ThinkBusiness.ie
ThinkBusiness.ie, powered by Bank of Ireland, has been created for Irish business owners and managers who are seeking information, resources and help on a range of business topics. It provides practical, actionable information and guidance on starting, growing and running a business. It offers guides, tools, templates, checklists and a wealth of other content tailored to meet the needs of Irish business owners and managers. It also showcases case studies, with insights, tips and advice from Irish start-up entrepreneurs and business owners. Bank of Ireland is regulated by the Central Bank of Ireland.URL: www.thinkbusinessTwitter: @thinkbusinessieLinkedIn: ThinkBusiness.ieFacebook: ThinkBusiness.ie
ThinkBusiness.ie
Episode 354 - Tim Grant, executive chair, TensorX
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Artificial intelligence is transforming how businesses operate, but as adoption accelerates, questions around data control, privacy and digital sovereignty are becoming increasingly important.
My guest today is Tim Grant, Executive Chair of TensorX, the Irish company building sovereign AI infrastructure for Europe following an €8m funding round. We'll discuss why businesses want their AI data to stay within European jurisdiction, what digital sovereignty really means, and whether Europe can build its own trusted AI future.
Visit www.thinkbusiness.ie for more news and supports for start-ups and SMEs in Ireland. If you want to start and grow a business, ThinkBusiness.
Hello and welcome back to the Lady's Think Business podcast powered by Bank of Ireland. Artificial intelligence is transforming how businesses operate, but as adoption accelerates, questions around data control, privacy, and digital sovereignty are becoming increasingly important. My guest today is Tim Grant, Executive Chair of TensorX, the Irish company building sovereign AI infrastructure for Europe following 8 million euro funding rounds. We'll discuss why businesses want their AI data to stay within European jurisdiction, what digital sovereignty really means, and whether Europe can build its own trusted AI future. Tim, great to talk to you. About a month or so ago, we wrote about an AI, Irish AI infrastructure business called TensorX, that launched with an 8 million euro seed funding round led by Darius Cubed Ventures. Now, the interesting thing about it was uh first of all, it's an infrastructure company providing private sovereign inference on dedicated NVIDIA GPOs. Now that sounds very talk complicated, but basically what it boils down to it's um it it's basically AI future infrastructure that isn't sitting in the vaults of some some uh let's say US tech joint. It seems to be uh the location of the infrastructure is in Dublin and Helsinki. Um can you first of all tell me a bit about why TensorX was established and what it proposes to do. Um the thing that the idea I'm getting completely is that it it gives businesses an opportunity to host their data and applications in a non-sovereign environment. Is that fair?
SPEAKER_01Yeah, exactly, exactly. And uh look, I think the the the story of of sovereignty is now, I think we've we've we've fortunately we've hit a moment, fortunately for our business, we've hit a moment where it's becoming the talk of the town. Everybody wants to talk about certain uh about sovereignty, and we'll get into that and what that means. But I think to to really understand the story of TensorX, we've actually got to go back a year when we came up with the idea. And we we so the the context here is is we aren't uh just some founders who've who've gotten some capital and and are building a company. We're actually part of something bigger. We build companies, um, regularly build and invest in companies. So um I'm I'm my uh my day job is is CEO of Deus X Capital. We run a very significantly sized fund which invests and builds and operates. And one of those companies that we've invested in is called Apex E3. Uh Apex has been around for several years, and and Apex was the reason that we realized that TensorX, lots of X is floating around, um, that TensorX would be needed. Uh because we and Apex, and I'm chairman of that company as well, we work with very large institutional asset managers in the financial services space. Um, and if I was to mention just one of those names, Vanguard, uh, which is the second biggest asset manager in the world and who we've been working with for several years, you'll get the scale that we're operating at here. Um and we established a year ago um when we were working on deploying AI solutions to that kind of cohort, that we were very quickly going to uh hit a hit a brick wall when it came to the data and where it was going and who could see it and how that would fly with the policies uh and procedures that exist at these large asset managers. Um and so it was a prescient call, I think we can agree a year ago for us to see that it was needed. But the fortunate element of it is that we we we we had a need right there and then. And so we didn't want to go building for the likes of Vanguard um without a sovereign solution because it was just so evident. Um and and by the way, I come from financial services as a career. So it I've been conditioned, and everybody in financial services has been conditioned to know that data, data retention is incredibly important. It's part of regulation, um, which could be both financial services regulation or something like GDPR in in the Europe, where where you actually have uh you know data regulation about where you know what you have to do with the data. So there's nothing new to financial services about this, but what is new is AI. So it took a minute for I think financial services to wake up and realize that it was going to need sovereign. And I think we were quite far ahead of it, John. I think we we saw it early, we started investing uh last year, and now here we are. As you mentioned, we've got our uh we we did our 8 million uh euro raise, which we've used very simply to get a great team in place and to buy and deploy GPUs and then build the layer around that that would allow our clients um to to go and and and come and use our service and actually achieve that that sovereign capability that they wanted. And in fact, uh Apex is it was our first client uh because Apex has clients who need this now. So we've we've we've actually gone from the the kernel of the idea to the deployment and the and the actual execution of the idea uh over that year period. But yeah, John, that's where it came from. That's why we've turned up because it just seems so obvious that this is going to be non-negotiable.
SPEAKER_00And and are you building or have are you part of a data center for this? Or how how does the physical I mean obviously the physical chips are wanting, but are these stored in a data center, or how do you how do you go about um preparing this uh non-su well this sorry, I suppose how do you say it, sovereign uh compute capability?
SPEAKER_01Sovereign compute, yeah. I mean, look, the the the critical, and you mentioned uh it the idea, I think, of of you know machines buried in large American companies, right? You know, and and let's just sort of call it out. I think you've got the you've got the classic clouds, which we're all familiar with, um, you know, Amazon and Google, etc. And now you've got this new term that's been floating around called neo clouds. And and I think it's worth, it's actually worth using that terminology if we're going to call cloud the the sort of 20th century uh version of that. Um and neo cloud is the 21st century, the more recent niche players that have cropped up, like like they they have a very different kind of positioning in the market. And what's really fascinating about the the evolution of cloud, and again, I can say this with confidence because I've seen this play out in financial services over the last couple of decades, is that what seemed at one point to those big financial services players a really smart efficiency play. I'll go, instead of having my my own infrastructure, I'm gonna go and outsource it to these guys, um, and it will be cost effective. Well, that was right at one point, but there's one thing you can be certain of with big uh you know monopolies or duopolies or you know, these large companies is they're gonna jack the prices up over time. And now what you've got is a capitulation that has started. And by the way, everything I'm talking about so far here is nothing to do with AI. This is just cloud in general. Now, large financial institutions are balking at this because the prices have gone up and it's starting to get a little out of hand because, well, they've got all your business now. You've got all your systems operating on their, on their on their cloud in certain ways. And so that is, I think, starting to make everybody think twice about going to the cloud, um, the classic cloud. Um, now, the classic cloud does have classic non uh, you know, the machines that really wouldn't use for AI, actually, you need don't, you know, you specifically need GPUs for that. But they do, they do offer GPUs, but they tend to be very, very expensive. Uh, they tend to be uh non-sovereign, um, i.e., you're, you know, you don't know where the data center is, you could be mixed up with a whole bunch of other people. You you can't, you can't, you really can't get that level of auditability on what's going on with a classic cloud setup. So people are starting to look at something else for that. Enter the neo clouds. And sure, you've got now a lot of players like us around the world who've cropped up. It's fascinating to see this new type of player come in, which is really focused uh very uniquely on GPUs. It's it's it's not uh it doesn't have a decade or two of history of building infrastructure on an old model. It's it's building it for what people need today. An interesting entrant into that market, um, and this is again something we understand, is actually from um from crypto mining. You know, there's a huge, huge amount of money was spent on building digital asset crypto mining capabilities all around the world, ended up being in the US because the Chinese um actually um uh banned it at one point. So the US is now the biggest concentration of crypto miners, and that requires pretty much the same sorts of things you need if you want to run GPUs. You need a data center, you need cheap electricity, you need to be able to cool them, and you need to be able to know how to run that sort of machinery. So actually, loads and loads of crypto mining companies, who some of them are very reputable and listed on the Nasdaq and things like that, have wholesale shifted their entire model to AI instead because the the economics of crypto mining has become incrementally less interesting. So they made their money and now they've completely replatformed uh their in uh their entire data centers to do that. And that's the neo clouds. And what they're selling is just my GPU capability or inference. And then I think we have to go further. Is it the GP? Are you looking to use the GPU or do you actually want to use inference? There's a subtlety there. Um, but ultimately that's that's the big picture story. So we're a neo cloud, but we're a neo cloud that was built from day one for sovereignty as the end game, right? So we don't have to go back and change what we've done to make it sovereign. We decided to build it sovereign out of the gate. That requires uh you know a different approach to make it so, to have zero data persistence. If you want to be sure that your data simply does not persist in any way, shape, or form in any models, um, anywhere in the process, you've got to be able to audit that and demonstrate that, which is precisely what we have. Now, where the machines go is the interesting question. We have we we are too small to go and buy a dedicated place and actually have our own in our own data center. The good news is there's a lot of very, very capable data centers out there who allow us to put our machines in their racks and we have complete control of them. 100% control, all the control you would need is effectively if they were in our own data center. Um, and we're using uh using one in Dublin for that, and we have access to others around Europe. Now, the long-term goal is though going to be having our own data centers over time. That's kind of the the longer term multi-year strategy. But in the short term, we're going to have, and this is now getting us sort of into what does sovereignty really mean? Well, one piece of sovereignty is that the machines are running in the jurisdiction that you need them to run for whatever the reason it is you need them to run there. Um, and so the the the the Dublin location is actually great for us because it's it busts open the entirety of the EU. We can now service any client anywhere in the EU, and we can demonstrate end-to-end that the that the hardware is in in Dublin, uh that the software is run only in Dublin, and that all of the data is not persisted, and that when all is said and done, we can't see their data, we don't use their data, the models aren't trained on their data uh uh unless they're doing something that they require that for. And and and then we allow that to happen. So it's uh it that that's the story of kind of where you put them and why you put them there, John.
SPEAKER_00Okay, so so what you're doing is um you're using data centers, but you're not fixed on any one. Uh you've got partnerships that and but ultimately the ambition is to have your own eventually. But what you what you can probably do is deploy according to client needs. So this person needs this jurisdiction, needs this capability, but the source of the power is the neo clouds and where that are connected that that that connect to all these different uh NVIA GPUs, GPUs.
SPEAKER_01Yes, exactly, exactly. And and again, the critical element is show me audit to me that that uh that I truly have AI sovereignty. Because back to back to the motivation, um if and again, I'll give I'll give an example of a of a highly regulated financial services institution.
SPEAKER_00So if you're in the if I was gonna ask you like to define what sovereign AI means in practical terms and why a company's CEO or board should care about it.
SPEAKER_01Well, I think um if you're a fan if you're a regulator, and this uh there are multiple um uh types of of community and business sectors that have this, and and the two obvious ones that everybody use, and because because we can all relate to them, are financial services and healthcare. Um but it's but that go it goes way beyond that. There are many, many companies, and I I I'd actually I would extend this to a more generalized case of of who would want their data to go and be taken into anything uh that they didn't know is is a separate discussion. Maybe we could pivot to that in a bit. Like we, you know, we look at the history of Google and Facebook, etc., where we all turns out we all gave them a lot of data for free, we didn't realize. But um, but but but ultimately, you know, what is sovereignty? Well, if I'm in financial services and healthcare, I'm already used to this. I'm already used to having very clear guidance uh in terms of where data for so me as an individual, for instance, or as or as clients can go, who can see it, uh, how long it's retained for, uh who can see it when they're asked to buy it. Um it's it's a pretty complex web that we've built. And it's it's a it's an artifact of the extraordinary um of the extraordinary ability now to move data around everywhere uh through the internet that's happened over the last 25 years. So so this is a this is this is entirely understood. The idea is we've never really called it data sovereignty. I suppose that's the term that's that's kind of evolved and AI, sovereign AI, but like sovereignty of data is has been around forever. And and it's not, you know, especially in the last decade or so, it's getting more and more uh regulated and more specific what you need to do. So that's that's really the key. What I need to know where my data goes, who can see it, that it's not persisted somewhere where I don't know that it is, and I need to be able to audit that in case someone who I riss who I'm responsible to asks me that question.
SPEAKER_00We hope you were enjoying this podcast. Bank of Ireland is welcoming new customers every day, funding investments, working capital, and expansions across multiple sectors. To learn more, Google Bank of Ireland Business Banking. Bank of Ireland is licensed by the Central Bank of Ireland. Would you believe that Europe has become too dependent on US hyperscalers for critical digital infrastructure?
SPEAKER_01Well, look, I think that the the the the existence of us as an example of Penser X and the relative explosion, I think we'll we'll say over the last couple of months, um, even just the last month, uh much to much to our uh much to our delight, actually, given what we've been building, we've kind of suddenly had a bit of an explosion of discussion around this topic. And it kind of started at the point at which uh you know Claude did a release, and then there was a statement that it couldn't be used outside of the US. And so everybody just felt that sudden pain of what can happen when a decision is made by a few people in another jurisdiction that's got nothing to do with yours, and suddenly you can't use that model anymore. And I think that was the beginning of uh of a new type of dialogue. Because if I I'm now really looking at business risk. So if I'm running a business in Europe and I've got my you know, my data now I'm running, I'm using a model like Claude, and but there's the optionality for them to turn that model off when they feel like it and me not know when I get it back. That's business, that's continuity risk. Yeah, that's it.
SPEAKER_00And also if you're if you're if you're a business depending on that technology, maybe you developers are using it or something, suddenly they can't, you know, that's that's that's a problem. But the what do the way to sum up digital sovereignty, would you describe it primarily as a security issue, an economic issue, or a geopolitical issue?
SPEAKER_01Uh you can't, I don't, I think it's it in the modern age, and people say data is the new oil. I think it's all three. You know, this is a this is a strategy question if you own a business and it's data dependent. It's a strategy question if you own a business and it like it's likely to use AI at all. Uh, and let's face it, most will. Um, it's about the people that you have. You might need new people who understand this, uh, which we should talk about, about the organizational change that will happen as a result of AI. Um, it makes it very interesting for us. And when we talk to clients, it's an economic question because if you have to protect your data, you don't want to overpay for that. So now you've got a commercial decision to make. So I think it's it's it's really quite pervasive at this point. And you know, my call to arms for anyone listening to this this podcast is if you haven't in your organization already started thinking about what it could mean to be using AI in my company and and that and whether data sovereignty matters to me in how I do it, then you're probably behind the curve and you need to be having that conversation.
SPEAKER_00It's kind of fascinating because if you've watched the headlines in Ireland in particular over the last year or two, uh you see these long-running battles between US tech giants and the data protection commissioner over fines for data that's been taken from Ireland and brought to the US, etc., are aware of things like the US Cloud Act? You know, how much of a concern is that for European businesses uh in terms of risk? And you know, if Europe fails to build sovereign AI infrastructure, what do you think are the long-term consequences for European competitiveness?
SPEAKER_01I I mean, if if you look at GDPR and the effect that that has had, and and the the the the necessary evolution that every company in the EU has had to make to take that into account, uh then then you know that it matters. Um then you add the AI element, which is you know, the the extraordinary speed with which these platforms are evolving to do things that even months ago would be would be considered fantasy and would be considered science fiction. It it is it's it's extraordinary, it's mind-blowing. And therefore, you you have to assume that this is all going to be very relevant for European companies. And I think you know, the EU in this particular instance, it it's always that it's always been a little bit more focused on data and privacy than many other jurisdictions. We know that to be true. Um, certainly in the US, it's just a little bit looser. I I've spent, I'm actually of American nationality and lived there for a long time as well. It's a lot, it's it's a lot less stringent in in the in the US. So the fact that it is more stringent in the in the EU, in fact, for us, it's it's it's it's the upside, right? Because we know that every company's going should think about this. Um, and sure enough, those that are seeing huge early value from AI, but also have sensitivity in their data are already realizing now that that's what they need, and that's the clients who are coming to us today. Uh, I don't I don't think it's negotiable. And I think competitiveness is going is going to be affected if you don't get on top of this.
SPEAKER_00I was thinking a lot about Ireland in this, because as a country, we kind of have prided ourselves on the last 30 years of 30 or 40 years of managing to be at the cusp of every wave. And you know, I was just looking today on the day we were we were talking, um, I probably did one of the early years of my career uh was defined by Intel, the big tech giant, investing in Ireland uh in a significant way in the early 90s. And just today they've committed another five billion to investing in the creation of more AI-focused chips of their own. Uh obviously they're going to go toe-to-toe in the data center market against the likes of NVIDIA. Um, but Ireland has always been a hub for global tech companies. Do you think it could also be a hub for the sovereign European AI infrastructure drive that you're on?
SPEAKER_01Well, it should be there's there's uh the way the way that it it when you really think about this, and we have that there's there's there's loads and loads of positives, and there's one big negative. So let's just focus on the positive real quick. I mean, the Irish economy has been has Been on an incredible run um and to see it being driven in many respects by that investment, the type of investment that you make. There is no doubt that Dublin and Ireland is on the global map when it comes to technology as well. And so what a fantastic achievement that clearly means that there's a talent, um, and B, there's clients in Dublin, uh, and C the fact that it's in the EU, it really makes it uh a fantastic beachhead for the entirety of the EU block. Um so it's so it's a it's a great place to do business. Uh the the biggest downside uh is actually power uh and the cost of the power. So for for really building at scale um data centers that are super competitive when it gets several iterations from now, and it will be a race to uh you know thinner margins over time. That's just naturally how it works. The power element is the biggest, uh the biggest long-term impediment. Um Having said that, for us, in the short run and the medium term, that doesn't make much difference. It doesn't change much when it comes to you know getting this stuff out the door. So I think there is a moment in time right now with a relatively small community who know each other and have built a lot of businesses and who are very facile with AI and with the large companies and with deploying you know this sort of technology and which is basically Ireland, um, it has an opportunity to steal a march on the rest of the EU block. Um in the long run, though. And again, it doesn't it doesn't really matter where all the the data centers are in the end. Um I can say with confidence that we will be headquartered in Dublin uh uh forever. Uh but it's a great place to build the business out into the EU. That's precisely our plan.
SPEAKER_00Yeah, because because I I get the sense there's a more nimbleness to what you are talking about when you talk about neo clouds. And you know, we have a very fixed notion of data in the sense that it sits in a data center, but actually it moves around and it's backed up and it's it's supported by you know redundancy, you know, everything. You know, it doesn't just sit in one place, it moves. Um, but when it comes to things like cloud services, data centers, the overall AI infrastructure picture could eclipse all that, right? It's it's like literally everything that we knew or thought we knew about uh you know cloud infrastructure, um, the cloud itself. It seems that this whole AI sovereignty uh question is going to overtake that.
SPEAKER_01Well, I think I I I think what it will do is it will it will change the the complexion of the end state in in where all the data centers are actually distributed. That it the the it it won't all be able to stay in the US. You know, it can't it just can't. Um whereas for basic internet usage and for basic server usage for a lot of companies, you know, the it does that level of sovereignty doesn't matter. Um the prediction I think that we're willing to make, and and and there's like very, very prominent people in the industry who are already starting to do this. I think if the other the other big news item beyond the the announcement from clout from Claude that then got pulled back and everybody realized you you know that could happen, um, was also Alex Karp, uh, who was one of the uh you know anthropic early you know CEOs or founders. And and he he came on CNBC or one of those outlets a week or two ago and and had a what can only be described described as a rant, but nevertheless, a very coherent uh call to arms for sovereignty. And he's basically saying, look, you have you guys have to realize out there this stuff that I built, this is from the guy that built it, keeps all your data. It keeps your data, it trains the models on your data, you can't get it back. Um, and by the way, this starts to become an even bigger issue when you move to coding. And this was this was actually where we where we where it began for us. Because you know, now everybody's using AI coders to do significant amounts uh of coding. That the some of the statistics we're hearing from from the large American manufacturing companies of how many, you know, how much coding is now done using AI coders is it's staggering. And I believe it because the the the efficiency is is is is so clear to see if you're a technologist. And so we built a coder um over in in Apex that needed sovereignty because for the same reasons a big institutional bank or asset manager is not going to want to use a coder where its data and its business secrets goes up into the uh into the cloud and they don't know where it goes and don't know whether they can get it back or can control it. Like that's non-negotiable. That's now even almost more important.
SPEAKER_00And so if I think in those terms, it's not as I imagine businesses are looking to do everything within a walled garden as much as possible. I mean, the data may be sitting in in different data centers and protected in the neo cloud, but ultimately it's still secured and enshrined within its own walled garden, ideally, especially if you're creating your own IP.
SPEAKER_01Well, I think this is the this is the fascinating artifact of the arc of the fourth industrial revolution. You know, I used to talk about the fourth industrial revolution a decade ago when I first got into the digital asset space, because that's one of the pieces of the puzzle. Digital assets um will, by the way, will come together with AI to create agentic payments in the background. That's a whole thing that's also being built that we that we won't d de go into today. Um but uh but the fourth industrial revolution is this moment from where the you know the silicon chip was invented, and then we had computing, and then we had the internet, uh, and then we got crypto, and now we got AI. Now you put them all together and you've got the ingredients that basically, and you've got things like biotech as well. You stuff that five years ago, a lot of people, if you mentioned the fourth industrial revolution, were kind of I don't really know what you're talking about. But I say it now, and I suspect most people will immediately go, oh yeah, that right, we're in the thick of that because I'm now seeing the impacts of AI more personally, more quickly than I ever thought was possible. You know, we also we most people are using Chat GPT at some point, right? Most people are using AI at work one way or another. Um, and many are actually using it now as a core element. Um, that's what the fourth industrial revolution is doing. But the the the thing is, we've already been through a bit of this. Uh we've all lived through the arc of what happened with the internet. We all have lived through, you know, if you're a if you're a procurement specialist in a bank, you've been through the arc of being told that cloud was going to be cheaper than then seeing it get more expensive. And you've realized as an individual, I think we've all realized somewhere along the line, you know, I can't remember when I signed up for Facebook, 2007, maybe, um, maybe earlier, I don't remember. It was somewhere we all signed up for this stuff. For years and years, we never even thought twice that everything we were putting on there was not owned by us. And every Google Gmail account that we used for free, which was great, we suddenly realized, oh, well, they can use all of our data. They're selling that data for to advertisers. I think we're all starting to wake up to the reality in 20 in this part of history. We we've had data's important. I don't want you to have my data.
SPEAKER_00Yeah, we've had some crude lessons in this and recently. One I think that comes to mind is Apple not releasing its latest version of Siri in Europe because of its spat with the regulatory authorities here in Europe. But the other one then was only in recent days uh Instagram pulling a certain feature or Facebook pulling a certain feature of AI that allowed people to do things with photographs that were stored there. And then a lot of people are realizing actually, once I put a photograph up on a platform like Facebook, I no longer own it. It's no longer my IP, it's someone else's. And this is becoming very obvious. Is it obvious to businesses? It's public domain.
SPEAKER_01Yeah, but like business. I think that's a great point. You know, we we it's it's it's actually extraordinary the sort of disconnect between the way we think about um our own data and and the way we think about it on our jobs. It's it's it's it's a i i they should be the same, but they're really not the same. Um we for some reason we're all we all have a different frame of reference and we're likely more sensitive about our own data in certain situations and less than we should in others. But uh, but I look, I think businesses are not ready for this. Um and this is where you you sort of get into the question of how much to what extent is AI gonna affect your business? It could be that AI is pure and simple, um, a way for you, this is kind of the way it's generally been sold, a way for you to do what you're doing now more efficiently with with less cost, most likely because I feel like I'm probably gonna be able to fire a bunch of people because I can replace a lot of what they're doing with AI. That's kind of the pervasive, very high-level view that I think we all have. Like, hmm. Well, the thing though, that we haven't thought through, and I think there's a really interesting data point that everybody listening will find interesting if you don't know it, is that Anthropic allegedly almost bought McKinsey recently. Well, now, if that's true, and I have it on pretty decent information that that's true, why would the you know one of the biggest players in AI look to buy one of the most prestigious players in management consulting? And that's usually where you have to look. I mean, you were you know earlier you were talking about capacity. Well, just look at what Elon's doing. He's thinking about creating capacity in space that tells you, you know, where things are going. You know, smart people way smarter than us are already thinking about that capacity and where it needs to be built out 10, 15, 20 years from now. Um, so look to the smart people. Anthropic nearly bought McKinsey. They nearly bought McKinsey because they realize that AI is more than just an efficiency play. AI may in many instances mean you need to completely change the way you do what you do. It's not just firing a bunch of people, it's actually changing your organizational structure, it's changing the way in which you actually end-to-end your value chain and what it might look like. Now, if that's true, then it there's there's a different kind of engagement required over the next few years from boards and CEOs and stakeholders to really understand what this is. And I suspect we will look back five, 10, 15 years from now and we'll talk about a lot more c a lot more blockbusters from this one than the last one. A lot more companies that just got caught out because they thought, ah, you know what, we're good at what we do, we'll keep doing it this way. People love us. Um, a lot of people might come unstuck if they don't really get into that key question now. How is AI going to change my business?
SPEAKER_00Avoiding the Kodak moment is kind of like, oh, uh, I was safe and now um I'm not. It's like it changed. That's it. Brilliant. Well, with that's it, and everybody should worry. Everyone should worry. Well, hopefully, hopefully, uh, we're entering a more optimistic period, I hope. Uh, rather than uh the problem with net with the AI narrative is that it's too easy to fall into a doom and gloom scenario, too. And I I hope we're uh more in Florence during the Renaissance as opposed to uh you know uh the sack of Rome. So let's see where it all ends up.
SPEAKER_01Well, I tell you what, my parting shot for this is is actually I am an optimist in all of this, and and uh I I think that what people will find by taking the time to look at their business, and by the way, there's some businesses where it won't make much difference, so it's not like it applies to everybody, but many it will. If you take that time, engage, understand sovereignty and what it might mean to you, understand what sovereign what AI might could do for your firm, uh, you'll find very quickly that there's a source of incredible opportunity to go and do things you couldn't do before. Uh, and if uh if you need to understand how to deploy sovereign AI in your organization, come and speak to us at TensorX.
SPEAKER_00Brilliant. Well, Tim Grant, thank you so much for your time. Fascinating new company, uh headquartered here in Ireland and uh leading the charge for sovereign AI. And it's uh, you know, it's just a testament to just how quickly things are changing, how all these new terms are coming about. And, you know, we've gone from cloud to sovereign AI already. So it's uh yeah, watch the space. Thanks very much, Tim. Thank you, John. I appreciate it. We trust that you enjoyed this podcast and found it informative. Bank of Ireland is welcoming new customers every day, funding investments, working capital, and expansions across multiple sectors. To learn more, Google Bank of Ireland Business Banking. Bank of Ireland is licensed by the Central Bank of Ireland.
