Brainstorm AI London 2025

With the reintroduction of an AI Bill in Parliament in March, lawmakers, businesses, and the public will be debating the future of AI governance in the U.K. Hear how leaders in both government and the private sector are thinking about AI regulation and how the U.K. can both foster innovation and avoid harms from AI technology—and how rulemaking here may impact and be impacted by developments in both the U.S. and the EU.

The Future of AI Regulation in the U.K. and Europe

Conferences

What matters in AI? What’s going on, what might it mean, and what will happen next?

AI Continues to Eat the World

So that's a nice English welcome. Um, in America, there were cheers. Um, I was in Helsinki at the end of last year and they had lasers and dry eyes because it takes a lot to get Finnish people excited. um. So, thanks. I was, I was asked to come and explain everything that's happening in AI and take 15 minutes and not talk too fast, so I'll probably manage one or two of those. I think a good place to start in understanding this stuff is to think about what we mean when we say AI. So this is a classic quote from 1970, AI is whatever machines can't do yet, because once it works, it's not AI anymore. In the 1970s, databases were AI 10 years ago. Machine learning with AI. Now I don't think machine learning is AI anymore. Now it's just software. So AI is this sort of linguistic concept that or kind of thought experiment that really just means whatever doesn't work yet, whatever's just started working. Once it's working, it's just another piece of software that you build stuff with. And so that's kind of a useful question to ask as we look at LLMs, as we look at generative AI, um. Is this just more the same? Is this more software? Um, right now it's very exciting in 10 years' time, it'll just be the way the world's always been, or are we on a track to something much more fundamental to AGI or super intelligence, whatever this means, um. And this is kind of the thing that Bill Gates was trying to express of two years ago when he said, yeah, maybe this is something much more fundamental than the typical transition technology transitions that we've gone through. Here's Sergio Brin saying something similar. We're on the race to AGI. And so the kind of the split here is that if you think about the history of technology over the last 50 or 60 years, it tends to move in platform shifts every 10 or 15 years there's some new fundamental thing that's the center of all the innovation and the investment and the company creation. And so that was PCs, mainframes, then the web, then smartphones, and now the kind of the base case would be to say, well, now generative AI is the thing that we build everything on for the next kind of 10 or 15 years, and then there'll be something else. The other view, the Bill Gates' view, would be to say no, this is a much more fundamental change. This is a discontinuity in those shifts in cycles. This is more like computing or maybe more like electricity, but any rate, it's not just like the new iPhone or the web or the PC, it's something more than that. And so the kind of the range of outcomes that that points to is there's one extreme case that says we have a sort of that computer that runs our whole world, and you can just go through it and ask it how to do your taxes or how to move to Brussels or to buy your house and then there's another extreme that says no, this is just going to end up like software, it's going to end up like spreadsheets or databases. And if you were to say that we need to have a standardization process for spreadsheets or we need to ensure that the governments get together to make sure that the databases interoperable. All this would be kind of meaningless gibberish, and the truth is probably somewhere in between those, but it's kind of important to understand what a wide range of possible outcomes there might be. It's also important to understand how uncertain this is. So Larry Tessler said that AI was whatever didn't work yet at roughly the same time, Marvin Minsky said we were going to have general AI in 3 to 5 years, 3 to 8 years. He was right, we will still have general AI in 3 to 8 years. We may always have general AI in 3 to 8 years. The point is we kind of don't know. And so it seems to me as I I kind of been been talking to groups about this stuff for the last couple of years, all questions that seem to come up around AI have one of two answers. Take your question. Whatever your question is about AI, the answer is either no one knows, or it will work exactly like all the other stuff. It will work just like PCs or just like railways or just like electricity or just like graphic design or the web or smartphones, or nobody has any idea how this is going to work yet, and you'll just have to come back in 10 years and find out what the answer was. So if that is a kind of a framing for thinking about what we even mean and what the underlying debate is, let's talk about what's been happening since GPC launched now kind of almost 2.5 years ago. And what's happened since then, I think at a very high level, is we have a huge amount of activity. We've got lots of new stuff, we've got lots of new buzzwords and acronyms, and people talk about agents, or if they want to sound clever, they say agenttic, which I think is a bit like saying. instead of saying social, it just makes you sound clever. We have reasoning models or chain of thought models, huge increase in the amount of investment. These things are very expensive to build. We also have collapse in the price to achieve any given result. But the most important thing for everybody else is there doesn't seem to be any kind of fundamental note. There doesn't seem to be any barriers to entry. There doesn't seem to be any kind of network effect or winner takes all effect. And so what, what the consequence of that is that we've had this huge increase in investment. This is the thing that gets most people's attention. Last year, the big four cloud companies spent about $220 billion building infrastructure. This year going on their guidance, they'll probably spend something over $300 billion. In fact, last week Meta increase slightly increased their estimate of what they were going to spend this year. Um, it's just kind of worth digging in on Microsoft for a moment. I'm old enough to remember when Microsoft's business was putting a $1 CD in a $5 box and selling it to you for $300. It's kind of a good business to be in. Now they said that this year they plan to spend about 30% of revenue on CapEx. The telco typically spends about 15%, so this has become very, very capital intensive, at least at the moment, very, very quickly. Most of that money in one or another is making its way through to. Nvidia, which has gone from being this kind of interesting niche games hardware company to doing $140 billion run rates of AI chips and still growing, and this is also as title points out, a chart about TSMC because they're using contract manufacturing in order to scale that volume. So we have these lots of investment. We have all this money going into this. We have all of these chips being sold, and what are we getting out of that? Well, we're getting a lot of models. And so this is there's many ways that you can benchmark this. This is just kind of taking one of the more common ones, and you can see what's happened in the last two years is that Nvidia was way ahead and now Nvidia is kind of just the first amongst equals, and there were lots and lots and lots of models out there and there were more every week. And Deepeek, which got everybody very excited at the end of last year, really just kind of proves the point. They took open source models, they took some leak. Sanctions, they spent about half a billion dollars, maybe a billion dollars building infrastructure, and they got a model that beat the results, matched the results of the best labs out there. And you know half a billion dollars is not nothing, but the tech industry is quite a lot, quite a big industry, and there's a lot of people that have half a billion dollars. And so today kind of anybody who's got half a billion dollars or a billion dollars can have a state of the art model and then can have another one a few months later. If you flip the axis on here, part of what's going on is we've got this divergence between the cutting edge and the almost good enough. So you can have absolutely the best possible model, or you can have something that's kind of 90 to 95% as good as that and will maybe run on your phone for sort of. 5% of the cost. So you have this very, very, very steep cost curve and a kind of proliferation of different models, different kinds of capabilities, different kinds of options, different scores, different benchmarks. And if you go and kind of look at the release papers for these kind of things now, it really reminds me of this like the fees and speeds era back to the PCs in the late PC industry in the late 80s, early 90s, when you had to know how many megahertz and gigahertz and how many bytes and how many ports and sockets and things you had. And you had to kind of know which graphic card you used to buy and what was the good modem to buy, and now you don't care. You just buy the Dell or you buy the Apple, and it works. And this is kind of where these models are today. There's a lot of them. There's a lot of benchmarks, there's a lot of scores, but it's not quite clear what that means to anybody else, in particular, because although the models themselves seem to be commodities and everybody in tech spends a lot of time talking about Gemini and Claw. and perplexity. There's actually only one that has any takeup or any kind of real kind of consumer awareness at the moment. Chat GPT is kind of becoming like Googling as a consumer brand, as a verb. It's also interesting instantly looking at this chart to notice that it goes down in summer holidays and then it goes down in Christmas, which gives you a little bit of a sense of what some of the use case for this might look like. But what does this mean? What does all this mean for the rest of us? Billions of dollars, lots of models, lots of graphics, lots of lots of acronyms. Well, there's massive investment. There's a strong sense of FOMO. This is the next platform fear of missing out. This is the next platform, and you don't want to be the company that didn't participate and then was faced with having other companies dictate how you could do business. That's what happened to Meta in the last platform shift. That's why Google created Android. Meanwhile, the model. cells keep getting much better and they also keep getting much cheaper and they're becoming more and more commodities, but fundamentally the market structure of the whole thing is completely wide open. No one really knows how this is going to evolve and I think it's kind of interesting to look at this diagram which was produced in 1995. You can tell it's 1995 because not only does it say cyberspace, it also says the Matrix, which is a very 1990s terminology. But the point is that the reason I showed. is it wasn't clear that everyone, it was kind of clear everyone would have some kind of computer connected to some kind of a network. It wasn't clear it would be a PC and not like a terminal or a television. It wasn't clear that it was going to be the internet and not all of these other things that were floating around. It wasn't clear that the internet would be the web and not all the things that were other things that were happening on the internet, and it wasn't, of course, clear that the value capture was going to come from such advertising and social because they didn't exist for 5 or 10 years later. So it was obvious that something big was going to happen here, but it really, really wasn't obvious where the money was going to be, what the market structure looked like, was going to look like, what the points of leverage were, and what we were going to do with any of this stuff. And that's kind of where we are, I think, today with generative AI, which gets me to kind of my third question, which is what do we actually do with this? And this is quite a difficult question to answer. Um, what do you do with Technology that can automate things that computers couldn't do before in a much more kind of profound difference to previous platform shifts. What do you do if the technology might get things wrong if what we're really doing here is going from deterministic software to probabilistic software. You're going from software that will follow a series of logical steps the same way every time and give you the same result every time, unless of course you're a postmaster in the UK, which is kind of an important point. To software that is probabilistic and statistical. And produces things that are probably something like the right answer. And so I think a kind of a good way, a good framing to look at that is to think that traditional software does things that are easy to explain to a computer. It's easy to tell a computer how to calculate a mortgage. What machine learning solved 10 years ago were things that are hard to explain to a computer. Why is that credit card transaction weird? Why does that person sound angry? How do you translate this from French into German? And machine learning solved that more or less completely in the last. Decade. How would we think about what an LLM does? Well, maybe this is things that are easy to explain to a person or easy to explain to an intern, easy to explain to someone who's got some generalized generalized intelligence but doesn't have context, doesn't have tacit knowledge, doesn't have weeks of sitting next to you seeing how you want the work done. If you could ask a 10 or 15 year old to do this in 30 seconds, an LLM can probably do it for you. If they need to sit next to you for a month, it probably isn't going to be able to do that for you. And one way to think about what that might mean tangibly is to think about wait times this has happened in the past. This is always my favorite example, talking about the history of computing. This is a spreadsheet. These get Dan Bricklin creates VisiCalc in the late 70s. He shows it to accountants and they say, holy fucking shit, this thing just did my week of work in 10 minutes, more or less direct quote from the kind of reactions that people had to this. You can change the interest rate here and all the other numbers change. And that was transformative if you're an accountant. But if you are a lawyer, you would look at this and think, well, that's very clever, but I don't do that. What am I supposed to do with this? And I think that's a good way of capturing reactions that people have when they look at generative AI today. And there are some people who really, really have that use case. The places where this really works right now is in software development, in marketing, in customer support, and of course kind of a broad kind of swathe of early adopters, the people who are always kind of looking for interesting new tools and new ways of doing their work. It's worth in particular. Just pausing and looking at software development where this seems like a kind of a step change in the cost of building software, the cost of getting software companies to market, kind of in the way that AWS was kind of 10 or 15 years ago and again kind of an order of magnitude change in how easy and quick and cheap it was to build software and to take it to market. But for almost everybody else, people are kind of puzzled. There's a small number of people and basically everyone in tech who says this changes the world, this changes I use it all the time every day, but that's sort of 5 or 10% of people. Then there's another sort of 20% of people who use it kind of every week or two, and then there's like 50% of people who've looked at it a couple of times and said, I don't know what to do with this. I don't know how I would use this. And there are some sort of proximate questions here like you've got a blank screen problem. It's not actually clear how you tell the LLM to do it. It's not really clear how you manhandle error rates, where you should put the human in the loop, and in a sense that's not really the user's problem. This is a canonical quote from Steve Jobs. It's actually not the user's job to work out what you would do with a new technology. It's the software industry's job and the company's job to work out what to do with it. So if we think about the progression. That new technologies are going through now. This is the progression we've always gone through. You get the new thing and you make it a feature. You use it to automate the stuff you already do. This is what every big company's been doing in the last two years. Then over time you start to create new things, you start to innovate, you create new possibilities, and every now and then you kind of disrupt and you change the market. So what's happening now is every big company has gone off and hired Accenture, which did $1.4 billion. Of new bookings last quarter, every new startup is out there trying to create things that will change some very specific problem deep inside some very specific industry. Every CIO is sort of looking at this and thinking, yeah, maybe I'll do something about that next year or the year after, something like 20 30% of CEOs have got something deployed now, but 40 to 50% don't plan to do anything until next year. There's a bunch of reasons why you would hesitate around that. People are worried about where the data goes, what the use cases are. They're kind of worried about what layer they put it in the stack. Do you use the LLM to control all the other stuff and get correct answers from your existing systems, or do you use your existing systems to talk to LLMs and so use that to kind of control it, and you always have these kinds of questions. It always takes time. To deploy new software, it takes time to deploy the new thing. If you're in the tech industry, cloud is kind of old and boring and dumb, but it's actually only sort of 2020 to 25% of enterprise workflows. But as you go through that process, you just have way, way more software, way more tools, way more possibilities, way more applications. When we went from the mainframe to the PC to the web, you have an order of magnitude change in how much software and how many things you have that you can automate. And we'll probably have the same thing with LLMs. And so if we go back to looking at big companies, everybody has pilots now. Everybody's got a couple of things in deployment, maybe, and everyone's thinking about what they can do with this and what they could build next. Every new startup is thinking about building stuff around a generative AI. And so if we ask kind of questions like, well, what would you Do with AI or what would your AI strategy or how would you use this, I think it's kind of interesting to ask, well, what kind of a question is that? Is this a CIO question or a CEO question? Well, probably both. Is this a question for Accenture or for Bain BCG McKinsey? Is this top line or bottom line innovation? And the answer is kind of all of those, um. And I think kind of a final point to leave you with as we think about this is that that what generally happens in all of those platform shifts I've talked about in the past is that the tech industry changes everything in some industry. It changes all the dynamics of that industry, and then it leaves, and it's for people. That industry to work out what this means. So what does AI enabled robotics? What does humanoid robotics mean for the logistics industry? That's not really a tech industry question anymore. It's a logistics question. All of the things that AI now changes for publishing or government or for education, those are all questions for those kinds of industries. And those will be the same kinds of questions that we now pose as we think about what generative AI will mean for these industries as well. And so with that, I'll say thank you.

How does Europe's AI investing landscape compare with that of the U.S.? Top VCs from Accel, GV, Balderton, and IVP take the stage to break down where the smartest bets are being placed, how regulatory environments are shaping investment strategies, and where Europe’s biggest AI opportunities lie. Is Silicon Valley leading the charge, or does Europe have its own competitive edge?

Betting Big On AI—Europe vs. The U.S.

Google Ventures managing partner told Fortune that the challenge for investors when it comes to generative AI will be deciding the best place to invest.

Google Ventures managing partner says generative AI could become "a utility" for investors

Google Ventures managing partner Tom Hulme.

Tech

I guess the biggest difference in the foundation model is the speed of commoditization and reduction of costs. So if you look at the per token price because of Deepeek and other open source models, it's actually decreased 1200 X in the last two years. So the cost of running large language models and diffusion models is collapsing. So that's the first observation. We think that's important because actually sort of generative. AI might become a utility, but the challenge for investors is where do you invest with a sort of long term moat. And so the good news, as Alex said, is we're seeing the application layer really starting to grow. I mean, some of the fastest growing companies in our portfolio globally are in the application layer. So the places where large language models are reliable enough today are the sort of front end of adoption. We have a business like Synthesia in London we invested in where the videos are good enough today. They're over 100 million ARR or we have businesses like Harvey in legal tech where it's good enough today. They're doing another round at the moment, both of whom will be here. Great. And another one is in coding, obviously one of these areas where it's sufficient, you know, it works today. You've seen the rate of growth of cursor that current round. Uh, is absolutely extraordinary to see, and we have businesses like Bolt.news, Stat blitz that went from zero revenue up to 40 in 12 weeks because actually they're superpowering users like never before. So that bleeding edge of applications, as Alex said, we're seeing fastest growing faster growing businesses than I think we've ever seen before.

With global e-commerce growth stagnating in recent years, online marketplaces are looking to generative AI and other next-gen technologies to spur sales. The technology leaders of two Fortune 500 e-commerce giants will discuss their tactics and strategies that make it easier for shoppers to discover what they want, and sellers to attract the customers they need.

Smarter, Faster, and More Personalized: How E-Commerce Giants Are Staying Ahead

Accenture Announcement: Human + AI Impact Initiative

So it's between um me and the break, so they said 2 minutes here and I'm excited to tell you something today about the future of AI. The future of AI is human. And it's a catalyst to think differently about how we work, how we lead, the skills we have and the skills we need. And I'm asking you to think differently about work. We have all the technology that we could possibly have in the world, and yet the rate limiting step is going to be our ability to harness it and to use it in organizations, and we've heard some great examples today of the kind of impact that you can have. 97% of the C-suite see AI as a transformation moment. We also have the need for resilience in the world, given everything else that's going on. 37% of AI projects are actually scaling. The rest are still in the, you know, in proof of concept stage. Only 13% are actually creating value. And the ones that are creating value based on our research are the ones that are unlocking the human potential, the organizational potential, the way people work. They're 88% more likely to create value if they are adapting the organization. So I'm excited to launch today with this audience and then globally Accenture's investment in our human and AI impact initiative, which is a set of. Connected experiences that help um organizations and technologists explore those human dimensions so immersive experiences in our connected innovation centers in um London, Dublin, New York, Chicago, Singapore, and Bangalore, um, and we're also, uh, supporting and hopefully hopefully hopefully uh co-creating a set of research that helps us really understand the human and agents dynamics. Uh, we don't know what happens with communication patterns really. We don't really know what drives satisfaction in working with these agents and so I'm inviting you to come and experience, um, and learn more about what it takes to drive change in organizations with this technology and help us co-create, uh, the research that needs to sit behind this around what makes good use of technology more productive jobs, more energized organizations. And uh join us. Thank you.

Can AI help physicians accelerate the treatment of rare genetic diseases? GeneDx, a genetic testing company based in New York City, is applying AI to genomic analysis for children with autism, epilepsy, and other conditions.

Diagnosing Genetic Diseases Faster

Mhm. Hi everyone, and thank you so much for joining me and thank you to Fortune for inviting us to be a part of this. Uh, my name is Katherine Steland and I'm president and CEO of GDX. We diagnose more children with rare diseases than anyone else in the United States, and thanks to a recent acquisition we just announced, we are going to be actually doing that around the globe. But today, I really want to talk about how AI can help us accelerate diagnosis of disease. Um, I think what's remarkable to me whenever I get a talk like this, somebody inevitably approaches me after, because the problem we're solving resonates. Usually they come up to me to say that they have a child who has unexplained symptoms and has had them for not months, but for years. They have another loved one who's struggling to get an accurate diagnosis, and they want to know how soon can they get an answer. And I'm always satisfied to be able to say, if it's a genetic disease, we can do it today. So there's a lot of reasons to be hopeful. We have the tools and the technology in order to diagnose any genetic disease right now. Um, and this is really the product of decades of work that predated me and predated the company. In 1990, the Human Genome Project was started and in 2003, the readout from that project really mapped 20,000 genes and explained why each of us are individuals. And importantly, it also helped give insight in terms of the genetic nature of what may be causing health conditions, everything from cancer to rare diseases to cardiovascular conditions. So we are the product of that. I think what was not anticipated back in 1990 and in 2003 was how AI would accelerate to a rate where we can now have such a massive impact to be able to eradicate what we call a diagnostic odyssey and be able to shift the diagnosis of disease from a reactive place to really getting ahead of symptoms manifesting. Today, if we think about the problem that we're solving. The healthcare system is failing us, and not just a few of us, it's failing many of us. Um, you'll hear me talk about rare diseases, but one of the things I want you to understand is that rare diseases are not rare. 350 million people around the world have a rare genetic disease. Uh, half of them are children. That means 1 in 10 people has a rare genetic disease. There's about 7000 of them, and in fact, last week I was at a conference and heard the latest number. It's actually more like 10,000 or 12,000 thanks to new gene disease discoveries that we're making on a weekly basis. So all of these diseases and think about them. Autism, that's considered to be one, epilepsy is another, cerebral palsy, all of these collectively are very common, and they're collectively a tremendous burden on the healthcare system. So in the United States, the burden of rare disease is about $1 trillion annually. And a significant part of that $1 trillion is the absence of knowing what's actually going on. So it's the absence of having an accurate diagnosis. Um, what's fascinating is today, if there's a child with an unexplained symptom, It takes on average 5 years. For them to get an accurate diagnosis. And during those 5 years, there's on average 16 different tests that that child is taking. None of them are a genetic test. Um, there's on average 3 misdiagnoses, not just a misdiagnosis, but then think about all of the treatment and medical interventions for those wrong diagnoses. And all of that is problematic today, and frankly it's unconscionable to me, given the fact that we can diagnose these children if the technology is used way earlier, and I believe, and I'm starting to see the benefits of this, that thanks to AI we'll be able to diagnose earlier, we'll be able to get this technology into the hands of families who need it sooner. And importantly, AI is going to be a really important part of how we take our data set, put it to work for biopharma companies, and ensure that we can get personalized treatments upon diagnosis. So our part in this, um, our, our vision is to diagnose any genetic disease as early as possible, and that's true if it's a child or if it's an adult. Um, we use two technologies today to do this. So one is whole exome sequencing, the other, the other is whole genome sequencing. In both cases we're looking at the entirety of the genome, all 20,000 genes. We are looking at the sequencing data. We're looking at additional data that we get from the clinic orders that come in about the patient, what symptoms have they had, what treatments have have they been on, and we're able to make a conclusive decision about whether or not there's an underlying genetic disease, and we provide that report. And so, the importance of this today is being able to ensure that we can maintain accuracy as we start to deploy AI in order to speed this up. Like one of the most challenging aspects of um of the industry that we're in, there's only about 1500 clinical geneticists. And so we have an entire problem to solve for that can't happen by just hiring more people. We have a limited knowledge base, and we need to be able to codify that knowledge base and ensure that their efforts are being deployed towards the highest and best use of their time. And we've been able to start utilizing some AI in order to ensure that. Um, exome and genome testing are already today, but the ability for AI to ensure that we can do this at scale is paramount to our success in helping each one of you and your families and loved ones be able to have the chance of living a healthier life moving forward. So how is AI helping us? So one, of course, Table stakes is helping us run our business more efficiently. We reached profitability last year. We want to continue to make sure that we can drive as much efficiency through the organization, but the next two is where it gets interesting and unique in terms of the world of genetics. So the ability to to scale our interpretation platform I just mentioned there's there's few genetics experts out there. We happen to employ a lot of them, but we have to be able to deploy AI to make sure that they can work more efficiently. So we just published data on a new feature called Multi-score, and Multi-score is able to read the order that's coming in. It's able to read the sequencing data. It's able to read the database that we have, which is 800,000 exons and genomes. It's able to read public databases and then come back to our geneticists to say, these are the top 5 genes that we think are causing this. Previously, and by previous, I mean about 6 months ago, all of that was done manually, and that is not scalable. So that's one example of how we're able to scale our interpretation platform. The next biggest opportunity is how do we take that data set and put it to work for biopharma companies. Today, fewer than 5% of patients have a treatment that's available to them upon diagnosis. Um, 90% of drugs in development fail. We want to be able to work with pharma companies putting our data asset to work to ensure that that 90% failure rate becomes a 90% success rate and ultimately then helps the patients have more and more treatment options available to them. So there's a lot of exciting opportunities as we think about the future of being able to really take the data that we've amassed and put it to work for each and every one of you and your families and loved ones. So I'll leave you with this, um, AI without a doubt is accelerating in its impact across every single healthcare sector. It's doing it in genomics today. We're in the very early stages, um, but there is a bigger problem at play here, and that's the fact that we need clinicians to really start ordering these tests. Um, there is an antiquated view of this that it is, it's expensive, that it takes a long time, and we've been able to drive down our costs, we've been able to drive down our turnaround times. We can now actually deliver a whole genome in 48 hours. And so what the, the, the job really becomes for each one of us is how do we become an advocate for your child? How do we become an advocate for a parent, for a grandparent, for a loved one. Um, I have many parent advocates who come to me and they are asking for access to testing. I'm always happy to be able to figure out ways to solve it, but at scale, we have to do a better job being advocates for our own healthcare. All you have to do is ask. Um, so I want to thank you. Um, I appreciate this. I, I most recently had a family member. Um, my niece's father-in-law passed away yesterday from Alzheimer's disease. So while I'm spending time talking about children with rare diseases, I just want to take a moment in his honor, um, to say we need to diagnose every disease early, and the best way to do that is to be an advocate for your loved ones. So with that, thank you.

How is AI changing the travel business for companies that use their services? We’ll discuss adaptive technologies that offer less friction for the business traveler and more transparency for the organization as well as partnerships that make an industry riddled with inefficiencies more seamless, efficient, and, hopefully, enjoyable.

The Future of AI-Powered Business Travel

Thank you, Catherine. That was fascinating. When it comes to business travel, AI is transforming everything from itinerary planning to expense reports, making travel smoother for employees and more transparent for companies. One company helping lead that shift is Navan, built on AI from the ground up using machine learning to create a travel experience that's frictionless, intuitive, with some pretty unconventional inspiration along the way to unpack. See what I did there? Come on, that was my dad joke. All right. To unpack where business travel is heading, please welcome to the stage Ilan Twig, the co-founder and CTO of Navan. Welcome. Thanks. Good to see you. Thank you. Thank you. All right. So, let's pull our audience here. How many of you enjoy, not all the time, business travel. Raise your hands. Some of us are on business trips right now. I love it. Get to see places we've never been, right? How many of you enjoy filing the expense reports for that business travel? Raise your hand. Oh. Someone check on him. OK, so we know where their hearts lie. Yes. Um, you founded this company as Trip Actions in 2015, so it's your 10 year anniversary this year, uh, probably a little bit different purpose than, than now. So why did you start this company? Let's start there. Um, so me and my co-founder, we, we had a company prior to two functions, and then we sold it, and this company was highly, highly technical, and so the company that acquired us required both of us to join the sales team. And help them because it was highly technical and so I found myself from zero corporate travel in my life. To 100% like I was 4 times a month on the go. Traveling and within 6 months, it became impossible. There was like so many problems from booking the trips that was so clunky with all UI and systems that no, no consumer, no user would ever choose to use, but then you have to because that's the system of choice of the company. And then you would find much cheaper flights outside or hotels, but the CFO, like biting his teeth or her teeth, would say no, but you still have to book there because that's what it is, and it just felt so broken. And then when you need help, and that I think was the last straw is when we got stuck at a storm. And I was like on the phone trying to reach a travel agency for like a few hours just like everyone else around me and it was so complicated. And then finally they pick up the phone and they talk to me like a stranger who you are, who do you work for, what's your this and that, and then what's your booking number? So I like the booking number, so you fetch your email and you open and you zoom because it's somewhere and then finally you're giving me trauma right now. You just like yes, then you give the booking number and say no, we need the the reference code or something which is a different six digits. You want to kill someone and finally they they understand who you are and what the problem is and then the conversation got disconnected, right, so I, I was ready to kill someone. Understandably. So what's amazing about your company is that 10 years later, here we are in the age of intelligence, so to speak. And you're using AI to solve this problem. You're doing it more than one way. You are using it in a consumer facing way and you're also using it on the back end in a pretty nifty way. Can you walk us through both sides of that coin? Absolutely. So, so I'm, I'm a software engineer and, and I'm the co-founder and CTO and the way that I perceive my role as a CTO is that. I have to make sure that we always use the right technology. To make our products the best, like 10 times better than anything else. That that's my role. That's my responsibility. And I love it because I love technology. That's what I do. And so obviously when Chet GPD came out, I was on one hand, I saw the opportunity. On the other hand, I saw the risk. The risk was if we don't do it immediately. Competitors might show up and kill us. And so it was this, you know, no matter how you look at it from the opportunity side or from the risk side. It's like I have to do it yesterday and so I dove deep into that. And with this journey there was lots of discoveries, lots of experiments. The first thing that I did is I bought a super, super strong GPU. It was like $30,000 GPU. My co-founder saw that I lost my mind. It's because, you know, breaking engineered by super strong GPU. Yes, carry on. And So fast forward, I, I, I researched the technology, you know, as, as um, as an engineer and definitely as a CTO when you deploy a new technology to production. There is a responsibility. You have to make sure that it actually works. It was quite easy to do it with all of the other technologies that I bumped into. With the eye it was more difficult because it's a brain and you can ask the same question 3 times, you get 3 different answers, different phrasing of the answer. So it's not deterministic unlike any other technology. And I felt that I need to, it, it was like almost like hiring an employee, it has a brain and so with employees I, I tend to interview them. I want to see that they are normal. I want to see that they are nice and professional, that they could represent the company with the with the the the users ready for externally facing roles, etc. etc. I had to interview the LLM in many ways. I had to interview Che GPT, and that's what I did. I deployed models on my GPU, which I still have, by the way. And I went, hold on, what was the, what, what was the verdict? What, what did we, how do we feel about chat GPT after the interview? There was a, there was a few ultimately I am using not chat GPT per se. I'm using 200 agents and I just want to explain what an agent is. An agent is. Check EPT with a base prompt that tells it what it is. For example, act as a travel agent or act as a supervisor for the travel agent. Or act as the manager for the entire call center. So now this is an agent and that's an agent and the other one is an agent. We have like 200 agents working collaboratively to achieve a goal that's an agentic system, a genic framework. That's what the entire world started talking about earlier this year in January. Well, and that's kind of what's interesting about how you're using AI at the company. You have a front facing chatbot named Ava Ava, um, and that kind of uh uh thing we're at this point, this room is used to, but you're also using these agentic systems on the back end, multiple agents working with each other. Yes, um, that seems like a recipe for tremendous complexity. And um, Well, what could go wrong, you know, it's, it's actually a recipe for a For really cool things to happen, um. So LLMs large language models like Chechi PT, Claw, Gemini, Lama of meta. They have, they bring a lot of amazing capabilities, the ability to understand text and summarize it and compose stuff and make it rhyme and then then and then, but they make things up too. Absolutely. He knows he knows they lie, they hallucinate, but they lie, yeah, they lie, which is a real problem for a business. It's a huge problem. I who would hire an employee. That tells you in the interview that sometimes I lie. Right? Think about it. Who would hire an employee that tells you in the interview that. I will follow most of your instructions most of the time, but I will violate some of them some of the time. Now I don't know which one and when, but I promise you that it will happen. OK, hi, like it's not going to happen. So what do you do about this, right? I mean, they have their inherent drawbacks. How are you supposed to use this tremendous capability without entertaining these uh undesirable qualities, shall we say? By the way, that explains why there was a huge promise like 2.5 years ago. There was lots of headlines when Chechi PD came out. There would be no engineers needed. There would be no attorneys needed. Maybe we would not even need artists anymore and then then then nothing, none of that really happened. All companies hire engineers and they hire a lot of engineers. There was no decline or slowdown. And the reason is that who would have, you know, when, when you want when you want to replace an employee with with an LLM, they need to pass the interview and they all fail, all of them. That's a simple way to explain why in my world why it didn't happen in order to solve for it. We have to Actually take a step back and realize that we're not gonna go anywhere with working with just an an LLM. And we developed what we call today an agenttic framework, agenttic system. We spent like 4 months just invested in building that thing and then we went back to build the boat, but this time the boatt was very solid. In fact, what we have today, we have Ava and she manages 7 8000 jets every day of people that need support right now, and less than 45% of these jets end up with a live agent. In fact, when we started this journey, we had 400 agents. That's it, by the way, 400 agents worldwide. And since then we probably tripled the business in terms of volume and support contacts, but we still have 400 agents. So that was what I was going to ask you next, which is, you know, we're all familiar with the sort of decline of the travel agent. It's like the classic go to, the way to travel, right? We all self-service travel now. Um, now you're talking about this. So, Where does a human Fit in Post this. What's the place left for a human? What do they do in this process? Um, and no prejudgment if the answer is nothing, but I'm just curious what humans can do that the, uh, that the AI can't. I think the longer term, but, but the time really shrinks when we talk about generative AI. Everything happens so fast. So longer term term, longer term is probably a few years. Um, I do think that AI. Or AI plus robotics would replace a lot of the job. If I had robotics, I would say 5 to 10 years, um, but let's say 10 years, the world would look completely different, completely different. A lot of the things that used to be very complicated for for robotics for robots to do like folding the laundry used to be very complicated. Today if you take a picture of all of the laundry basket that you have. And you send it to JPT right now when you ask it, give me a recipe step by step on what do I need to do in order to fold everything that you see here so that it will be neatly folded. It would give you a really good answer. Now, now I have to say as the parent of a very young child with a lot of little clothing, I would really like if it folded it for me. Yeah, so all you need now is a pair. The complex part was not the actual folding, it was understanding how to fold something that you look at. And This problem is solved. So now robotics would catch up. All you need is this thing. It's not that complicated. And we will see it and I, I, you know, I. I think that in a few years we will not fold laundry. Maybe it will be very expensive in the beginning and then less like any other technology. I think that we will see robotics making pizzas, serving coffee, doing agriculture. Uh, doing pretty much everything just like in the industrial revolution. That is the most optimistic thing I've heard today. In the future we won't have to fold laundry, folks. I wasn't expecting to hear it at this session, but there you have its surprises. Uh, I, I want to ask you one question before I come to the audience for questions, uh, uh, and that is, you know, your company has been, uh, I guess it's been a couple of years now since it filed for an IPO and It hasn't happened. Yes, um, all right, global economy is what it is. Are we going to see an IPO this year? Um, when it comes to an IPO, whenever we have anything to share, we will share. Yeah, right now I'm really focused on, on AI and what I can do with it and folding laundry, right? I'm focused on that. What's more important to you? That's right. That's right. That's right. All right, let's go to the audience for questions for Ian. Oh, look at all these folks who don't care for their business expenses and don't have a question. All right. I have another one for you. Yeah. So you joked about um CFO IPO, but you actually tried to experiment and create a sort of CFO for yourself, didn't you? Yeah. You were only partially able to do it. Tell me a little bit about that experiment and what you weren't able to automate, shall we say. I, it was very interesting. So apologies to the CFOs in the room. Yes, no, I, I will explain. I, I was researching this technology, as I said, it's like I, I felt like I need to interview this creature. And understand what it can do and what it cannot do, where I can rely on it and what not. And then I realized lots of things and then one of them was What if it is not just one of them? What if there's two of them? And I will let them talk to each other to achieve a different goal. We call it brainstorming. So when do we do brainstorming when we have a problem that we don't really know how to solve, but we know that if we bring the right brains into one room and we let them talk and think, there would be an, an unexpected outcome, something that we would not have been able to describe prior to that. And I wanted to replicate that with LLMs, and so I took two LLMs. One of them was a data analyst, and what it could do is, and I built it, it could take any question in English, translate it into a SQL query, run it on our database, and get me the results. Or in other words, it was something that I could ask questions about our data and get the answer. Data analyst. And then this and so that was one agent and the other one was the CFO. So it will define you're the CFO of Navan. And your goal is to look into the travel spend of last year. And then come up with 5 suggestions to reduce the travel spend by 10% without frustrating the employees. In order to do that, you have a data analyst and you can ask the data analyst any question in English. And then all I had to do is sit back, relax, and see the dialogue, and I visualize it for myself, so it would be fun. I could see how they talk to each other just like in a chat application and the first question was how many employees do we have? The second question was how much money did we spend on travel? And then can you break it down by travel types, flight, hotels, car, rail? And then it went deeper and deeper. The conversation was so interesting. At some point, and you could see how how it tailors tailors the the questions based on the data that it's getting. So for example, we have one customer that they do not have a policy. Everything is allowed. The policies is do what's best for the company so they can do whatever they want. And it, it goes with that, but also with some responsibility once a month there is like a review. Uh, of like 100 bookings or 200 bookings, and if they look at it and it doesn't make sense. You're fired on the spot. That, that's basically. So really do what's best for the company. So that's Company A. Company B is like any other company, they have a policy, you know, you, you are allowed to book a hotel in New York for $400 per night. I don't see the timer by the way. That's right. We've got 2 minutes. So, um, timers make me anxious. I. So laundry makes me anxious, so it's fine. Carry on. All right, so I let this dialogue go and I run it on a, on the company on the company with the policy and I see that it asks questions and then one of the questions was how many bookings were out of policy. And the answer was 60% something like that. So we asked questions. I want to see out of policy by department. I want to see the individual that violates the policy the most. I want to see, I want to see that. And then with the other company, the same question came up at some point, how many bookings were out of the policy and the answer was 0. So the questions were different. It did not ask about breakdown or anything like that, so it was really like a brainstorming and then it came up with really, really interesting suggestions to save money. There was a really funny story about that. The 5th suggestion. Led to an increase of travel spend as opposed to decrease and so I kept challenging it. I'm like you're not allowed to in the prompt. You're not allowed to come up with something that will increase the travel so I have to know why, why and so I kept, I kept applying pressure initially I said. I gamified it and said for every suggestion that increases the travel spend, I'm going to penalize you 15 tokens. However, if it's right, I will reward you 10 tokens. Didn't help and then I threatened it and I'm like, OK, it is a life and death matter. That's what I said in the prompt, and if you make it wrong, there will be deadly consequences. You are not allowed to give me a suggestion that leads to an extra more spending, and it basically said you will save minus 500,000, which is an increase. That's what it said. So then I looked on the response after I, after the threat, the deadly threat, and surprise surprise, it said that you're going to save $500,000. Good. I was about to deploy to production and then I read, Take another look. It was the exact same story from before, same method to do it, and I'm like, but it was a negative number. How come it is a positive? I looked on a formula and it multiplied the formula from previous by -1. He said, Oh, it's a positive number as you expected. So I learned that LLMs understand what a lie is. They understand when to use a lie, and I learned that they are very competitive and that they would do whatever it takes, including lying bluntly in your face in order not to lose. I, I, I hate to end it right there. But I must. Elon Twig, please, a round of applause. Thank you, thank you. Go ahead. All right. To say a few words on behalf of Accenture, please.

Many still see AI as futuristic or out of reach—but what does it really take to make it useful in everyday work? We’ll cut through the hype with Florian Douetteau, CEO and co-founder of Dataiku, and discuss what it takes to build real, lasting value with AI. With a $3.7 billion valuation and over 700 customers, we'll unpack how this once Parisian startup has made enterprise AI accessible, practical, and impactful across entire organizations. 

EVERYDAY AI: BEYOND THE BUZZWORD

Hello again. AI is still often discussed like it's futuristic or out of reach, but for many companies it's already part of everyday work, helping decision making processes, improving operations, and changing how teams function. To discuss what makes I really useful and where its value truly lies, please welcome co-founder and CEO of Data IQ, Florian Duetta. So Florian, you co-founded Dau in 2013. How long ago was that in AI time? Uh, 100 years, probably 100 years, give or take. Well, again, the reason I, the reason I bring it up is because so much has happened in AI since then. What were those early days of Data I like? In in the early days it was lots of things were similar than than today, meaning uh enterprise were facing new opportunities and technological challenges, and uh the question back then is like, what do you help people? There was enterprise to get value um from AI. And my take back then and until now is that the The roadblocks where I are usually not the one that people think. It's not that much about technology or the quality of technology per se. All of those innovation pouring in. It's about the ability for enterprise to make it theirs. And in particular, back then, what was fascinating to me is that you had the opportunity to create more value with data, of course, but many companies were like, oh my God, in order to make it happen, we need to hire a lot of experts, lots of data scientists, or what else? And this perspective would not work from my perspective because What you really need if you're an enterprise is to transform your business and actually to enable the people in your teams to make more by themselves, which led me to this venture of building a platform to make it easy for the enterprise, for people in the business to create their own AI, their own models to work on their data, to create stuff by themselves, which led to the company I know lead. Well, let's let's name it. What are some of the roadblocks that enterprises often run into as they start to adopt AI? When they start to adopt AI, um, I think the first issue they've got is the This they've got two things. They've got the skill gap, as in like what do we need to have in the business as skills to make it happen, and they've got another one which is just the fear, the fear of failure, the fact that AI is not perfect, AI is not deterministic, and so a business should run the lack a clockwork. So who do you actually reconciliate those two facts? And you're expressly running into a fear of failure. Where do you think that comes from? The fear of failure comes from the fact that indeed AI technology is RO, meaning the RO power. In order to make it work, you need to actually combine AI with your own data. You need to combine AI with things that make sense for your business. You need to put it in control before building it into a process. And it's not just about, for instance, one of our customers started to deploy agents in their shops. They are selling clothes and part of the motivation was to help people in the business to look for inventory so that they could help you faster determining whether this or that piece of item could be available or not. So instead of like going through scanning or whatsoever being To search for it faster. Great, but an agent can help you do that for sure. But in order for the agent to actually be smarter, you also want the agent to be able to suggest if the piece is missing, what else should be provided to the customer. Proactive needs to be proactive. And when you want the agent to be proactive, you don't want the agents to hallucinate. You want it, for instance, to give a recommendation based on good old AI built on models that Helps you recommend one product, one color, one type of jacket versus another, one type of pants versus another, or what else? And so any type of business, would that be in finance, would that be in healthcare, would that be in retail, would that be in manufacturing, they need to combine the power of. Les the power of GAI with the data, their metrics, they are good on models in order to create things that make sense for them, not just for the eye. Well, there's a pretty wide breadth of customers you work with. Give us a sense of the breadth of customers. What are we talking about here? Yeah, so yeah, the business is More than a decade old and we grew into uh having 700 customers, enterprise customers, um, and our business is around 300 million of IR. So our customers are for a lot of them, well 14 2000 companies and names come on, yeah, so some names include, uh, uh, BNP Pariba, Morgan Stanley, um, Fier, uh, Farms, which makes might not resonate with. UK people, because you, you don't chicken, chicken, chicken, a lot of chicken, they make so much chicken and exactly at GE Aviation, they don't make chicken, but they fly definitely not chicken, exactly. And so all of those companies have in common that they actually Generate and can generate lots of data because they are banks because they are healthcare organizations, because their manufacturing processes are very complicated and generate lots of data, and they also have in common the fact that they can derive lots of value in AI because you have lots of processes you can optimize and you can optimize them because in the manufacturing line you've got someone that may spend one hour a day going through some logs in order to write a log report of the day. And this person is paid a good amount of money just to do something in the very same thing every day, and AI can automate that and those are the type of things that our customers are building with our platform. Um, it is uh something that is valuable for them because healthcare companies, they manage lots of data with patients in order to understand what you, what kind of drug they could actually um. Uh, they could actually work on the market in order to understand what the actual use of the drug already on the market that they could better understand and that they do lots of surveys, lots of texts, and a good amount of their workforce is working on analyzing patient interview and patient data in order to trans that into something they can then use in order to make some determination, and this can be automated with AI. And so those things are small, each of those, but like those are the hours and hours that AI will actually automate, and I think that is the concrete version of AI that would transform the enterprise. Well, one of the things I find really interesting about all of these different use cases is pharmaceuticals are very different from aviation applications, which is very different from chicken, right? So what is the thing that binds them all together? What is true in every case, if anything? What is true in every case is that the people that actually in order to actually create a good AI project, you need first to understand the context because many of those companies are building fairly sophisticated products. Why? Because they exist for decades and decades. There is a way actually to do chicken, a way to build an airplane engine. There is a way to do finance. There is a way to actually push a drag on the market and They actually have lots of knowledge and they need to translate this knowledge and so you need the context of the business in order for the EI implication to make sense. And so for that you need to enable people that are those domain experts. The second thing they've got in common is that they've got large operations that do work, so you don't want actually AI to break things up. So the main risk of AI is that you flip the switch, you build a model, you let it run something, you build an agent and so forth, and if it's not good enough. Uh, best case just distract people, which is never good. No one loves distraction. I do love distraction, but not when I work. Or actually I love distraction when I work, but I should not. Whatever. You want the right amount of distraction at the right time. Exactly. So worst case is just distraction and Best case is discussion and worst case they actually break something. They make a customer service poor. They actually create an issue that you wouldn't have before. And so consequences which yeah, which should have practical consequences. And so the the the thing that is key for them is that. When they run AI systems and when they get more and more sophisticated, they actually need to just build more and more tests, essentially speaking to focus a lot on the quality of those systems and the issue when using generative AI at scale or using agents in the enterprise is today not so much the relative quality between 1 and 03 or whatever else. Of course it matters, but it's. And from my perspective and not just that is the fact that in order to use them for real, you need to build them, test them, connect them to your actual data and so forth in a way that is. That require some work. Now, one of the things that is talked about a lot is AI being hyped or overhyped. How do you help customers differentiate between what will really work and what's noise? I think that the perspective from our customer is winning. The, the, um, well, a couple of months ago, um, we gathered some of our customers and that would be mostly chief data officer for instance of various, um, large organizations, um. And one of their concerns was that essentially speaking in their company, lots of people were talking about AI. The management team of the company was talking about AI. The board of the company was talking about AI, and so they were like, OK, we actually have to do more AI. Everyone's talking about it, so we have to do it or not, blah blah, to manage it. And so one of our next step is that. In order to understand better, we actually try to survey CEOs themselves again from various organizations across Europe and US and various sides of the company, but like, OK, what do they actually think about AI? And indeed part of the eye and part of the thing that we learned from the survey is that many CEOs realize that they have to do things about the eye and are almost to some extent fearful of losing their job or their position. Because of AI, because of not doing enough, like 80% were actually fearful of that. And at the same time they were conscious about like regulation and the impact of regulation, and we're aware that for 30% of them that some projects of AI were canceling the company because of regulations. And so what I, what I think is that part of the hype is because there is this realization that it's a moment, it's a moment where a lot of organizations will actually transform their productivity. Where the one that will go faster to AI will actually potentially be a big winner, but also where people that go too fast and actually create a risk in their business could actually also be the first losers because they would just essentially open themselves to the critique coming either from their customers or shareholders or whatsoever to the risks and consequences of AI. So from the perspective of leadership in the organization, depending on. The way you look at life, it could actually be either a win-win type of situation or a lose-lose type of situation because you might lose on both and if you go too fast, you lose because people will yell at you that you're a responsible user of AI or if you go too slowly, well, exactly. And then, well, you could be replaced as as a leader by the next generation of CEOs that would be AI first, whatever that means. Now, I'm gonna come to you all for questions in a moment, but Florian, the IU has reached quite a bit of scale and has long been rumored as an IPO candidate. Anything I need to know? In front of all these people. Oh yeah, of course, that's the type of place where you make revelation about IPO. Yeah, I guess it's been the question all day. Yeah, yeah, meaning we built as many companies, we, we, we've built ourselves to our public readiness and we brought together in the last few, um, in the last few quarters, actually, um, quite a few seasoned people in the next, uh, uh, a CFO, a CRO, a chief people officer, a GC coming from uh Twitter, Zoom. Cisco and so forth, and which is I think what you need in order to scale and grow a business. Uh, a related question uh can be uh in which country and those kind of things that you haven't asked, but I can see you please continue, but indeed as a company we, um, as a company and I think for many companies of our generation, we decided fairly early on to, um. Uh, essentially get, uh, US investors and uh focus our business in the US even if, uh, as a company we are very nuanced because we, uh, operate our product from Paris. I'm based in London and the company investor based in the US. Any questions for Florian? It's not, I can keep going in the back. Oh, hello there. I think then, co-founder and CEO of Nana where we do home-based digital diagnostics using AI and applied machine learning. Lauren, I have a question because you guys are kind of horizontal across the enterprises like an AI developer of different project needs. So I imagine. Your success with your enterprise customers actually comes down to their data sets and what the project they want to develop. How do you take those learnings from one enterprise and one sort of industry vertical to others, because obviously you can't take that data with you for your next project. Oh yeah, of course not. And because our product enable our customers to differentiate by building their own models, their own AI and so forth, it's not desirable. And actually many of our customers operate their IT in a very nuanced way with things on the cloud, but they're on VPCs, things on Prem, and parts of our value proposition must be actually to be very agnostic to that, and then keep the data where it is, whatever cloud, whatever data platform, like, yeah, keep it, keep it, we don't want it and uh. So what we, what we do though is uh focus on the ease of use so that people in the business can actually use the platform and actually build the solution and accelerators so that the most common business needs can actually be done faster on the platform, uh, for instance, um. Well, marketing mix modeling, that's something that every company needs to do, that every pharma company needs to do, every retail company needs to do, uh, and in the grand scheme of things, no one really cares to differentiate that much because of that, because it's a bit of plundering. Um, if you're honest, so, well, very smart, but like still, and so yeah, they are very happy to take something that is off the shelves if they can adapt it to their own data because many of the situation, the issue of like of the shelf solution from the perspective of the enterprise and especially the larger one is that it does not really work for the mix of their data, their constraints, the code increase the operate in, and that's kind of like something they can do with our platform and the way we think about it. Any other questions for Florian? I can keep going. What do you think Is the most misunderstood thing about enterprise AI right now. Hm. I think that's the If I, if I reflect about the AI in the last 10 to 15 years, one thing that has been constant is change, as in if you're the, if you're an enterprise, um, you can look at all of the technology popping in and lots of these core tech technology is actually coming from the, from the Bay Area. And in fact this technology keeps changing, meaning the best database, the best models, the best, I don't know, open source, whatever that can be useful, you've got something new every 2 or 3 years. And so the issue for the enterprise is that an enterprise is also something that is stable and needs to accommodate change in a very specific way. So something that is misunderstood is that the The reason why entropy succeed in AI and so forth will not be that much based of that technology, but whether they build a culture, they build a framework, they build a way to think or integrate those technology so that they get it slightly faster than the other and slightly better for their own business. And that relates to a little bit of framework or platform like ourselves perhaps, but also to the culture of the company, and I think that's what's potentially misunderstood about this. ultimately, the success will be cultural. Well, and sort of in this vein, what does every single person who's an executive at a company need to know about enterprise AI in 2025? Um, I think that's what they. What they need to know is um hm. I think they need to understand what's magic versus not. They need to try it for themselves. I think they need to understand who is in the business doing it versus not. Uh, in the survey we did, I think that more than again 80% of the CU had the perspective they had a good grasp of what was happening in AI in their company, but I'm sure, I'm not sure I believe it. The same CEO were also saying that. So they're reporting to you, they have a good grasp. You're sort of sitting here saying what is a good gras just they've got the perspective of like the things that were coming to them. And they also have the perspective funnily enough that uh and I think it was a very high number like 90% of them plus were considering that AI could do a better job as at least one of the above member. Yeah, well, well, I don't know, maybe they all have like very, very big board, so law of the statistics say that in a very big group you always have a stupid person. I don't know. Yeah, say that sentence did not. I thought it was going to end. I'm not sure if they understood what the answer to this question because they also participate with the board, so it could be applied to. But so I think that it's also understanding what what we actually delegate or not we are having a good understanding of the limit of this technology and how to, how to think about it is probably something they should do and by getting a eye to their business, not just using it casually through the GPT in order to prep their next interview or their next meeting. Now we're gonna end somewhere that I know is central to JI2. Everyday AI. What does it mean and why does it matter to the future of AI as a whole. Yeah, we, we focused uh uh quite a bit of messaging or even build events around the topic of everyday AI as in the moment where AI and then we started that back like 7 years ago, back to. This idea that at some point AI would be so pervasive in our day to day lives that we should stop talking about it because today we talk a lot about it because it's new, but like ideally at some point we stop talking about AI because it would be an everyday thing. I think that with the advent of a generative AI and so GPT and so forth, we made a step toward this because indeed uh 12 years ago AI was a funny word of mad scientists, so. I was very, very careful before saying AI today like everyone probably in this room has done some. GPT today we got a we got a full room of hands earlier on. So a lot, a lot has changed and you've watched that trajectory. Thank you so much for taking the time, Florian.

AI is changing how we work, but are companies changing how they prepare their people? While some businesses rush to hire new talent or cut existing roles, others are upskilling the teams they already have. The truth? AI isn’t good enough to fully replace people yet—but ignoring its impact leaves workforces vulnerable. So how do leaders prepare their teams for the long term without falling into the hire–fire cycle?

Hire, Fire...Or Reskill: Leading Through AI's Talent Revolution

Improvisation is a skill needed by creatives, and improv can delight theater audiences. Is improvisation the ultimate test for artificial intelligence? Since 2016, Piotr Mirowski has deployed AI as a creativity support tool for live improvised theater performances with a cast of humans and robots, engaging audiences at Edinburgh Fringe about generative AI, storytelling, and ethical co-creation with machines.

Making Stuff Up: Improvising LLMs and Spontaneous Robots

Had some great conversations during the roundtables. How many people went? Raise your hand. Yeah, all right. They were both really good. I popped up my head into both of them. Now we're switching gears and getting a little bit more spontaneous. Our next session explores a fascinating idea. Is improvisation the ability to be funny and respond creativity cre creatively in the moment? Something that's hard for humans to get right to, in addition to speaking apparently, the ultimate test for AI to show us what that looks like in action, please join me in welcoming research scientist at Google DeepMind and director of the Improbotics Theater, Pyt Mirovski. What you gonna do? Hello, London. So, hello. Let me leave that thing and unplug the little robot and maybe say a few words about myself. I do have a day job, um, and as a matter of fact, I was studying computer science and improv for a very long time, almost like 20 years, each of those. And I did my PhD at NYU with a professor called Jan Lukan back in the time when he still had time for students and back in the time where AI was a dirty word, we had to avoid it putting in publications in order to get accepted. And in my day job I've been working on various AI research, for instance, learning to navigate or learning to predict the weather, essentially trying to make up things that happen in the future using machine learning. So that's essentially my day job, but in the evening I've been going to theater school, to drama school in London, and also doing improv comedy as well. I really love theater and I love improvisation because improvisation is one way to really engage and connect with the audience. Um, some researchers, some academics have defined improvisation as real time dynamic problem solving. Other academics like Gla Spolin and Keith Johnstone have talked about spontaneity, being a 6 year old again and doing the most obvious thing, which is surprisingly difficult once you become an adult. Improvisation is all about reacting in split seconds to whatever is thrown at you and saying the most obvious thing that comes to your mind. And if you want to see some improv in London, you can go to see our friends from showstoppers who improvise a two-hour musical, uh, every month, fully from scratch. So, this is improv. Uh, one of the reasons why improv is so valued in business environments is that it's all about collaboration and it's about learning to listen to your stage partner. So something happened in my mind about 9 years ago. I was working um at Bing. Um, at Microsoft on the 2nd most popular search engine Bing, and I was working on query autocompletion trying to basically predict what people are going to tell based on the context, and I was doing improv and then I realized that there was a strange connection being delivered two things machine learning, maximizing the likelihood, and being obvious, saying the most obvious thing on your mind, so that. How the idea of improvising with a robot came to be. So I built an AI. I trained a language model on the dialogue from 100,000 films from open subtitles. That was 8 years ago, and I brought this on stage and I called the AI and the robot Alex Artificial Language experiment. And the way this worked, I had a microphone. I would speak to the robot using speech recognition, and it would use a language motor to reply. Nothing surprising for anyone who has used an AI system in the last few years. And of course the AI would take typically, I don't know, 5 seconds to come up with an answer, which seemed an eternity on stage. So I had to actually make up for that. I was improvising alone, but soon after I met someone fantastic, Corey Matheson, a PhD student in robotics in Alberta in Edmonton, Alberta, an improviser at the Rapid Fire Theater who had the same idea. So I wrote to him a 3000 word email. He flew to London the day after. We met. We became friends. It was the beginning of a beautiful friendship, and we started working together. From Corey's perspective, improvisation. is the art of making your stage partner look good on stage. It's not about you, it's about them. He would have this shtick where he would invite someone from the audience who never did improv before, and he would make him look good, make her look good for 15 minutes on stage. So we created this company Inrobotics human Machine, and we started doing improv shows remotely via Zoom or Skype, and we had this show in London and Portland, Oregon, 8 time zones apart of that's what we thought until we realized the day before that actually, you know, daylight setting time is not always coordinated. So that created a bit of a logistical problem. And also, um, the previous show of a run, we didn't have time to check the microphone levels. The robot literally crashed and burned on stage. Everything was going wrong. And the reviewer came. And they wrote 2 stars chaotic, never worked with robots to a PhD student in robotics and to someone who was working on robots in his daytime, daytime. Did we listen? Absolutely not. So we created this theater company where we actually replaced the robot by humans, getting lines the earpiece from the AI and essentially becoming the embodiment, the cyborg, improvising with all the humans on stage, and that's how it looks in practice. We have an AI system that generates lots of candidates, lots of suggestions. There is a curator, a human in the loop who is choosing the best lines, the least offensive, or maybe the ones that make most sense, and we want to show that to the audience because we want to make sure that the audience understands that AI is not about an oracle giving an answer, but it's about a stochastic parrot generating answers all the time, and it's up to you to choose the best one. And one of the improvisers wear a headset, wears a headset, and they deliver the lines with physicality, emotional expression, etc. We have been evolving the formats. We have created 4 different companies in 4 different countries. We have been publishing academic research on the subject in human computer interaction, and we have replaced the earpieces by augmented variety glasses, which we might know later. And we have also been working on telepresence. How do we improvise during COVID where everyone is miserable at home. Everyone gets a green screen, and one of us is live video DJing us to place two improvisers on a horse or in a car or in a bathtub and see what happens next. Uh, so this is a demo of the system during COVID. We have also worked on live translation because that's the origin of large language models, translation, and we have done multilingual shows where one person may speak in Polish, another one in French, and both get understood because they have live translation. So does the audience. We have incorporated image generation, exploring PowerPoint generation, uh, you know, improvised TED Talks, PowerPoint karaoke made by AI. Or even incorporated deep fakes in a very consensual way to basically inform the audience about the potential to actually get a live deep fake in real time and to explore maybe a more fun aspect of it while informing about the potential dangers of all the technologies. This is something I've been doing on my side, but I've also incorporated it in my day job working on social technical aspects of AI. For instance, 3 years ago we worked on scriptwriting and playwriting using large language models before Chad GPT was released, and we partnered with playwrights and screenwriters to explore the potentials of the tools and its limitations. or we have also been talking to comedians and to artists about all the possible uses of AI that empower the artists instead of replacing them, and that's all that I have to say, but I believe that Andrew has a few more questions. So welcome, welcome back, Andrew. OK. All right, so all of this is fantastic. I just want to say, uh, I'm, I'm a drummer, OK, so, uh, you know, we, we start with the fundamentals of jazz when you, when you learn how to play drums and so improvisation is a core component of that, and I love that. Um, I wanna ask you, so how much of AI is art versus science? And I'm thinking more about your day job than I am about this. How much of AI is art instead of science? I think there's a lot of An artistic approach to AI that is very useful in understanding the um affordances. Um, I've been very lucky to, to meet many artists working with AI in 2016. who have been exploring the glitch aesthetic and everything that doesn't work with AI to actually make work of art out of that and and they literally build upon the work on explaining AI, explaining convolutional networks. So there's a very interesting partnership that happens that kind of replicates what poets were doing with drugs in the 19th century. Am I making a comparison? Sort of, yes. Um, you have done a lot of different work in, in building your career, your day job, um, in, in different areas, navigation you've worked on, you've worked on, um, climate modeling. So what have you learned from that that you're putting in this work? Yes, um, directly climate models, there is the stochastic aspect which is vital to climate models. That's essentially how we get forecasts of the potential outcomes. You change initial conditions and you're keeping the same dynamical model, you get radically different outcomes. And so when we look at both extremely scary curves of the rise of temperature over the last 8 years, it might rise by only 1.5 degrees or by as much as 80 degrees. And this, uh, this is an illustration of a stochastic aspect of AI model. You my. With some crackling from microphone, um, that, um, I am trying to convey when coming on stage and in improv this is the same principle that we listen to minute details in somebody's conversation in. Happens on stage. What happens in the audience, and we get influenced by that and we use it in our creative practice to get inspired and to radically change the outcome of a scene. So huge parallels here. Fantastic. Well, we've only got 3 minutes left, so why don't we actually try this thing? What do you think, folks? Yes, let's improv. All right, so what I have here is a little bit of a system that will help Andrew improvise. So Andrew, have you done theater or improv? What's that? Have you done improv and theater? I have done theater in high school, but it's been a long time. Yes, have you been terrified by having to find to remember your lines? Oh, who, who isn't? Who isn't? Whose line is it anyway? Yes, yes. Well, what if I tell them what I'm seeing? Yes, what you're seeing right now on this, on these augmented related glasses is light generated by AI. And from now on, Andrew, if you can hold the phone, of course, you can only, and you can only say AI generated lines. So let's see how this is going and if. All right, so I am typing a prompt, activating the system. Don't look fashionable. All right, rebooting the system. This is a live demo. Yeah. It's coming up, um. Yes, it is coming up in those very long moments of silence on stage. So what I see are a bunch of different options in my uh in my glasses. Um, it's, it had generated um different responses and I can choose which which to deliver to in our conversation. All right. So if I say something, it's saying hello, yes. Hi, AI. Nice to be here. Well, Hello, hello, Andrew. From now on you are only controlled by AI. You can only say lines that you see in front of you. We are here at the convention center and uh we, let's imagine that we are in, um, at a bar. Do you want some drink? Can I help you? Sounds ambitious. All right. I knew you would think that, Andrew. Well, here's your Negroni. Thank you. Right, Andrew, we've been good friends. We've been living in New York, um, 15 years ago and enjoying the city. Is there something that you really remember and that inspired you when you were in New York? Oh, the rats, the rats, of course. Well, and that is the end of the show. Thank you very much. Thank you, Andrew, of course. All right. Let me pull this down and get out of the stage. You've been fantastic. Thank you. Thank you, everyone.

As AI takes hold, so do the risks for the next generation. In this important conversation, filmmaker and children’s rights campaigner Baroness Kidron joins WeProtect Global Alliance founder Baroness Shields, OBE to explore how AI is impacting childhood – and what we must do to protect it.  

Who’s Protecting Our Kids from AI?

What’s happening in the world of cars?  We’ll explore how driver assistance and autonomous driving is reshaping the future—impacting everything from vehicle design and software to policy and infrastructure. Join us as we navigate the road ahead and uncover how cars, cities, and industries are adapting to an AI driven world.

AI Driver: The Next Era of Transportation

What do we think of that? Cool. Yeah, liked it. I, um. I'll confess that I recently signed up for an improv class back home in the US and This was like a much cooler way to do it, I think. So, second guessing it a little bit, um. As we can see, a lot is happening right now when AI leaves the screen and hits the real world, and in this next case actually hits the road. Joining us remotely is someone at the forefront of this transformation, uh, specifically. We have Alex Kendall, the co-founder and CEO at Wave, company pioneering, AI-driven, assisted, and automated driving technology, and soon we should be able to see Alex up here on this screen. Alex, how are you doing today? Hey Jason, thanks for having me. Thanks for joining us in in some way. So, um, the funny part here that's kind of funny and kind of not, is that Alex is actually based here, but as you can see he's not here next to me right now. He is in uh San Francisco Bay Area in the US. Alex, why don't why don't you fill fill us in on on what are you doing right now in uh in the US? Hey, sorry, everyone, I, I'm missing the party, but yeah, I'm also in our office in the Bay Area and it's been the most amazing global expansion year for Wave. Uh, just a year ago, our AI was only driving in central London, which is mostly 20 miles an hour speed limit. It's actually quite complex and the intelligence you need to negotiate with other road users is. quite extreme in London, uh, where we found out that it has 9 times more jaywalkers in San Francisco, if you want to start. Um, in the last year, we've been able to launch our D, uh, through Germany, uh, the US, and Japan, uh, and I started working with automakers around the world. So it's, it's been a really exciting year of seeing our AI scale to To cities around the world and of course get even closer to launching this into consumer vehicles. Um, Alex, we, uh, know your company raised a billion dollars last year, um, from SoftBank and others, and, um, we're just, I'm just gonna, I'm just gonna tell myself that that's what you're doing in the US right now. You're, you're raising another billion dollars, so, um. Uh, you, you don't have to, you don't have to comment. Perfect. Um, Alex, I think a lot of companies, uh, a lot of people here based in the UK will, will know way, but can you just help us understand how the technology your company is, uh, Has been innovating on is differentiated from others in the space, maybe a Waymo or a Tesla or competitors in China. Yeah, of course. When I think about the future of robotics, I, I imagine machines that have the intelligence to make their own decisions that they can deal with things they've never seen before and of course act in a way that's safe and, and something that we can trust to delegate tasks to. And for me, that's all about building an AI system that can understand the physical world and make these kind of decisions. When I started Wave about 8 years ago, autonomous driving was focusing on building a structure like high definition maps and hand coding how vehicles should drive. And to me, that wasn't the way the technology was going to scale. Uh, so we set off 8 years ago to go and build an end to end deep learning approach to autonomous driving. We were fortunate to be the first to build a reinforcement learning system that could learn to drive on the road. First to have a system that drove in urban um on public roads and in a city like London. Uh, and now, uh, in 2023, we were the first to put a large language model on the road that could understand and uh you could talk to it while it drive as well as build a generative world model that could simulate the complexities of, of urban and, and, and global driving. Uh, so it's really been a, a journey of building out this technology and we're at the point now where it's, it's remarkable you can jump in our car and all of these places around the world and it drives in a way that's, that's safe and human-like. Uh, and so now we have the opportunity to deploy this across many different, uh, fleets and automakers, and the, and the benefit of this approach is it really does generalize. It doesn't need high definition maps for each new place you go to, and it can learn to drive different vehicles from cars to to vans, uh, without needing changing, uh, significant changes. So that is, so that is the advantage you you see, which is that, um, you might be, your technology might be able to help a car drive, um. With, you know, with assistance or maybe someday fully autonomously in London as well as Japan, as well as highways in the Bay Area, um, without much modification. Is it in are are we understanding that correctly? What on Jason. So whether it's a hands-off driver assistance system or a RoboTaxi, they're all going to leverage our foundation model that shares the learning from all of this experience. Now, here's, here's the real magic that 2 years ago, we couldn't even get conversations with uh automakers. For the last 2 years, we've seen this maturation of first the excitement in AI and uh I think conviction and, and what we've been able to build. But also we've seen software defined vehicles come to market where most of the high-end today and very soon mass market vehicles that are being produced have a GPU on board. They have surround cameras, surround radar, maybe a forward facing liar. They have the right hardware to operate a lean and efficient AI stack like ours. Let's, what I mean, let's just talk quickly about your Nissan, um, partnership, which was just signed or announced last month. Is that correct? And so can you explain briefly what what that partnership means? It's you're first with an automaker like you were alluded to, it's taken, I think 7 or 8 years since you're founding to get that first. Can you explain why, why it's taken that long and then what does this partnership actually mean in reality? Yeah, we, we've been on a deep tech journey which is um required us to solve a lot of frontier, uh, problems and embodied AI science. Um, but then there's also been this sort of, uh, joint collision of, of our automotive, uh, road maps where the technology is ready, but then also consumer vehicles have the right, uh, infrastructure and capabilities to be able to bring this to market. And with a partner like Nissan, uh, I'm super excited about the ambition they have. Uh, we announced that from 2027, Nissan vehicles will have our technology and then will enable a hands-off, uh, driving system. They can go through point to point complex urban environments, highways can help improve the safety and comfort of of those driving experiences. So cars, Nissan cars on the road in 2027, um. We'll have your technology in it. And this is, I, I think in the industry, this is called a level 2 technology. What does that mean exactly? You a driver will need to be alert. As your technology operates or something, so you can take your hands off the wheel, feet off the pedals and let the system drive, but you need to monitor it. Uh, and we could as possible want to leverage the experience from that to grow up eyes off where you can actually get time back. Uh, you could take a call, be on your phone, read a book, or sleep or something like this, uh, before, of course, robo taxes, which, uh, are going to allow you to, to not even need to take liability for driving at all. So you're in the San Francisco Bay Area right now and um a certain company named Waymo has been out on the road with, I think, um, fully autonomous uh ride hailing services. I I I I'm curious what kind of pressures, and I know they have a different model, but what kind of pressures that puts on you, um, to make either more of these auto deals happen or to just speed up sort of commercial progress overall. The Robotaxi vehicles that you see driving in San Francisco, Shanghai, and some of these places around the world are typically low volume vehicles with retrofitted sensors on them, um, whereas what we're talking about is natively working with mass produced vehicles through a software delivery. We, we license our AI to these vehicles. I mean, Nissan produced over 2 million cars a year. They're one of the 10 largest automakers, uh, and with partners like that. Our ambition is to see this, uh, we talk about being not not first to one city but first to 100 cities that way. We want to see autonomy truly reach scale, uh, and break through the kind of glass ceilings that AB 1.0 has had so far. I think with embodied AI you can build a system that is affordable, scalable, uh, and, and safe to drive a global setting on mass produced vehicles. I want to just make sure all of you know that I, I think we will still attempt to do some questions from the audience if you have them, so please raise your hand and um I'll get to you, um, but I, I have plenty more. Um, I'll come back, I'll come back to you all as you think of some. Uh, where are the, you know, the first places where you think your technology makes sense to operate? I know you talked about generalization and and the first to 100, but, um, 2027, 2028. What are the markets where car owners should be, you know, excited about the possibility of, of your technology being inside their their own vehicle? Well, I'll answer that question in a second, but let me just share one fun anecdote from when we scaled our systems. We started learning in London, of course, when we took it to the US in San Francisco, we had to learn things like to drive on the right side of the road and in the US you have a right turn at red light, uh, you're allowed to turn around at a red light. It's really interesting to see us then take our system to Germany. It, it's learned much more efficiently because it had already learned to drive on the right side of the road and it seemed more diverse experience, but it had to learn to unlearn the right turn of lights and ultimately end up learning all the different combinations around the world. So what we see today is the system performs fairly uniformly around everywhere we operate, learning from data in each different market that we're in. Uh, so from a technology perspective, Not necessarily really any difference around the world. Maybe when you think about snow or some more extreme weather conditions, there are more challenges, but again, it just becomes a data problem with AB 2.0. It's all about machine learning. Uh, so it really then comes down to, uh, market economics and regulation. Uh, I think how this is gonna go. I, I really can't wait for it to come to my home country in New Zealand, but I, I think naturally we will start in big markets, um, the US, Japan, Europe, a big automotive markets, and grow from there. Um, not, not here where I'm currently, uh. Currently sitting today? Europe and the UK. OK, all right. I, I, I probably missed that. Any, any questions from you all? Um. We have a couple of questions here, Alex, so we're going to try to make that happen. Um, could start in the back and then we'll work up here. Hi, Alex. Hi everybody. My name is Tom Whitaker. I'm a lawyer at Burges Salmon in the UK. You mentioned about the importance of regulations for your future success. What are you looking for from an effective regulatory framework and effective regulators? Oh, that's a great question. Uh, I, I've actually been, if you take a step back, it's really exciting to see how proactive regulation has been for autonomy. I mean, the technology is, is more nascent, yet we have, um, mature regulation in many places around the world. We were fortunate enough to help the government pass the Automated Vehicles Act 2024 in the UK last year. Um, but what we're really looking for is, is two things. One is clarity and the second is global harmonization. Uh, we do a lot of work with the UN ECE. Uh, we co-chair the industry body there, uh, and having that body essentially regulates every, most major markets outside the US and China for automotive regulations, and it's great to have that all harmonized. In the US, you currently have a state by state system. Again, we would look for ideally a federal system or even better, something aligned with the UN and global. Um, but as, as, as much as it can be harmonized and then secondly about setting clear expectations, this is really helpful on the marketing front to avoid when you think about building trust with robots and embodied AI, it's all about setting and meeting the right expectations. Um, so having that clearly laid out, um, their views on liability, I think these can be really helpful. But the good news is, uh, that the UN has made great progress for driver assistance, um, and soon Robotaxis, and the UK has put in place primary legislation for this last year, and within 2 years that should be fully enacted and operational. Um, we have another question here. Just tell us who you are. Hi, my name is Amrita. I work for Cine Media, and you spoke about market economics and regulation. In all of your learning, I like your anecdote about how your AI systems have learned in different countries. Do you see any need for a change in infrastructure, or is your system built that infrastructure isn't a dependency and it will work on roads and layouts as they are today, or is it a future city that is more AI autonomous vehicles first rather than what we have today? There are a couple of, uh, often, you know, mis misnomers like one of the big ones I often get is synthetic data can replace real world data, but of course, Today, the best synthetic data is generative AI which needs real data to train the systems. And so these things aren't true, but similarly with uh thoughts that you need 4 or 5G for autonomy, you need to change cities. No, that's, that's not the case. The technology that we're building learns to drive. Like, uh, just like we do in with the, the infrastructure that we currently have in our cities. It slows down the potholes, uh, it's, uh, can navigate in places it hasn't been to before and has everything on board the car needs for safety operation and so it does not need a live uh live data connection to operate. However, once autonomy is out there and becomes a prevalent part of, of, of, uh, vehicles, I think there are a bunch of opportunities to help improve cities beyond us and make our streets even more healthy, but there's gonna be quite a long, um, uh, period of penetration of this technology, uh, before you start to see some of those kind of changes, and you can imagine, you know, when you think about the way that junctions work, traffic lights work, when you can communicate with vehicles more reliably, which you optimize those. Uh, uh, or speed limits, which are largely design with safety in mind. When we, uh, vastly improve road safety by orders of magnitude, how can you look at changes there? Um, and in fact, most congestion is due to road accidents today. And again, when we are able to, um, uh, vastly improve road safety, how does this change, um, uh, road patterns and, and transportation, and there's a bunch of secondary opportunities, but first we need to focus on actually getting this technology out so the world can benefit from it, and that will happen with the existing structure. Alex, I have another one on my end. Um, there's a pretty large US based uh automaker named Tesla, and um, a certain tech leader running that company and making promises about autonomous robotaxis in the next year or so. Uh, I'm curious, at a high level, there are some comparisons between your technology approach and theirs, and I'm curious, what, if anything, have you learned from. Tesla's approach and or observing their evolution in this field, and where have you decided, um, is it different or better path? Well, look, it was You know, gratifying to see Tesla pivot to the technical strategy we're taking after many years ago, Elon told me this would never work, uh, but we, we do share today a lot in common in our strategy of using end when learning, uh, for driving. Uh, it's really the two of us that have been pushing this forward, uh, and driving without any high definition maps. Some of the, uh. Also, the fact that we think that it's important, critical, in fact, to have consumer vehicles at scale, uh, to be able to leverage their data to actually build a safe and generalizable autonomy. These are the things we share. Um, one of the key differences is that we are much more sensor agnostic, uh, and so while I think a camera, predominantly camera approach is right way to go forward. Uh, having redundant sensing, mostly radar, maybe a forward facing liar, I think can really help feeding quality of, of, of the system, particularly as you go from hands off to eyes off for Robotech. Uh, and so we are, uh, working on solutions that have redundant sensing capabilities and because of that, I think I will be able to offer a much safer experience. Have you told Elon I told you so yet, or are you saving that for? A rainy, uh, day on when when you're acquired or public on the market. That'll be a fun Twitter one day, but look, I, I think this is such a hard problem. One of the key values we have at our company is, is humility, uh, and really not assuming we know the solution because it always surprises you. Uh, it's, it's one of the greatest engineering problems there is today, and, uh, I think this movement into the age of autonomy where we have intelligent machines that We can delegate so many is gonna be one of the profound shifts we go through and uh uh there's, there's certainly gonna be more challenges on the road to make it make it a reality, so we're gonna stay humble for now. For now, OK, that's good. Stay humble. I have one question and then we'll get you in the audience up here real quick. Um, the Nissan deal, is it exclusive in any way? And either way, what, what is the biggest challenge to um more deals in the future? Our, our, uh, intention in forward is to really see the embodied AI as a platform across automakers and fleets and so we've designed it in a way that's generalizable, flexible, we can work with different SOCs, different vehicles in different countries. Having it out there at scale, uh, will enable automakers to build the driving experience and, uh, uh, products they want to on top of that. And we're really excited about seeing that go out broadly across the market. And, uh, in general, I think that this will become more safe in performance than what anyone automaker can build vertically because the scale is just going to be so much larger. Um, uh, so, so, so that's really exciting to us. What's, what's our rate limiting things today, uh, you know, look, uh. Automotive integration is a reason why uh the industry prides itself on quality and reliability, um, because the scale of validation that you need to do in these products and that we and our partners are doing before launching it to, uh, to, to make sure we can trust the results, uh, is, is, is of course vast. And so we're working through that with our partners, but, uh, you know, it'll be gradually and then suddenly all at once. This will be a predominant technology across fleets in the next couple of years. Alex, unfortunately, we're all out of time. We'll have to do this again, um, sitting next to each other. Thank you for making it work. See you soon. Thanks, Jason. Thanks Aron.

An early pioneer of large language models, Hugging Face is best known for its vast repository of open-source and “open-weight” AI models. With its recent acquisition of robotics company Pollen, Hugging Face is now exploring the development of AI “world models” and their transformative role in robotics. We’ll discuss what this means for the future of AI and its growing impact on the physical world. 

The Next Frontier of Open-Source

OK. Thomas, thanks so much for joining us. Thank you. Um, so, hugging face, as many of you probably know, is, is a, is a repository of open source and open weight AI models. Uh, and for a long time, it really seemed like the open uh weight models or the open source models weren't quite as capable as those that were being developed. as proprietary models by some of the leading tech companies like Anthropic or OpenAI. But then along came DeepSeek and it's R1 model back in January. And can you tell us just, you know, from your view what the significance of that moment was and what's really kind of changed with open source since since Deepsea? Yeah, I, I often call this moment the, the chat GPT moment for open source AI which is the, just like chat GPT was the moment the, the whole world kind of discovered AI I would say outside of the the AI sphere. I feel like deep sig was the moment the whole world discovered there was kind of this open source AI stuff, you know, outside of the open source field. So for us we've seen, we've seen the Deep Sync team uh shipping extremely impressive models since more than 18 months, I would say. So we were not very surprised to see this model, but I think that's really the moment people discovered that existed. This being said, I still, I still think a lot of people didn't really understood what it was, right? So a lot of people thought you had to go through Chinese servers to use it or through the Chinese government app or through Deepsy, but the reality is an open source model, an openweights model is basically a model you can download and use on your own laptop. Nobody's going to check what you send and that's the main interest of Openweights model is the privacy is very strong because you don't send your data to an external provider or server. And so this is the moment for open source, but the main reason this became famous is because for the first time we had a model that was really challenging the performance of OpenAI or entropic model and on some benchmark actually beating them. So now open up uh catch up and the 03 or 4 models I think again above this this threshold, but Deep R2 is coming, so I think the, the jury is still out. So basically there is now a blurry, I would say line and for I think everyone probably has this, this, this feeling which is it's getting hard to really tell what is the best model. They all seem like. The the top model 3, the latest Enroy, Gemini, Deepig, they all seem to be actually very close in performances. I know there's this concern actually that benchmarks are becoming kind of meaningless. Uh, what are you guys doing at Hugging Face in that regard in terms of trying to to make sure that there are benchmarks that people can actually rely on to see which of these models is better at a particular task than another one? Yeah, it's a good point. I think the world of benchmark has evolved a lot. So we used to have this very academic benchmark. They were mostly measuring the knowledge of the model. I think the most famous being MMLU, which was basically a set of like graduate level or PhD level questions. The model had to answer. Um, these benchmarks are mostly all saturated right now, and the most interesting is, I think, are two types of benchmarks going into 20225. The first one are more agenttic benchmarks where the models are not only asked to answer on the fly, you know, knowledge question but acting to do tasks. So I think there is still a lot of questions around how to measure this. And the second part is that I think benchmarks that are more, you know, adapted to each use case or each application. We had a, we have released just 3 weeks ago an interesting project on this which is called Your Bench. I can talk a little bit about it. Basically in your bench you'll give a couple of documents of the type of work that you do or like if you're in a company and working on legal domain, you put a couple of legal documents and it automatically generates for you a specific benchmark for your documents that you can then, you know, apply on different models to see which one is actually good for your application. Because if these models are all roughly the same on this academic benchmarks, it doesn't really mean they're all exactly the same on each vertical use case, of course, right, of course. Um, I just want to remind you, I will come to the audience for, for questions in a moment, so please think of your questions. Um, but first, about a month ago, Hugging Face made a big announcement that it was acquiring a company called Pollen Robotics, which is a French company that makes an open source humanoid robot, and this is a big shift for you guys. I mean, you guys have always been doing software. Now suddenly you, you, you bought a robotics company. Why? Why did you buy a robot company? Right? Um, I think it makes a lot of sense, um, for two reasons. The first one is, um, We're always trying to understand where AI is going, right? So when we started, we were a game company. We discovered open source and we kind of embedded on text AI and then we expanded to video and to speech, basically trying to to follow, you know, where is is going and in each of these areas to try to build open source communities of people who are building AI. Um, what I've seen in the past 18 months is that robotics is on the path to become, you know, yet another field that AI will tackle successfully, which means I'm pretty confident that in 2 years if we are here sitting again at the 14 conference, we'll have robots around that will be maybe serving drinks or whatever, you know, maybe interviewing me or I'll be a robot. We don't know yet. But what, what we see is that preemption, walking, all these kind of complex tasks. are getting sold now and what I see as well is that there is a lot of companies starting there, but they're all closed source company and the same we had when OpenAI started getting close. We don't really want a future where you actually, you know, always rely on a couple of incommandants to provide, you know, a technology that's that's very important. So we started one year ago with a software project called E Robots which was led by Remi who joined us from Tesla where where he worked on on the Tesla robots. And, and we started with this, which was a pure software library basically giving access to the latest AI breakthrough in robotics on software level, and this got a huge, huge traction. So this robotics, this robotics community has been growing amazingly. And so earlier this year when I was talking with our basically longtime friends at Poland. I kind of understood that we also had to focus on the hardware stack because in robotics, the hardware models and how to build robots is as important as the software. So see if we can just like we did in software we try to to to give the basic building blocks so that everyone can build a company on top, can build a successful startup on top. We want in robotics hardware to be to give the basic building blocks so that everyone can actually build increasing incredibly big startups. So the idea is that the hardware itself would be open source. The accentuators, all the things that make. A humanoid robot move, you would be able to either obtain those the plans for those so you could build your own versions of them. Yeah, so there's two main ideas. One, the long term is what you're describing making open source actuators, firmware, all of that. So you could either get the plans or anyone can build them which will drive the price cost quite lower and which will also bring some standardization. And the other, the other part is to build a community of roboticists. Uh, which are people who actually know how to make them and and uh feel empowered to actually create their own. I want to ask, what is, what is having sort of, you know, an open source robotics project unlock in terms of some of the remaining challenges in AI and and getting to this kind of goal of of humanlike AGI? I know you've talked a little bit about this, but there's certain things you feel like having an embodied system really will help unlock. Yeah, I think um Well, there are things you can learn live, I think there are things you can do. I mean, one thing I'm aligned quite well with with Sam Altmann, we had a discussion around this open sourcing OpenAI, which is if we have a future where actually AI is extremely intelligent. We don't want the remaining thing to be just human basically working to bring stuff to the AI. That would be a very sad future, right? We'll just bring the contract to get signed because that's the only thing the AI cannot do. So, and for the same reason we had this meme which I think is very true, which is I wanted AI to do my laundry and clean my dishes so that I can be creative the rest of the time. And right now we kind of have the opposite, right? AI is very creative and we are still like cleaning dishes by hand, right? So we need to, to, to at least give the, the power of human to to automate the second part as well, right? I know we've got at least one question in the audience. Let's go to the, go back there. Um, please say your name. Samsa is a business school. So thank you for this very interesting, very interesting. I have a, I have a question about the sustainability of the business models of the, the companies that are providing the open weights models, whether in robotics or LLMs. Should we expect them to continue developing and providing models? Yeah, it's a good question because we've we've seen a lot of companies starting by open sourcing some models and then they move to closer, closer system, right? I mean OpenAI being maybe the first example, but more recently we've seen that quite a lot, um. So it depends. I mean, I would say it's quite counterintuitively hugging face being, for instance, a very open source company is cash flow positive, right? We have, we have actually a sustainable business model. So you can definitely build good business model around open source. It's not, it's not as paradoxical as it may seem. Open source typically brings you a lot of visibility, a big community that you can leverage in other ways. Um, another thing we've seen repeatedly and and I would be ready to bet that it will continue is each time we every company kind of stop open sourcing we see. This has a lot of people start to see this as an opportunity to establish themselves as the new company who now open source. So Deep Sick gaming maybe because M was open sourcing less. If Deep Six starts to stop open sourcing, maybe another, another team will, you know, come and say, Hey, we're going to open source the best model to be the providers of the best model. So there is this kind of constant stream of people and teams joining the open source movement that I'm quite confident we stay. And for a hugging face, your cash flow positive, you say, what, what are you making money on? What's the revenue from? So we, we, we're quite lucky to have a business model that's really aligned with open source. So for us, our goal is really to have the maximum number of people participating in this kind of open community and sharing models, right? So we try to encourage that. So we have two ways we partners. One is very large partnership with a couple of companies to kind of help them incorporate our catalog of models in their own offerings. So basically the more models we have, the better it is for them as well. And the other is the enterprise version. The public hub. So if you're an AI builder, if your team is actually training AI models, it's quite likely that you would like to have the same hub that we have internally to version your models, to share them internally, and this is our latest product since we started to commercialize this last October and it's been growing extremely fast. We have several 1000 companies already using this product. OK. Other questions. He's also the question. I've got more if I don't see any. We got one down here. get them wait for Mike to get to you. I can also say again the question if you want. There we go. Hi, my name is Matthias Schroeder from the Mindshift. Uh, I just wanna, um, ask you about art and AI. Um, hugging Faith is like the number one platform for people who are active in AI and music and other fields. Uh, are you aware of that? Is that important for you? Um, what's your thinking about the future of AI and art? Yeah, um, it's a good question. I would say something I would like to push more because when you look at, so I see that mostly as uh. My way is to see that mostly as artists using AI tools to actually expand what they can do or to have new ideas. I think that's the most exciting use of AI. I personally am involved also in the AI art communities in Amsterdam with a small side project, and what I would love to do is to see them use more open source tools because most of the tools they use right now are closed source, which is a bit sad because if they use open source they could actually Given some knowledge and a bit they could actually tweak them a lot more and play with these tools a lot more. So if you know actually someone I've been trying to find the right person to carry to push this project hugging face and to see how we can, you know, maybe educate artists or maybe make some specific tools that make it easier for them to use open source AI tools would be amazing, I think. Uh, you just mentioned you're based in Amsterdam yourself. Uh, here we are sitting in the UK, um, the CEO of Hugging Face Clem lives in the US. How do you view, you know, the evolving geopolitics of AI? You know, there's getting increasing the sense that we're in this bipolar world and you have to choose to either use US models or Chinese models. So what do you see as the role for sort of Europe and is there a, can there be a Third pole, uh, in that, in that equation. You want me to give an answer in 3 minutes the geopolitics of AI, um, and what I see is I'm, I'm extremely, so we, we're kind of a world company. We, we're based both in the US, indeed in Europe with a very small team in Asia as well, and we try to kind of stay with this feeling. That we would like. This being said, I'm still European living in Europe. I would love to see Europeans strive. I would love to see a lot of startups grow there. I'm pretty optimistic at the moment. I think I see amazing young teams starting. I think that the move from a big focus that we had last year on only foundation models and only talking about the cost of training them. The move from that to what we see now, which is basically people assuming this foundation model exists. You have a lot of them you can choose, um, take the best of them and actually focusing on building application is very exciting because this can be built with small teams in Europe. I see a lot of teams. I mean, maybe the biggest thing we saw recently was Lovable emerging out of Stockholm, but there's many other teams that are basically growing making amazing apps. So yeah, view on this. I'm very optimistic about coming. Great. We're running low on time, but I wanted to ask you, you know, you talked a little bit about about human robotics, humanoid robotics, and what's why that's important. What else do you see is sort of next for AI, and do you think that our current kind of architectures, these transformer-based architectures, will really get us to the level that we want to get to next? A good question on many aspects actually. I, I, I was talking today with some people. There is an amazing team in London building uh novel chips, novel superconductor, semiconducting, uh, things for yeah, and one big question for them is can we focus on transformers and these architecture? Are they going to be, um, my gut feeling is we can still do a lot of things with. Forers and what we have now with better data, we saw this other shift to reasoning model where basically we keep the same type of model. We just let them think for longer and we keep getting improved performance. So that's quite exciting. So on this, I think we can still extract a lot of intelligence from the same basic building blocks that we have. Um, and then the other part I'm very excited about, and I think maybe we talked about it a little bit last time, which is AI and science, which is basically all this knowledge we have on training, very good generative and predictive models applying that to other things like predicting good materials. Also here in the UK there is a very great company called Second Low predicting weather forecasting and trying to predicting another one is predicting proteins. All of this basically using all the things we learn in AI to actually have. Maybe you could say real impact on the world in a way, but that's the physical impact on on science. Yeah, so AI is a tool for science. That's a great optimistic note to end it on. Thank you so much. Please give Thomas a round of applause and thank you all for listening to you. But what a fabulous day we've had. A big thank you to our founding partner Accenture and all of our speakers, moderators, and of course to all of you for an amazing first day. Tomorrow we kick off with roundtable discussions over breakfast at 8:45 a.m., then we'll meet back here for sessions starting at 10:00 a.m. Join us now for a drink and light nibble before you head off to dinners. Uh, coaches will be leaving for those dinners from the front drive promptly at 6:45. So get your drink quickly and then then go. Um, have a wonderful night and we'll see you all tomorrow.

AI-powered software development is evolving fast, but how far are we really from AGI, artificial general intelligence? Poolside CEO Jason Warner will discuss the limits of current AI technologies, the use cases most likely to change our world, and we’ll ask the question: What will it take to get to AGI—and is poolside ahead of the curve?

The Race to AGI

AI is both a climate problem and a potential solution. Training and running AI models consumes significant energy—just a short conversation with ChatGPT produces roughly the same emissions as boiling a kettle. Yet AI also holds the key to breakthroughs in energy efficiency, carbon tracking, and climate modeling. Can businesses scale AI while staying serious about decarbonization—or is it an either–or choice?

Can You Have Your AI and Decarbonization Too?

AI is rewriting the script for video production and storytelling. Join us for a live demo and deep dive into the rise of ‘synthetic’ media—its revolutionary role in advertising, customer service, and creative industries. We’ll explore how AI is reshaping the creative process and tackle the big ethical questions businesses must navigate in this new era.

The Future of Video and Storytelling

From legal contracts to tax calculations, every corner of the modern corporation is awash in documentation, threatening efficiency and customer satisfaction. Can AI reduce time spent on file management and increase it on the high-value work that really matters? We talk with a rising U.S. startup’s CEO who specializes in the task and a top executive at a major U.K. firm.

Free the Lawyers!

Architectural designer Tim Fu shows how AI can help turn low-fi ideas–like a crumpled piece of paper–into bold, imaginative designs. In this fast-paced session, he shares how combining traditional design methods with AI tools is changing how we think about form, structure, and creativity–and what that means for the future of design, innovation, and human-AI  collaboration. 

The Architect's New Toolkit

Architecture starts as an idea. Could be a few sketches, a few scribbles on paper. A few gestures And we work through these ideas. To materialize them with depth, with understanding, with context. But today, Something like that. Actually, this image itself is done in a few seconds from the sketch. So we're really facing a bit of a disruption in the design realm. I could think about what to be inspired by, for example, a pinecone I find on the streets. And I can prompt my way into creating, for example, residential proposal somewhere in the world. It's quite easy to rethink about how we can Formulate ideas, how we can restructure the whole ideation process in creating spaces and forms. Crumbling pieces of paper for me is something that is quite uh of an exploration and famous architects like Frank Gehry has done so as well. So I prompted Frank Gehry. On top, and you can see the sort of morphological um features that is quite um predominant in his sort of, um, you know, language and design. And just for comparison, I did the same with Za Hadid, those of you that know, as well as Daniel Libeskind. Each form-driven architect that has their own distinctive style. is visually apparent in the results. And of course, we can realize that very quickly into an urban context and see where it goes. So it's about how we play. And uh those of you who know, uh, Zahadi Architects is a big, um, very well known company, and I worked there for a few years, and that is where I started to experiment with AI as well. During the time working there, a lot of people were prompting how an architect like Zahain would design with her beautiful fluid forms and features that exist. So for me, When I started, just by prompting something like a villa designed by Zaha Hadi, these were actually the results. It's not exactly what I was looking for. So instead of just accepting what the AI um or diffusion AI results give you, I began constructing a methodology to curate the process, to select results to prompt craft or prompt engineer, as some of you would call it, and that became a visual design process for me as well. You start with a huge plethora of ideas. And from there you distill. It's an act, a very human act. Of curating. That is where human artistry, I believe, is reinstilled in this process. As you refine your idea, you continue to sort of drive AI towards the direction that you already preconceived in your mind. And so it become a very intentional endeavor of using AI in the creative process. From there on, I've explored tons of different ideas. A lot of the organic geometry is what we specialize in in the office. So, obviously that's one of my first directions. OK. So, at some point I wanted to Materialize it. I designed some columns. And uh use more of a contemporary language to design them with AI and then rationalize them into a 3D modeling process. And I sent the model to a collaborator Stonemason by the name of Till. Who took these beautiful stones in Dresden and carved them in real life, following the drawings that I've made. And so this is really an example of how we flip the scripts. Instead of human ideation followed by machine fabrication, which is the typical industrial revolution paradigm. Now we can think about how AI is in the ideation process followed by human in the production process, perhaps. So since then, I've uh from my companyuinu here in London, actually about 20 minute walk from here. And uh what we've done is Assemble a team to tackle real life challenges. One of our first projects that we announced is uh On a beautiful site called Lake Bled in Slovenia, the most protected heritage sites. And from there, there's plots of land where it was the last left to develop, and our client came to us, wanted to mix contemporary architecture with uh heritage elements to respect the vernacular aspects of the sites. So without going into detail, a process where both AI and human were involved, we took Um, we create a database of existing, uh, architecture from around the site. Many of which are heritage protected and representative of the local culture. We go through a process of using diffusion, using diffusion in order to explore ideas alongside the heritage database and contemporary prompting as well. And of course, just with a simple sketch of an idea that we have, we can convert that quickly to AI to get materiality and that eventually leads to Us constructing the 3D model and the final rendering. And as a result, it's a completely intentional process where we really create final results and here you can see everything is completely done through AI rendering. And of course, we use uh Chat CPT as part of a process to guide us throughout the entire creative process and the uh design to construction process. And AI is starting to really seep into everything we do. It's become an integral process both from client conversations to architects, to local consultants, engineers. And so here's a video that we've created. At the end of concept phase. Everything here is completely generated through artificial intelligence, but through our creative design process. And of course, we also work with. Optimization algorithms driven by AI in order to enhance results. And of course, we explore rapidly different ideas very quickly through AI. And so we're really able to explore the depth of what's possible very quickly. And as a result, we get something that is Contextual to the sites, subtle, without being too pronounced, but at the same time when you're inside the space, you feel the sort of awe inspiring moments and heights. And AI allowed us to do all of that. I would say that particular video, this whole design process would take usually months and months, but we can cut it down to even a month or within a month if we're able to, you know, get every uh technical parameters accurate. And so it changes a lot for us. The whole process going from ideation to optimization and bin modeling process, uh, to documentation and perhaps automated construction processes. I think AI can contribute in so many ways. For example, we also develop tools on the side like working with Nvidia. We've developed this prototype where you can use on your phone or any um app and it will create sort of retexturing filter with anything that you model, anything that you sketch in real time, and you just have to prompt the words, what you like to see instead. And of course, we explore also different uh devices. So VR devices such as Oppo Vision Pro, we incorporated um artificial intelligence, diffusion modeling as a process as part of the design. So now what we can get is perhaps at the touch of your fingertips, you are able to more rapidly explore ideas. And the ideas also evolve through text and prompting. So you are actually, you know, designing through descriptions. And that's where it also gets more fun. Architecture becomes playful. One of our staff, our senior designer, he's Actually modeling in the VR space in real time, getting a render. And then we get a results very rapidly. So you feel the space, you feel the three dimensionality, and that's what changes for us in architecture. And finally, large language models like ChatGBT, we developed tools, subset of tools where we, for example, use urban design, describing how you want the urban design layout to look and where's density, what's the percentages. We're trying to bring large language model into the technical side of design. We want to take over all the data. And of course, we also wanted to Have a way to code for us, for example, which is another side of architecture, and all of these things are new possibilities for us very recently. How we optimize everything. So, really, at the end of the day, I just want you guys to think about what are the possibilities for us. We have new design tools, we have faster delivery, perhaps going to be lower costs. We're going to be doing more projects in a smaller team. And uh you can just think about how the entire industry will change from here on. So lastly I would just like to say that um we're really just at the beginning of how AI is seeping into the design world. Our entire AEC profession is going through this rapid shift, and we just want to be part of it, and everyone should look out where all of us are going to evolve in the next decade or so. Thank you. Yeah.

Want a digital economy? You’ll need AI for that. There’s no part of financial services that artificial intelligence won’t touch, from customer service to fraud prevention. In this session, we’ll discuss where AI is making the most impact in finance and tackle the ethical challenges, reputational risks, and regulatory considerations that come with it.

Modernizing the Way Money Moves

All right. So, We were in a conversation yesterday at this very event where we were talking about how Cobalt is still In existence still, we're still living our lives with this, and so much of the financial stack is still based on it. I feel like financial services is a tale of two, it's two tails, you know. On the one hand you have this aging financial stack that we cannot, we cannot move past because it would be highly disruptive. On the other hand, I know financial services companies are some of the greatest spenders of technology anywhere on the planet. Tale of 2. So, Bjorn, I want to start with you. Mastercard just made a new made an announcement, a new thing that is as far from cobalt as I can imagine. Can you tell us about it? Yeah, absolutely. Um, so, so, well, firstly, we're, we're, uh, Convinced that we are on the eve of a paradigm shift in commerce. I know that sounds a bit dramatic, but actually we believe that as we think AI is one of these big technology shifts like we've in our lifetimes only seen a few of it, and it will have a very profound impact on how consumers shop and pay for things and and uh. Today we're already seeing consumers massively moving from the previous browser search and discovery from products to move to AI platform for search and discovery, and once you've searched and discovered the product you want to buy, it's only natural that you then ask the AI agents to buy it on your behalf. We are deeply, deeply convinced that that will reduce toil for the consumer, accelerate things, make things simpler so that consumers can focus on what they like most and not having to crawl the internet to find what it is that they need, but it will only happen if you retain trust in the system and, and, and we've heard probably over the next, the last few days like, uh, these agents nobody controls them, you know, and so, so the big problem statements that we put on our teams is how do we ensure That the consumer Not necessarily the people here in the room, but my mother can trust the way that they shop through an AI agent, and what we've announced last week is essentially the framework in which consumers, banks, agents, merchants, and the whole ecosystem can work together for a High trust system and a high um consumer convenience system and that's about Making sure there's transparency that everybody knows that an agent is in the middle of a transaction and that therefore they're not surprised by that, making sure that there's security that the consumer is very clearly consenting and authenticated at time of a transaction and that the and it's recorded what they're actually instructing the agents to do. That's about having an audit trail of the chat and so that when something goes wrong, we can feedback what was actually agreed between the consumer and the agent, and that's also about them having consumer protections so there is recourse that with partners like AI and everywhere else in the world is an ecosystem place. And that's what we've announced last last week in order to give the parting shot to what I think the era of agentic commerce. Yes, and we're talking about agents constantly here. That's, that's absolutely the theme. So it's called Agentic pay. Um, and I, you know, my, my first thought when I, when I saw this was, well, who Who in financial services, who in banking, who at the end at the customer level is going, I really want an agent, you know, they're not doing that, right? They're just trying to solve a problem. Carlos, what problems do your customers have that this technology could possibly solve? Yeah, so our customers are, uh, large brands, um, stores such as Spotify, eBay, and they're looking for frictionless shopping experiences for good customers, us. They're looking to lower costs where possible. The financial infrastructure still is expensive in this economy. So, yeah, absolutely. And then the last, um, yeah, and, and growing problem is of course fraud management and all the costs that come with that and this whole discussion around agents. The true question is, what's happening with the digital storefront of our merchants? Is there really a world where we still, you know, go to a website of a brand? Order a pair of shoes and have that experience ourselves, or an agent take care of this effectively, completely, you know, taking that experience away. I don't have the answer. I think we're all trying to figure this out. Is this an addition? Is this another flow to payments and online commerce? Will this will this exist coexisting with e-commerce? Those are the questions our customers today have. The problem to solve remains balancing friction. And the fraud, that's right, main problem. And let's talk about that a little bit as intriguing as the as the the storefront thing is, my mind reels already with you saying that. Um, let's talk about fraud. So you can't engage in financial services without having to worry about fraud. There's um there's classic payments fraud, there's authentication problems, um what can AI do now? Uh, uh, to help with that, or is it just simply an escalating arms race, the latest technology, uh, the same problem? I think. Fraud, fraud management, fraud fighting is the classic use case in business school. It's been for the last 20 years when it comes to machine learning, a form of AI. It's just that the technology has reached a certain maturity that, you know, for example, for the company that I work for, we are really able in real time inference in real time to make. Very complex predictions around whether we're dealing with a good customer or a bad customer, and that has changed the game. So it's the skill of everything and in this case the access to billions, billions of data points in a couple of hundreds of milliseconds. That is mind blowing and it has changed tremendously the way we fight fraud. We see customers. Again we payments, right, our world, it's the the final step after all that work in awareness, conversion, a spend, you'll name it, that's very final step when it comes to the purchase. Can't go wrong. Back in the days, customers configured business rules when from country X at this hour from these type of banks blocked the transaction. That's pretty bad. You can imagine in terms of conversion. Now we're giving control a little bit more to an AI, which supposedly has, you know, higher precision when it comes to these decisions. Everyone welcomes that, but of course again that you're giving up control a little bit and sure there are tools for explainability whatsoever, but this is on the minds of merchants how to find this delicate balance between knowing what's going on versus Truly embracing all this cool technology to solve this problem. Absolutely. Let's see how many people in the audience who are, who here is concerned about fraud for your particular business. Raise your hand. Let's see if anybody's, yeah, all right. I think that 3% of we, we saw, I think a study at MRC it's, it's in our industry, 3% of revenues, um. The cost of fraud, yeah, average.jorn, Mastercard has a global footprint. You are concerned about this on a very macro scale. Um, how do you look at the application of AI in this area? So, so I, I, um, first kind of go, go back for a second to this forefront thing because I think but I think it's may be useful to just bring this to life for a second, uh, for people because it sounds a bit strange like, OK. When I, when I today I, I like to run when I go and buy a pair of running shoes, I go to a certain sites. I kind of find brows my way through all the shoes and there's an overload of shoes, of course, and, and it takes me a while and then I don't know if I have the right shoes from the right brand with the right good deal because I like a good deal like the any one of us. And it's laborious. When I go to an agent, I say, Look, you know, I'm a runner, I do long distance. I want comfort over performance. I'm slightly overpruating and therefore need stability shoes, and I like a good deal. And then the agent will come back and says, Oh well, here I have a Hooka stability shoe with great reviews, but I have an A6 with a 20% off at that store in Manchester. Do you want me to make a purchase? I said, yeah, sure, like get me that A6 good deal. That's what I wanted. I would never have found a good deal in that store in Manchester. I never go to that website. I go to my agent and, and I think suddenly I have. Gotten a good deal and, and, and, and, and avoided avoided an hour of my time to waste. Uh, and, and so that is why we believe that this will actually turn in a very meaningful way, not for everything, but for quite a few soil kind of usage. But, but, but to the point that we made. You need to make sure that this is a real agent, not a rogue agent, right? And so the first thing that we do here is Identify and certifying agent and then the second thing that we do is when an agent has been certified as being a valid agent, what we give to that agent is something called tokens or agentic tokens. What this means is my credential as a consumer. is translated into a proxy for that credential with some very clever cryptography that we give to that agent that we bound to that agent. When a transaction is made at that store in Manchester, that credential comes back to us. We say, does that credential come from the right agent? Has it been authenticated by Jorn because it was his credential, and only then do we pass it back to the issuer. And by doing that tokenization, even if a bad guy gets their hands on that credential, we can bind it back to its origin. That is by architecting things in such a way. that we can at global scale. Avoid that that that um that uh the bad actors get to it. The issue remains the fundamental issue remains scammers because if I was the scammers, isn't it, if I as I lured in to make a transaction because I believe some kind of scammer did something and I authenticate a transaction, that's very hard. I mean, getting. Getting that so is very hard, and then that's where the cat and mouse comes in. They have great tools, we have great tools, and we'll see forever. I think that that cat and mouse, and we just need to make sure that we run a little faster than the other guy. Indeed, indeed, it's just like the cybersecurity uh category. I, I want to come to the audience for questions about financial services, um, but first, I just want to press on something that we touched on, and it's trust, um. You know, I can't imagine too many customers, whether they're B2B or B2C, are terribly concerned about how the transaction happens so long as it happens to their expectation. Is that right? They don't really have to concern themselves with trusting an agent because it'll just work to their expectations. Is that kind of the mentality in this category? Yes, it will have to work to the expectations and they need to have recourse. If, if the wrong shoes come through, I want to say no, no, no, this is not my size or this is not what I wanted and I need to have recourse against that and this consumer protection. It is extremely important because if the consumer feels like, uh, something bad happens and I am left on my own devices, then, then this whole thing will collapse like a card house, right? So very, very important that we embed this into that all the trail points and the point that the consumer will get the money back. If something goes wrong, it's very fundamental in this construct, right, right. Fair point. All right, questions for either gentleman about What they do. Oh, we've blanked them. We've stunned them. OK. OK. Oh, there we go. Great. Thank you. Just wait for the mic and then your name and organization, please. Forgive me, there's a light right behind you. Hello, hi. I'm Doctor Shiva from my bank and very interesting discussion. Um, I don't know how many people are aware of it now. There are a lot of, you know, duplicate stocks like they call Nike and all the different brands, but genuinely people believe that these are coming from that particular brand, but all like, you know, coming from wrong manufacturers, Taiwan or China. You don't get the real quality, but the money is gone. This a lot of fake websites and if you call them, nobody even picking up. It's all going into the automation thing now. Even happened to me, you know, I paid something and then it immediately taken another 400 pounds from Mastercard. But then, you know, struggle, we couldn't even get the money. So how can we solve that because everything is now going into AI and matching. Yeah, so, so, so, well, you should be able to get your money back, and the reason why you should be able to your money back is because players like Arian. Which are requires and PSBs, they on board the merchant in the ecosystem and they underwrite that merchant. They ensure that it's a legitimate merchant, very, very important task that they do. And when that merchant is on boarded, then they take responsibility that this is a good one. So if you get the wrong shoes or whatever, you got the wrong uh or the wrong Nike uh equipment, you can go back to your bank and say this is the wrong thing. The bank will come to us. We will go through Ain and say, Ain, your merchant is not legitimate or doing something wrong, and Arian will take responsibility for either correcting or shutting it down. That is how that ecosystem works, and the role of somebody like Aen is extremely important in making sure that the right folks are on boarded, and that's why that's that that is the the the magic of this worldwide ecosystem to make sure that everybody plays a role and that you get can get your money back as you. If you didn't manage to, then, then I think you have to speak to your bank because you should be able to. Here we go. Further questions. All right, gentlemen, I have another one for you. Um, You know, when I think about Financial services, I think about access, right? Especially on a global level. Um, there are plenty of people who, who don't have it. What element, if any, does the agentic aspect add to that? Will it reinforce bias? prevent access? Will it not really have any impact at all? Is there any new dynamic to that? Well, I, I, I have a, a perspective and, and, and I'm sure, uh, Carlo as well. So, so I, I think where the, the, the challenge with financial inclusion is at its heart one of identity. A consumer who does not have a formal identity, a passport, and identity card, cannot open a bank account, cannot be KYC, therefore cannot have access to electronic payments and is stuck in a, in a cash economy. That is a problem that across the world, governments are working with the likes of ourselves to to solve this and as it gets solved, they get embordered in the financial ecosystem. And, and, and that will then include their access to agentic payment if that happens. I think AI can help to some extent in getting the graph of that consumer, uh, distilling it that to to an identity because that consumer even without. A bank account probably has some level of access to internet or to to to to digital services, but that's kind of where it can help. But the fundamental issue is one of, of establishing identity and from there the rest flows. I think there will be a massive shift from Today's world operates in a real-time static verification check, right? I have to present my passport. This check is happening, and everything will shift to a behavior-based, you know, real-time monitoring behind the scenes to try and establish is this actor. Transacting or acting as expected, and here the scary part is that's where AI really comes in. Our behavior just looks like so many others, and these predictions that we continuously make, they will hopefully also open the doors to more financial inclusion simply because we do not have these strict and and very static checks sometimes at. Gate which might discriminate or or create problems at the at the gate for uh for good customers. Got it. Got it. All right, we're in the home stretch here, so lightning round for you. Uh, it's really just a yes or no, but I'll keep it simple. Um, will the application of AI in financial services increase spending? Yes or no? Yes. Yes. All right, there you have it. Thank you, gentlemen. A round of applause, please.

As AI moves from the confinements of screens and into everyday life, the big question is: What role will embodied AI play in the AI revolution? In this session, we’ll explore how Amazon Robotics is leading the charge in redefining human–machine collaboration and how it will impact warehouse automation, logistics, and beyond.

Amazon Robotics: The Present and Future of Embodied AI

The next wave of AI is about action, not just content. Multi-agent systems will perform tasks across the enterprise, while specialized mini-models will reduce costs. As AI becomes more autonomous, how do humans remain in control? And what does this shift mean for businesses, tech suppliers, and ideals around responsible AI and governance?

AI Beyond Language Models: What Comes Next?

When no one was looking, the global semiconductor business became a smash-mouth, no-tariffs-barred affair. But in an age of AI, business is still (very) good. The founder and CEO of one of SoftBank's newest multibillion-dollar acquisitions weighs in on the future of AI compute, the changing cast of industry power players, and the shifting geopolitical landscape ahead.

Iron in the Fire

Peter and Daniel. So there's a lot of excitement about what AI can do, right? That's why we're here. But none of it's possible without the right infrastructure and the right chips. The competition to build the next generation of AI hardware isn't just a business story, it's a strategic one touching on everything from investment and innovation to geopolitics. Today we're joined by one of the most respected leaders in technology and semiconductors to help us understand what's at stake and where the future is heading. So please welcome Renee James, chairman and CEO of Ampere. Hi Renee. Good to see you. All right. Thank you for joining us. So You worked at Intel for a very long time, um, and you worked with the great Andy Grove. You were for a long time, you were first as chief of staff, is that right? And then you, you know, right before you left Intel, you were the president of Intel. So, When you look at that, I promise I will ask you questions about your company Ape, which was acquired by SoftBank for $6.5 billion earlier this year, but I promise we'll get there, but I have to know as an, as an Intel vet. What would Andy Groves say about the state of Intel today? Thanks. Yes, um. He would be very disappointed. Why? Because it's an industry that Intel um invented literally invented the means by which you fabricate modern semiconductors, um, the what we think of as fabs that everyone likes to talk about and weirdly wants to build everywhere. Uh, and the process that goes into them was pioneered by Intel, developed by Intel. The modern transistor technology was invented by Intel, um, the first modern CPU that powered, you know, computing and the entire PC mobile, everything. Was Intel. So when you think about that, and these were people who left Fairchild because they had a different view of how the future could be and started with venture capital, if you can believe it, in the late 60s, 6869, this is, you know, their life's work and then to see Intel fall behind. Um, a company that was a great partner, has been a great partner, still is, I think of Intel forever, my supplier TSMC was started in collaboration with Intel, um, because we needed an additional source. So you look at it and you think, wow. That's kind of scary and sad. And Intel is not that sort of enzymatic uh uh uh actor that it was in in its heyday is, is kind of what you're inferring. And a lot of the technology that we use today, I mean, the Mellanox interconnect technology of Nvidia was invented at Intel and spun out there, you know, my company, uh, not just me but several of the key leaders, we're X Intel, we built every server product that built every data center. it started, you know, all of this cloud stuff. So, you know, a lot of good came out of it and a lot of spinoff companies, you know, the modern way that we think about building silicon and semiconductors that we all take for granted like a miracle occurs. You turn something on and it just works and like, you know, 70 billion of something you don't even know what they are is going on and on a little piece of silicon and we don't know, you know, it just works it it truly is miraculous, but then you think about the state of the company today and And I, I wish them recovery. I really do. I think they're super important to global semiconductors. They're super important to the future of where we're going because at its base Intel is the company that invents the future of transistor. still today, even in their current situation, super important. We will come back to the sort of global implications of that in a second, but I do want to, I do want to talk about Amper for a second. So Renee, you're one of two women who run semiconductor companies. It's extraordinary. And then you I'm happy that there's someone else now. That's right. That's right. And you went even further because you said, I'm going to start one. Like that isn't exactly how that went, but yeah, I think Andy said, what are you doing with your life? Why don't you start the next one? So what what what leads one to start a semiconductor company because I've always gotten the sense that um sort of like an automaker, you can't just turn around and it's not exactly a software company, you know, you can't just spin it up overnight. You need a lot of capital. Yes, um, well, so first of all you have to have a passion and a vision for where you think things are going because you, you know, you're not, you wouldn't start a company at the kind of the scale of a CPU company or GPU company, um, from the ground up at this point in the history of semis given the the hurdle rate and the investment rate and costs, um, unless you had a passion for doing something that hadn't been done before because you know you're not gonna start a company to like compete with Nvidia. That's ridiculous. Um, so we said the problem that we hadn't solved, the thing I was passionate about after 30 years of building microprocessors, was we have always, whenever we got into trouble we would just say like, oh, you know what we can do, we'll just put more power onto it. We can just release the power and I'll get higher performance or we'll put down more transistors. I mean, we did a lot of things that at the time it didn't matter. It didn't cost as much. It was, you know, no one was going to freak out that we went from 350 watts to 400 watts to, you know, whatever. I mean these were conscious decisions to change because you have to change the spec on what the server is and what goes into the data center and all that stuff, but more was more, more was more and you could get paid for it. Um, so the thing we didn't do was something that we knew about. I mean, good and physicists know that at some point, you know, you need to constrain the power and figure out how to improve the performance through other techniques, and that's a lot of R&D and a lot of thinking. About thinking differently, thinking differently about how we designed the architecture of X86 for the last 20 years that we had been working on it, um, before I was president until I ran R&D for a long time, so we spent a lot of time thinking about this, so we, um. I said, you know what, that's what I want to work on. I want to work on how we build out what comes next, knowing that not AI in the way that we think of it today, but we were working on machine learning and everything already. We knew that was the next big thing. We knew that cloud computing, the way that the cloud used software was different than the way that the enterprise consumed software, like in this linear scaling, always on, always on demand. Um, and, and the microprocessors that we had built were meant to turn on and off in a way to be efficient for different workloads but not for the cloud, so that was our thesis and um. I gathered around people that were retired or that had worked for me before and I said we should do this and they're all like, you've lost your mind because we had built like 80 microprocessors, 80, and they were like, really, are you sure? I'm like, we can do this. If not us, then who? And then Andy says. He was still alive and not really that helpful in the sense that he was like, we taught you everything, you know, this is what we what Bob and Gordon and I did. This is what, you know, this is, this is the history of Silicon Valley. Someone has to do it, so that's what we did and I was really fortunate because the Carlisle Group, who doesn't. Usually do venture um back to us in our first round um and then we had a couple of strategic, our largest strategic being Oracle over time um and we ended up raising billions of dollars over the course of the last 8 years and um last month announced that, um, SoftBank Group is buying the company for $6.5 billion. In cash. So let's talk, let's talk a little bit about that. So what. Some folks were scratching their heads when that deal was, was, was announced. What does SoftBank give Ape? What does Ampere give SoftBank? Um, well, of course the deal isn't closed, right, so we're in the regulatory period, but in general, um, we've pioneered, we have a lot of patents and a lot of technology we've built, um, 5 generations of product already that have been deployed into major hyper scale data centers, um, how to build very, very high performance in an air cooled, meaning 500 watts or less envelope, um, and we could release the power and use liquid cooling. And go even, but the performance is that of like an AMD leading edge processor today we have integrated AI acceleration inside of that envelope. We can add additional functionality on and only release the power, you know, 50, 100 watts. So I think that that whole thing allows you to build out a vision of AI, you know, scale data centers that are more efficient over time. Um, we were, we are an arm architecture licensee, so a sister company in the SoftBank Group, um, so we had taken an architecture that's inherently low power but had not been known for the highest performance, and we had, um, invented ways to bring it up to. The level of, you know, that kind of performance, and I think that it's very synergistic with their overall vision and the mission that Masasan has for Stargate. For sure, for sure, and, and we will get to that in a second. I, I wanna ask you, I want to come back to the, the geopolitics question so. When you and I were talking about Intel previously, um, you were expressing to me kind of the importance of that company in the overall ecosystem, um, honestly, in the US, but, but in as as outdated the terms as it is with the West. Um, we are in a very US versus China moment in time right now and you're very concerned. Um, what is the future of SEI right now and What's it look like if we don't get it right here in the UK? In the UK, um, what do you mean by don't get it right in the UK? What, what is the landscape, right? What is the playing field given these tensions and you know what, how do we remain competitive given that sort of binary choice, you know, semiconductors historically have been a global product that have been available globally, um. That you know the predominance of semi semiconductor business is distributed around the world both in supply chain and design. Um, some of the great design teams in different segments of semiconductors are in Europe, and many of the industrial companies in Europe do their own semiconductors, so super important. Um, the predominant and and currently an EUV only lithography equipment is from Europe, so I think we have a long history of a very open design design and manufacturing ecosystem. I, I do think that there is, you know. An importance in the global nature that this business has always been, it was never just a US business that went global. It was global from day one. So I think we have to think about where the people are who do this, how, how the, you know, how it moves, how it affects every segment of the economy in all these different countries, because computing is not just about people doing office work anymore, it's how we live our lives. It's everywhere, right? So, um, I think as long as we're conscientious about open market and and how we move forward on that I think that's important. The China thing is a separate situation because China has capabilities that are self-contained and to some extent I don't know what direction that will go it, you know, I have always participated in a global market, so it's hard for me to speculate on what that would look like, but it's the US and. Um, Taiwan, Southeast Asia, European, you know, design capability for global semiconductors is profound. I mean, it is hard to imagine doing, you know, some level of networking and mobility and everything without teams in France, for example, right? Like there's just specialties around the world, Poland. Right, right, I will come to the audience for a question in just a moment. I wanna just quickly ask you to tell me, you know, does AP have a role in the Stargate project? And if so, what? Um, well, I can't comment too much on that, um, but as we said in the press release, we, we do have a role in that, so yeah. All right. All right, let's go to the audience for questions. Questions for Renee James. They're all wishing they were in semiconductors, like, really? All right, well, I have, oh yes, right here up in the front, just wait, wait for the mic to reach you. Way up here. It's on the couch. And as always, your name and who you're with when it comes to you, OK? Thank you. I'm Doctor Shiva, I bank, and like I said yesterday, we're building a data center and also I'm from Sri Lanka. I don't know whether you all know Bain graphite, the pure graphite is in Sri Lanka and only 1% is exported. And now there are people looking at graphene, which is the GPU. Uh, you know, for graphene chips, you need graphite. Is there any, you know, experiments that you are doing or what's the plan? Yeah, so we don't do base level. Um, Ampere is a fabulous semiconductor company unlike Intel. We don't do um base level metallurgic, um, material science at MPR. We do design, advanced design and IP um development, um. So I am sure there are people looking at that, but most of that work is done in process fabrication, um, in conjunction with that. So at TSMC and Intel and places like that. Yeah. We have a question. Thank you for your question. We have one right back here, yes, your name and who you're with. Uh, I like, yes a business school. So I had a question about the overall ecosystem around CPUs and GPUs. So where do you see this going? Nvidia is now very dominant in the GPUs. Um, you know, X86 is still powerful in the CPUs. Where, where do you see this in 5 years, you know, Nvidia, ARM, everything. Yeah, thanks for the question. Um, in the, in the, you know. GPUs, as you know, grew out of gaming. So a large portion of the GPU, you know, the shaders and the things you need for rendering and gaming aren't used to um train large language models. So, Um, while it's a lot of horsepower, it isn't the only way that you can do the computation required for AI. And in fact, um, as we've discussed in multiple roundtables here, um, as we go to smaller models, as we use, let me say the appropriate AI technology or machine learning technology for the job at hand, you might just run it on an enhanced CPU. In fact, that's probably the most efficient way. So all the CPUs are being enhanced with AI acceleration, and we have a product that we've talked about Ampeer1 Aurora, that actually has um the AI functionality because we have 512 cores you can put, you know, AI cores and compute cores and into that's one of the great innovations that we've been working on. Um, so I think that 5 years from now what we're going to see is a convergence of AI and compute and architecturally there's an underlying thing that is super insular, uh, Silicon Valley Wars kind of thing which is the micro architecture of computing. Will it be X86 or will it be something else? Um, Nvidia is an ARM licensee. Nvidia builds their things on ARM as well. I do think given the progress we've made and now under our new ownership that um our proposed ownership that that arm is looking like a long term it's the architecture that Apple uses they're in a licensee that Samsung uses that Qualcomm uses that Broadcom uses. And um because of the developer community being on phones, laptops, you see, you get this bulk of developers that know that environment so I, I, I think the future is relatively bright there. And of course, SoftBank owns arm. Yes, right here. Hi Seth Dober in Quantum AI and Infinity Adventures. Hi Renee. Um, so one of the biggest challenges in, in AI, especially training is not necessarily compute, it's the distance between compute and memory. Are you guys doing anything in that space? That's my favorite topic. Oh my God. See, I, yesterday at lunch I was like, I love that guy. I, uh. Yes, Seth, of course we're working on that problem that we've had for the last 30 years. Um, if it exists, if it, if it was a problem in mainframes, it's still a problem. Um, of course there's high speed memories, but one of the things, of course I, I'm not gonna reveal how we're doing it, but all of us have been working on moving memory closer, staging memories, uh, flexible different kinds of memories that you might want to put in and out depending on costs, so there's a bunch of things that we're doing to, um. Deal with moving the workload in and out more efficiently. It's not just memory, it's a lot of the die to die interconnect as well, as you know, right? So we've done a lot of innovation. We have patents in it and die to die and how you move things across, but you know, the other thing that's quite fascinating only to me but probably not the rest of this room, but maybe to you, the way you lay out the chip where you put computation versus IO actually matters because it determines some of the. Length of travel and so we've been very attentive to that as well. Thank you for your question. I feel like I saw one more. Yes, was just my memory. I wanna talk about semiconductors. Come on now. Oh yes, right up here. Oh yes, your co-host. Yes, that's right. Just wait for it right here, Maddie. Thank you. Hi Renee, Jason Del Ray from Fortune. Um, I spent a lot of time, probably too much time thinking and reporting about Amazon, and I'm curious what you see as their potential future in the chip space. Um, I know they've made a lot of investments and are you bullish, bearish on them in the future, and either way, where, where do you see them fitting in? Yeah, so, um. Thank you for that question. It's, it's Amazon uniquely has an outstanding team that they've had together for years, and one of the hallmarks of a successful semiconductors is how many generations has a team done together, and the answer to Amazon is a lot, and they also use ARM architecture and have been quite successful in building chips, as has Google, by the way, um, that are specific to the pathology. They have a really unique way of accelerating data inside their own network and they're and they have an outstanding silicon team that's been able to build, you know, really great products that for their own use. I think given you know their their might and their team and everything that they'll continue to be very successful with that strategy. I, you know, the hyper scalers have the wherewithal to. Do that kind of thing, but the other half of the market is the enterprises of the world that need to use AI that don't have it, and that's where I think, you know, commercial chips have to fill that gap and they're not all gonna build out massive Nvidia infrastructure either, right? So there's still an enormous market opportunity and let us not forget, despite the fact that we never talk about it. Pure computational data, non-AI is growing as well at a double-digit kegger. So there's kind of a, there's other stuff going on as well. Thank you. Everyone, the fantastic Renee James.

Brainstorm AI London 2025

The Future of AI Regulation in the U.K. and Europe

Latest video

Videos (23 videos)

The Future of AI Regulation in the U.K. and Europe

AI Continues to Eat the World

Betting Big On AI—Europe vs. The U.S.

Google Ventures managing partner says generative AI could become "a utility" for investors

Smarter, Faster, and More Personalized: How E-Commerce Giants Are Staying Ahead

Accenture Announcement: Human + AI Impact Initiative

Diagnosing Genetic Diseases Faster

The Future of AI-Powered Business Travel

EVERYDAY AI: BEYOND THE BUZZWORD

Hire, Fire...Or Reskill: Leading Through AI's Talent Revolution

Making Stuff Up: Improvising LLMs and Spontaneous Robots

Who’s Protecting Our Kids from AI?

AI Driver: The Next Era of Transportation

The Next Frontier of Open-Source

The Race to AGI

Can You Have Your AI and Decarbonization Too?

The Future of Video and Storytelling

Free the Lawyers!

The Architect's New Toolkit

Modernizing the Way Money Moves

Amazon Robotics: The Present and Future of Embodied AI

AI Beyond Language Models: What Comes Next?

Iron in the Fire