to go in and pursue and accept and give people comfort that they actually not just are willing to accept those risks, but they encourage people to think outside the box. Of course, there are ways to mitigate the risks as well, boxing experiments in, doing things in a controlled way. But fundamentally, it's really important that we go out and try these things and have the courage to really dare to imagine what could be possible with the technology. Go out in your organization, talk to each other, get inspired, try the technology. If you see something really cool where the AI model did something that you might not have thought was possible, share it with somebody. Share that excitement and share that moment and start talking about, if this was possible, what else might actually be possible. If this was possible, what else might actually be possible? That's really the most important and where we really fuel a lot of the innovation. So let's hear from somebody who's been in Wonderland longer than most. FTFA is a Danish unemployment insurer who had the profound experience of a motivated employee bringing a really interesting case for them when he showed to them that generative AI was able to write job applications. And their CEO, Casper Kyd, really leaned in and saw the opportunity and had the courage and imagination to see what JobPartner could potentially be, the solution they finally made using the technology. Today, JobPartner has written over 200,000 job applications for FTFA's members and has really profoundly changed and reshaped the way Casper views their company and their role in the market. And I think that's why Casper is so interesting to talk to, because he sort of lives a little bit in the future compared to the rest of us, thanks to his experience with AI. So let's bring in Casper from FTFA to tell us more about JobPartner and how it changed his company. The reason why we dared to imagine JobPartner was because we thought that it could create a competitive advantage for us. So if we were the first to go to market with a service like this, then we could be more attracted to future potential customers. Well, basically, JobPartner is making it easy to apply for your next job. So what we do is that we take your resume, and then we take the job offering, and then we make an application for you. The idea for JobPartners was actually not ours. We were handled by one of what we cooperated with, and he came with the idea because his mother became unemployed. And she said, well, I have had the same job for 25 years. I don't know how to be a job seeker. And this guy, he was very much into AI in the early stages with ChatGPT3 at that time. And he said to his mother, well, I can make an application for you. And this he showed us. And what we realized was that his mother's problem was an ordinary problem for the 10,000 of people who are unemployed each year. So we thought, OK, maybe he couldn't make a business out of this. But this is definitely within our core value proposition to make it easy to get the next job. So then we said, let's do it. And we knew, because we have a lot of skilled people, that we could make it. So it was pretty easy to say, OK, we can make it. So what was more difficult was to discuss with ourselves, should we make it? I actually had a longer discussion with one of my colleagues. And I said, well, we need somebody else to make this for us. And he said, no, it will be a competitive advantage if we make it ourselves. Then we can sell it and we can profit from it. And I said, no, no, because if we make it ourselves, then we are moving beyond being an insurance fund. Then we are moving to becoming an IT company. Took me actually some time to realize that it was not me who was right. It was my deputy who was right. So I took the role more to do what he had said we should do. You have to look upon job partner not as being something about D&T AI. It's also something about D&T AI. But what you really want to do as a CEO is to look upon your business and say, OK, can you transform your business model? And then D&T AI comes in as a solution to my problems. But really, you have to look at your customer needs and then find your solutions to your customer needs first and foremost. And then at some point, D&T AI can be the solutions. It is for us, with job partner, but for others, it's other things which are the solutions for the customer needs. So a business should be driven on satisfying customer needs. What is really important is that job partner was not my idea. It was not the idea of any C-level people in my business. It was just the idea of a normal, good, hardworking employee. And he was not afraid to say, well, I have a good idea. And then we listened. So it's really a matter about getting all your people to think about the business you're in and getting ideas forward and then discussing. And between those hundreds and thousands of ideas, maybe one of them will have the potential to redefine the industry you're in. I think that D&T AI is the one thing which will have the most impact on businesses that we have seen for at least a generation. So anybody who does not dare to log into this is really taking a risk with their companies, both short-term and definitely long-term. success. I think that a lot of people are held back because they don't really grasp what D&T AI is all about. They only see like the surface, which is trying to prompt something in chat-GBT. And they say, oh, it doesn't give me the right answer. It's hallucinating. And then they just don't get it. I've seen a lot of benefit for us being this transformation from one company to another. And that is clearly that when we see the world with an IT company's glasses, we can imagine some things which we couldn't do before. So it makes a lot of other things possible. For 150 years in Denmark, when you had to apply for a job, it was a newspaper. Then about 30 years ago, you went from the newspaper to a job portal on the internet. But still you have to apply for it. Now we have disrupted that because now everybody can make a good application. So we are in the middle of a transformation. And in a couple of years, I believe that nobody would apply for a job anymore. Our disruption is just really destroying how the job market has been for a lot of years. And I think actually this is a good thing. Because why should it be a good measure of how you'll be performing in the next job that you're good at applying for it? Why? Thank you very much to Kasper from FTFA for sharing his story about how they've worked with the job partner solution to spark their AI transformation. Now we're heading over to a conversation between Adam and Mette Muldvi from the Novo Nordisk Foundation. They're going to talk about how the Novo Nordisk Foundation have embarked on their organizational transformation. Thank you so much for taking the time and be with us today. We'll have a conversation about how AI changed your organization, how you work with it. Initially, could you tell me a little bit about how did you begin working with AI at the Novo Nordisk Foundation and what has sort of prompted you guys to begin thinking along these lines? I think as any other organization, we were quite surprised with the release of ChatGPT. And I would say our CEO was very interested in this technology. I think that's the primary reason of why we acted so quickly on this. We started out with the Novo Nordisk Foundation and we started out with the Novo Nordisk Foundation and we started out with prohibiting any use of ChatGPT for work-related matters. But quite quickly, we faced that we needed to get this technology in the hands of our employees to actually start exploring how this could provide value to the organization. I think at the same time, we faced some exciting challenges. Our previous growth, of course, but primarily our future growth, internationalization, many drivers for looking into us as an organization So, in the future, how can we do, how can we do this technology in the future? How can we do this technology? How can we do this technology? How can we do this technology in the hands of people in the organization? What kind of things happened? Like, were there any moments that really inspired you? Were people working with this technology? Yeah, I would say so. I think the primary experience I had was that if we didn't, not control it, but if we didn't kind of set the target in which we wanted to, achieve something with this technology, achieve something with this technology, it was kind of like a grassroot movement and there was no coordination of how do we and what do we want with this technology. So, we experienced that several departments actually started exploring a bit different options for developing a chat, but actually at the same time, maybe uncoordinated. So, I think if we didn't set the right target and the right training and supported the employees, it would have evolved in a very different way than it did. And I think it benefited us a lot to have a clear, ambition and application areas where we wanted to apply the technology. And this was both in terms of how we are scaling our operations. Like any other organization of 260 employees, we needed to look at are there any efficiency gains or productivity gains? Can we raise quality with this technology? But I think one of the very interesting perspectives that we found out when we developed this strategy was it nudged us into asking these really high-level questions of what is the role of responsibility, what is the role of the role of co-watching our foundation? When our applicants can all use this technology to write the best applications, how do we find the best ideas? So, where I started was in how can we in smaller ways are low-hanging fruit projects, how can we enhance efficiency? But where we ended up was asking these huge questions around what is the role of our foundation? I think there was a lot of interesting things in what you just said. And I just want to launch on to a couple of them. One, I think that idea of the grassroots movement is really, really cool because we see that in a lot of places. This is a really intuitive technology. And when you put it into the hands of people, people start working with it. And we oftentimes see that the same idea crops up multiple different places in the same organization at the same time and having to balance that kind of thing. And at the same time, that prompts many organizations just like yours to ask these bigger, more strategic questions. That, okay, if we can do this, then what else might it enable? And you start to see, and I think that's some of the cool things working with you, that seeing those early shifts of what it enables the role of you guys in terms of working with the applicants and your positioning in the market and the ecosystem. Finally, I would like to touch a bit upon sort of recommendations to other people who might want to get started. If the technology is a few years down the line now, people want to get working with it, maybe they have something in their organizations already. Have you had any experiences that you want to give on to people who might be jumping into it now? Sure. I think we have a lot. One of the most important choices that we made was actually making this technology available in a secure manner to everyone in the organization. And recognizing that we would not be able to come up with all the great use cases ourselves in a closed room in our digital department. We needed to engage the entire organization to start thinking about how can this technology improve my work? How can it have an effect on the entire organization? So I think supporting this grassroots movement and really supporting the ambassadors around the organization. Can we can we even use this technology and still have a solid reputation of trust? And can our applicants trust us to evaluate their applications in a manner that we can, that aligns with how we perceive ourselves as an organization, as a foundation? And I think balancing this, the fear that all the questions around sustainability, security, the ethical questions with the whole interest and potential and ideas has been the tricky part. And this is, I think, our experience is very much from our governance model and how we structure this. So we open up for all the questions. We have established a sounding board of some of the important leaders in our organization to discuss these ethical measures. And especially if we're in our own manners and especially if use cases are high risk or could affect our reputation as a foundation you can trust. But at the same time have a really swift operating model in how to turn ideas to proof of concepts, to see the value, to prove the value for the organization. I think that is really, really cool and interesting. I'm looking very much forward to following you on where this journey goes next. So thank you so much, Mede, for your time. Thank you for having me. And now we'll have the opportunity to hear from one of the architects of this AI future. Nadja Kholsten is the manager and operator of the Gepheon AI supercomputer. That's one of the most powerful AI supercomputers in the world. And it just recently arrived here in Denmark. And as such is really one of the flagships of where we're going with this AI revolution. So let's hear it from Nadja. We're going to be able to do this AI capability, this huge computer and make it available to multiple types of stakeholders. It's available to academic researchers. It's available to startups. It's available to large businesses. Really anybody who's looking to do something really big with AI. So we're helping build an ecosystem. We're bringing different parts of the ecosystem together. The academic part, the business part, the startup part, and the smaller businesses part, all coming together to do something really big. And they're all bringing something different to the table. Some of these stakeholders might have very interesting ideas. Some of them may have very interesting models. Some of them may have very interesting data. And we're bringing all of them together to do something really special and help build for the future. AI is definitely transformative. And it's going to transform a lot of industries. Some people talk about AI like it's the new electricity or the new internet. And it's definitely going to be as big as those things in terms of how much it transforms society and how much it transforms society. industries and different companies. But those are just analogies. So it's not enough to just say that AI is like something else. We actually have to understand what AI is actually doing. So for example, on Gipheon right now, some people are using AI to do mass sequence prediction of different types of enzymes. Some people are using AI to figure out how to best optimize a quantum computing circuit. Those are actual workloads that people are doing. And it gives you a sense for what AI means practically and what AI means. And what people can actually do with AI today. Some of the topics within AI are extremely complex. And it takes quite a bit sometimes to understand how AI is actually doing certain things. But not everybody needs to be an expert in order to use AI and to be able to create value out of it. So the important thing to remember is it's important to be curious. And it's important to ask questions about what kind of industries can AI actually have in the world. impact in. And whatever field you're working in, you can just ask yourself, imagine what I could do better using AI. Imagine what I could do faster or imagine what I could do more efficiently if I had these AI capabilities and actually knew how to use them. So I always encourage everybody to just be curious about learning more about AI, not necessarily because you have to be an expert and understand every piece of it. But it's really important to understand how it will transform society and how it will transform many businesses. I want everybody to be thinking more about AI. I want them to be thinking about how AI is transforming their work. I want them to be thinking about how AI may transform the future, including the society that they live in. I want GIFION to be a place where ideas are accelerated, where people are building towards the future and are doing research that would not otherwise be possible, where they are being very innovative in a way that wasn't possible at this scale before. So let's build together. Let's come up with an AI future that we can all be proud of. I'm starting to get this, right? So there's at GIFION, which is the supercomputer, we have DECAI, we have NAJA, organizations around that and figuring out how and who is going to be using it. And I think that's exciting, right? It means that the opportunities are so large and now so close to home. But it also makes me a little bit curious on all of the power dynamics that come into play here, right? How we organize, who gets to play with it first, what data sets get in there. That's going to have a big influence here on how the development goes. I think that's some of the important next steps that we're looking into. I think right now there's a lot of excitement, especially in Denmark, that we have this computer. And while it's not unique in the world, it is a remarkable machine. And it does enable lots of research and also enables lots of businesses to do things they weren't able to do before. And of course, that also begs the question, like NAJA invites people to work on the machine and we need to do that. Like we need to step up, like as excited as we are here in Denmark to finally be able to be part of the architects of this AI future. It also forces us to actually have an opinion about these things and it forces business leaders and researchers and to some extent civil society to actually lean in and understand some of these really complex things that we're working with. And this also just opens to me that the architects of that future is, of course, partly the architects of the technology, but it's also the architects of how we organize around it, right? What do we prioritize? Where do we start? And what sort of gets to influence the straight? Which leads me to another thing that I wanted to hear your thoughts on, which is this gets so abstract, right? For me, it sounds more like almost primary research of saying, what is it that we can feed this computer? And I'm struggling to understand what exactly it'll spit out or do with it. So as leaders and as organizations, how are we supposed to understand and see the possibilities, right? What's going to make us able to imagine what we can use it for? I think that's totally fair because it has been a step change in what we're capable of. And the complex thing about a machine like Gafion is that there will happen primary research and not everybody needs to care about the finer nuances of quantum simulations at this point, maybe someday. But there is also going to happen things on Gafion that is pretty directly applicable to businesses. And six months ago, that didn't matter. We didn't have a computer this close to home that could do these different kinds of things. But now it actually matters. And I think that Nadja and Dekai is great at inviting in both businesses and academia to start thinking about these things. There is a community around it where you can lean in. And if you know what to do today, I know some companies, they have been excited by this for day one. And of course, they know what to do. And for everybody else, here's an opportunity to listen in and see what can actually be done with this business because it is really hard, but it is also really exciting. Let's move on and see the last part of your conversation with Benedict Evans. I know you also get into these levels of change in that context. I think five years ago, if you had a planning horizon as a company that were three or four or five years into the future, you were probably all good. If you have a planning horizon as a CEO of a major company today that is that long, you're probably wasting your time because you have less of a chance knowing what the future holds in five years because we're looking at bigger changes in the coming five years than we did in the previous five years. Well, yes, but I mean, the question is changes for who? You know, if you are, you know, there are some kinds of company where this will change how your IT org works and that's about it. There are other companies where this will completely change the nature of your competition. And I think a lot of people will recognize the status quo scenario, the increased intelligence. Well, I'm not sure I would say status quo. I would say there is a view that says, look, this is a platform shift. This is as big as the internet. And that's not a funny thing to call status quo. Like there's a base case. There's a base case is this is as big as the internet. Then there is a more aggressive view that says, no, it's a rather bigger shift than that because this stuff is going to scale a lot more. And so this is where you get people saying it's kind of like inventing computing or electricity. The problem is, as an aside, like all conversations about AI seem to turn into hunts for analogies. There's a base case view that says this is as big as the internet. Then there's a view that says it's more than that. What would more than that mean? Well, no one really knows, but like it's a way of communicating. No, it might be much bigger than the internet. And I think that thought is, to me at least, a little profound that it is the base case scenario for a lot of people is as big as the internet. And concluding that we can draft up not insane sounding scenarios where we just admit our imagination doesn't stretch to that. Like we are hunting for analogies and we're having to go as far back in history as the Mayans or imagined space sorcerers. And I think it's a totally fair. I'm in the same boat as you there in saying, we don't have the capacity to imagine if it's a third degree scenario of transformation. We might be able to do something in the second degree. The challenge here is that like, and I should kind of go back and ground all of this, which is like, we don't have a theoretical model that will tell us that these models will or will not scale. The fact that we don't have a model that tells us this is impossible doesn't mean that we should spend an enormous amount of time wondering what would happen if it did. Because A, you've got no, just because you've got, A, you have no model that tells you it can't happen. B, you don't really have a model that tells you it will. And secondly, you can't really say anything particularly useful about it anyway, because you've got no idea what it would look like. And if it did happen, anything we're worrying about today is kind of irrelevant anyway. What's going to change now? Well, everything. Then what? Well, what does that mean? How can we possibly answer that? So if we've had AGI, then we can't, you know, all bets are off. And we have bigger things to worry about than the adoption of enterprise software or middle class unemployment. You know, and I think that's a totally fair question. And I would be inclined to agree. And I think thinking just to that extent about the third scenario and saying, okay, if AGI were to arrive, all bets are off. It would be like the Spaniards invading the Mayans. Like we don't even know what it looks like. Also tells me that, okay, that is not a relevant scenario for businesses to consider. I know in my own advisory, I sometimes get to points with clients where I say, if that were to happen, I'll personally invite you to go to the next level. We'll be able to roast marshmallows over fire and we'll reconsider the world as it is. Let's not spend our time there. Well, so there's an urban legend that says that during the Cuban Missile Crisis, there was a rumor that the missiles had launched and everyone starts selling stocks in the New York Stock Exchange. And this veteran trader goes out and starts buying. Yes. And he says, well, it's binary. Either the rumor is true, in which case we're all dead anyway, or it's not, in which case the stocks are cheap. But it's also funny looking at something like Star Wars. Like that whole movie wouldn't work if they had phones. And yes, phones that can connect across the galaxy. That would be a kind of tricky network infrastructure problem. And in fact, the hilarious thing is, is one of the new Star Wars movies, suddenly everyone has a mobile phone. It's like, wait, wait, A, where did those come from? And B, that completely changes how the whole movie works. But you don't seem to realize that. So you've given them all mobile phones. But still act as though they need to fly places to talk to people. You actually have to get on a ship and go over there to talk to somebody in person. Like the people. And meanwhile, it's the same point now is, why aren't the robots flying the ships? Here we are, however many billion years into the future, and you've got a guy sitting at the front of the ship with a steering wheel. Like, we're not going to have that in airplanes in our lifetime. Never mind by the time we've got into space. So there's a sort of challenge here in how much has happened that we've already taken for granted. And we don't see around us anymore. So you don't see elevators. You don't see that it's automatic anymore. It's just an elevator. And I think that perspective of what does that look like for AI, what does the, like, I think it's a plausible thing to imagine that will, the assistance of the world, the series and the Google assistance and whatnot will grow immensely more powerful. And as you say, that eventually that just becomes automation and becomes a natural thing. It was interesting. It was AI because it didn't really work initially. It was weird. It was human-like, but not fully human. So we called it AI. And then as these things start to develop, it will be the most natural thing that you have to book a plane ticket or if you have to go somewhere, you might just ask your AI. I think that's the favorite demo of all the tech CEOs. Yeah, and it's also the worst demo because it's the one thing that doesn't work because that's not how you book holidays. You don't anymore than, and this is the thing, it's like my point about interns, you wouldn't say to your EA, please book me a family holiday. This summer, whatever you want is fine. Like, that's not how holidays work. Father gave my son a radio a while ago. And my son said, it's an automatic speaker because you just turn it on and music comes out. It's not like the speakers we have where you have to choose what music you want, which of course, I think of it as completely the opposite. But no, you just turn the radio and there's a dial you can choose to have different music. So the technology comes and it changes everything and then it disappears. We don't see elevators anymore. We don't notice cars. We don't think of how cars changed the world or how low waves changed the world or how cheap air travel changed the world. And I think that's a wonderful place to end it because I think the point is great that while the world might look strange and weird at this point and we're looking at a technology that we don't quite know how to grasp, we most definitely look back at it in a few years and think, okay, a lot of this was super obvious. I think that'd be the words. Thank you so much, Benedict. Thank you. It was a pleasure. Okay, Adam, we're nearing the end. What note do we want to end on? I think it's been a tremendously exciting hour that we had here. We have had a lot of interesting conversations. And what I really hope people take away from today is that while things are really strange and weird and AI technology is developing super fast, it also means that we have to envision multiple different futures. Some of them might be ridiculous, some of them might be transformative and some of them might be somewhat mundane. But this is one of those situations where we need to hold multiple thoughts in our heads at the same time. And that's a real challenge. But we have to imagine those possible futures in order to anticipate what's coming. I agree. And it's going to challenge us, right, to be able to balance both a combination of sort of grounded realism and common sense, and then on the other hand, very, very ambitious and bold imagination of what could be possible. And I think that what a lot of our guests also spurred on today was that sort of urge of saying, well, you cannot predict the future, you cannot control it, but we can influence it and we can take part in designing it. So the question becomes more, what can we do as organizations to make us able to do that? What capabilities do we need to be able to be a part of designing our own futures? Agreed. And I think some of the stories we heard today really gave some good examples for that. I think both FTFA and the Novo Nordisk Foundation's transformations and Nadja's invitation to be part of that design. But that's the conversation that we have going forward. That's the conversation we want to have. So after this, there'll be a lot more to come. We'll have a lot more interesting industry viewpoints and concrete cases. This was very much the beginning of a bigger story that we want to tell. So we hope that you stay tuned and you want to keep following us as we have a lot more interesting. constant coming. So with that, all that's left is to say thank you very much for joining us today. We really hope that we've left you with some inspiration and sparked some good and curious conversations out there. It's been an absolute pleasure and we are so excited to see you out there. Thank you.