How to Design AI for Scale From Day One with Tom Greenlees, Intelligent Core [090]

Speaker A: The Business of AI Podcast. Exploring how businesses are using AI to build services and tools, transforming organizations and delighting consumers. Produced by UKAI and hosted by Tim Flagg. AI is our business.

Speaker B: Welcome to another episode of the Business of AI podcast and I'm delighted to be welcoming Tom Greenlees from Intelligent Core to the podcast today. So Tom, welcome to the podcast.

Speaker C: Hi Tim. Great, thanks to be here.

Speaker B: Well it's great to have you here. I think there's a lot we could talk about today and I'm looking forward to getting into some of the detail. But to start off with, can you tell us a bit about your background, how you got into what you're doing now and what Intelligent Core actually does?

Speaker C: Yeah, absolutely. Um, well I mean I actually ah, started. I spent 20 years of my career predominantly in the consumer landscape, the consumer space. So predominantly marketing, uh, market expansion, product development, uh, growth strategies, working m for brands like Coca Cola, Red Bull and most recently British American Tobacco where I was global VP of Consumer experience there. And it was at this um, and my last role where I really started to experiment with the power of artificial intelligence, predominantly from a generative AI perspect. So we, we use a lot of Adobe Firefly suite for example, so creating uh, brand assets, marketing assets, um, and when you're working for a, um, kind of big corporate that has presence in over 150 markets around the world and you're trying to build global brands, consistency of message and consumer experience is absolutely paramount. And that's where what really piqued my interest in how I can help big companies and operators ultimately operate more effectively. Um, and so I actually left British uh, American Tobacco in April uh 2024 and I kind of took the chance to re educate myself. So I went back to Oxford University, did an executive Diploma in Artificial Intelligence with a focus on heavy industry. Um, and it was during my time there that I connected with my now co founders Adam Gillis and Robert Girvin, um, both of whom have deep experience in heavy industry. Adam in particular spent 20 years working for um, major oil and gas operators like Exxon, Swire Pioneer, um, Deep Blue. So he really understood the kind of challenges but I think more excitingly the opportunities that uh, heavy industry in particular oil and gas provides, uh, for AI to make real meaningful impact. Um, and so that's really where the Intelligent Core journey began.

Speaker B: Amazing. And at UK AI we often look at um, the whole energy, um, the importance of energy, um, and how we can actually ensure that the UK um is using um, energy sustainably. Um, but Also as an AI industry that we are building the infrastructure to be able to connect um, that energy into um, the data centers which are going to power all of the businesses which uh, which are using AI. So I think this whole sector is critical to understand that. One of the areas which I think is less understood is how AI can be used to optimize uh, some of those grids and to improve the processes. Um, but I know this is an area that you've been working on in heavy industry with oil and gas. So could you sort of help us to understand more around um, why some of these industrial projects fail and how to make them scale?

Speaker C: Yeah, look, I think it's a really good question, um, and I think ultimately it goes back to the legacy and of a lot of um, how these operators were set up. So I um, would say there are kind of several reasons why a lot of them fail. I would kind of choose to highlight five in particular. I think the first is lack of operational ownership. Right? So a lot of the time currently AI pilots are sponsored by uh, innovation strategy or digital teams. Now whilst those teams are really important, they're not always the people who own the operational problem. Um, if the person responsible for uptime, cost, safety or compliance is not directly involved with any potential pilot, that's where, that's really the starting point of where these pilots can become disconnected from reality. Um, second reason is um, I guess misaligned incentives. So a lot of organizations who we speak to say they want transformation. Okay, they say they want change, but ultimately leaders. And it's funny, I did a LinkedIn post on this the other day. But leaders are still rewarded on short term cost control. Okay? And if those incentives are only tied to quarterly cost reduction, then anything that requires kind of process change or requires new workflows or upfront investment can look risky. And you can't really ask people to transform the operating model whilst rewarding them only for protecting the current one. And so, you know, there's a kind of paradigm shift that needs to be taken at that sort of level to change who owns those pilots ultimately. Um, the third is data fragmentation. This will be a massive one. I'm sure a lot of your um, kind of uh, community will be watching. This will be familiar. It's a common challenge across a lot of different industries, but it's highlighted within predominantly heavy industry in oil and gas because a lot of them are uh, businesses of asset acquisition, um, predominantly. So industrial data is really messy. It can sit in SCADA systems, it can sit in DCS systems Uh, maintenance platforms, spreadsheets, sensor contracts, sensors, contractor reports, um, and a lot of the time actually sometimes just in people's heads. Um, so you know, a lot of AI solutions actually assume clean data, but the real world and the reality is that it's not clean. Okay? And the real world is not clean itself. Um, and so that's why at Intelligent Core, as part of our service offering we offer not just software but integrated hardware solutions that allow us to plug those data gaps that a lot of our customers and operators have through sensor packages and our long range drones. Um, and that allows us to plug those missing or plug those blind spots that operators just can't see. Um, the fourth, uh, is probably workflow failure. Um, so whilst AI might generate a prediction, who then acts on it, who owns it, when do they act, what authority do they have to act? And how does it change the kind of shift meeting or the maintenance schedule, uh, the field inspection plan? If those questions aren't answered effectively, that insight just becomes another piece of insight that's sat in a dashboard somewhere. It doesn't become action. And that action is where the value really becomes, um, really starts to deliver impact. Uh, and the fifth and finally is lack of trust. Field teams, they have experience, um, they know when something looks wrong. If AI is presented as a black box that tells them what to do, they're naturally going to be skeptical. So you really have to earn that trust. You have to show evidence, you need to explain recommendations and you need to start by supporting human judgment before trying to automate decisions. And that's what will really start to move from small focus pilots into large scale, impact focused programs of transformation.

Speaker B: Yeah, thank you Tom. That's a really good list. Um, and to hear you talk through them, um, from the background of hairy industry, I, um, think there's a lot of those process, um, and operation challenges there which they face. And maybe they're better at thinking about processes because there's a lot of engineers there who, you know, famously good at the process thinking. Um, but the same, uh, the same challenges are very evident in all organizations who trying to go through this AI transformation process. But maybe they're not thinking about it in quite the same process led way. Um, I mean I think there's so many things I'd love to sort of ask you about here, but I think I'm going to go with the um, question around the aligned misaligned incentives because that was a really good insight there. Um, because what we often see, you're right, is that people are acting out of fear or FOMO or some sort of short term uh, need. And you know, often that's the C suite, um, because they'll see the competitors doing something else and there's immediate pressure to catch them up and they will then um, drive or request some change to happen. And then everyone else also acts out of a bit of fear because the boss has told them to do it sort of thing. Within a large organization though, M. Um, you then also have what you were talking about. Those uh, those metrics, uh, those um, those targets which each team has. And you know, we've all been in, uh, the organizations we've worked in, we've all had KPIs uh, that we're trying to hit that will lead to some sort of incentive as well. So I can kind of understand why people are reluctant to change for those metrics because it might mean they're not going to get their bonus or they're not going to get that promotional whatever. So how do you, how do you actually start to change that? Because you know, to use the oil tanker analogy, changing the, the cultural, um, the culture in the organization is going to be hard, um, particularly in um, some of the organizations you mentioned.

Speaker C: Yeah, look, I think it's a really good question and I think a lot of leaders at the moment are struggling with the answer to that question. I think ultimately it comes down to reframing of I think what are known quantities. So let me give you an example. So, um, instead of saying the targets are based on cost cutting, that same premise can be articulated through value creation. And as soon as you reframe it within the context of value creation, you start to be able to open up more KPIs and measurable metrics that will align much more deeply with successful AI pilots. Okay, so things like uh, reducing staff downtime for example. Right. Or you know, making sure capacity of the staff is, is fully utilized. You know, those are, you know, through view to one lens there they can effectively be, you know, cost cutting views viewed through another lens. You're creating value, right? Value to the company. And actually the long term value of shifting to value creation means that those cost cutting measures are going to be permanent, not temporary. Okay? So then the question becomes what is it? Starts with the problem. It doesn't never start with the technology ever. You know, successful pilots will always start with the problem, the pain point. You know, what is the operational pain, how much does it cost, who owns it, what happens today, what decision would be like to improve? And then you start to define success in Those measurable terms. So the value creation terms. So these could be things like reduced downtime, lower chemical spend, fewer truck rolls, um, fewer emergency call outs for repair, uh, teams. Faster leak detection, for example, for oil and gas companies, better asset utilization, lower energy use is a big one. Reduced compliance risk. You know, all of those will have cost impacts, right? But you start to measure, you start to add those KPIs into some of these pilots. That's where you start to see actually the impact that the AI can deliver. Okay, so first thing is starting with that problem. Second thing is defining those success in measurable terms. Then you need to design the pilot around the workflow. Okay, so who sees the output? Uh, what does the recommendation look like? What does the operator do with it? Uh, and most importantly, is it integrated into an existing system? Okay, so, you know, those are the questions you need to be asking. And I guess third and finally you need to be thinking about a route to scale from the beginning. Okay? So I think too many pilots now that we see are designed as demonstrations ultimately. Um, and the question really should be, if this works, what happens next? Who pays for it? Who owns it? How does it move from one asset to 20 assets? From, you know, you take oil and gas, for example, from one basin to another basin. Okay? So, you know, so scale really has to be designed from day one. If you only start thinking about scale after the pilot, you're already too late. Okay? So I think, you know, that's if you're thinking about, you know, how do you reframe it? You need to reframe it in value creation, not just cost cutting measures. And then in terms of what success looks like as a pilot, as I said, you need to be looking at it holistically as opposed to in isolation and actually create a roadmap from pilot to scale, um, and understanding. And this is one of the things that we did at Intelligent Core when we were building our product. You know, our product is not built by researchers. Okay? Everyone in the team has domain experience in heavy industry. And we worked very closely with some of our customer engineers in the Permian basin, you know, VP of operations, um, in the field, to really understand what do they want to see, how do they need to see it, and what actions do they need to take. So what we have delivered is not just another dashboard. It's this kind of operational layer that sits on top of what they're currently doing, that gives them that control. And that's where impact really starts to be seen more, more effectively and more quickly.

Speaker B: Yeah, uh, thank you for expanding on that. That's really interesting to see. And I think uh, the thing you said um, around designing it as a demo, I think that is really key insight there because so often um, people will be thinking in that again very short term way about just being able to prove something and it's important to design the pilot uh, correctly and think about that a lot. But the point you made was actually if you're not thinking about the longer term how uh, you actually scale it and not designing that, that's going to have a massive impact and then that will lead to the short term pilotism. Um, I'm sure there's a better way of putting it, but we know there's all those stats which are kicking around about how many pilots uh, fail, etc. And that's probably why, um, but also wanted to expand on what you said around the action because that was also really interesting as well. The workflow failure, uh, as you called it, um, the fact that you um, need to actually have someone who owns it, who can actually take it forward. Because I think we, you know, a lot of us would have all been there where we've seen um, a strategy slide, uh, deck, you know, this is, this is the strategy. Um, but then when it kind of actually gets out into the, the organization, what reason do people have to pick it up and to drive it and to make it theirs? If they don't have an ownership, if they have the KPIs, um, they're not going to go and take it and actually drive it into action. Um, so can you unpack that a little bit more and help us sort of understand where um, have seen um, ways to get people more engaged with driving this action?

Speaker C: I think it's a really, really good question. I could talk about that for hours to be honest with you, because it's such a poor topic and I think there's a lot of, you know, I think at the moment if you, if you were going on to LinkedIn, right, and you looked at LinkedIn just in general or just generally across all of the kind of news that is out there at the moment, you'd be, you'd be forgiven for thinking that every man and his dog is deploying AI, right, and everybody's getting the benefit of it. You know, every company, if you haven't got it now, you're miles behind your competitors, already have a, you know, step ahead of you. The reality is far, far removed from that. You know, a lot of companies, small, medium, large are still understanding the impact uh, you know, from a cost, from an operational perspective, from an output perspective, how and where they need to start deploying AI. And I go back to my kind of previous or answer. My previous answer is you always need to start with a problem. So when you're talking about, you know, who owns it, who needs to get involved, you know, how do you then tie that back to your specific KPIs? You know, ultimately it goes back to what I previously said is it always needs to come from, you know, there needs to be a strategy put in place from the leadership perspective. If you, if you don't have that, that kind of thinking up front, which is, um, this is our strategy for AI deployment, this is where we're going to start, this is how we're going to scale and this is how we're going to manage the kind of uh, operational adoption of IT within our teams. You're already fighting a losing battle. So I think that the question there is not about the kind of broad answer question, which is who owns um it? Because you could quite easily say, does it sit in it, does it sit in operations? Does it sit in leadership? The answer is none of those is right. Um, but all of them are right at the same time. Right? So each person's going to own a different part of the output that AI is going to provide. So if we take oil and gas, for example, the output for the operators in the field, right. Is going to be real time operational insight into how the assets are running, what's going to be happening, when and how do they move from reactive operations. That is costly, you know, uh, increasingly, uh, costly and poor use of resource to a proactive mode of operations where they are predicting what's going to happen before it happens. Okay? And that's where you start to really allocate resources more effectively. So what those guys want to see is that operational insight and that data. And so they then own the KPIs around operational performance. Whereas your, your kind of corporate, um, team and the kind of executive team, they want to see the kind of cost implications or the benefit or if we go back to my point, the kind of value that's created out of that. So how is AI really helping our operational model evolve into a point where the benefits that we're seeing from them become permanently embedded into how we operate. Um, and so I think it's not an AI pilot or a company that says, go on IT team, go and run me an AI pilot and come back to me and tell me what it can do for us. It's never going to really succeed or take a long, long time to succeed. Where you really need to be focusing on is what's that end to end journey, what the problem is, who's who has a role and a kind of operational role within that problem, within that kind of pain point that we're addressing. And then how do you then tailor that output to ensure that every person within that workflow is seeing what they need to see, when they need to see it, and is their kind of performance is specifically tied um, to the evolution that AI can provide to them. So it all goes back to, as I said, to having that value creation, having measurable KPIs around those pilots and around the scaling of it, and then ensuring that everybody's invested in it from that kind of end to end, um, workflow perspective.

Speaker B: Yeah, um, well, great. Well thanks for sharing a little bit more uh, of the experience of going through those kind of uh, change processes. Um, you'd mentioned drones uh, at the beginning there and I wanted to understand a little bit more about that because I know you're quite involved in um, sort about, of talking, developing uh, drones for a range of different uh, activities. Um, and one of the areas that I um, think politicians care disproportionately, um, high amounts of mail they get on this is around uh, potholes, um, in streets. Um, so I don't know whether you've kind of used drones that might impact upon that. But I'd love to hear more around more generally how you're using drones and sensors and some of the examples that you might have seen where they're now also bringing in the broader AI technology.

Speaker C: Yeah, absolutely. Uh, I mean it all goes back to one kind of insight which is you can't optimize what you can't see. Um, in many industrial environments there are massive gaps in the data. And you know, pipelines may be remote, infrastructure may be underground, inspections may be periodic rather than continuous. Um, existing sensors might not capture the right signals. And importantly, particularly as it pertains to a lot of heavy industry field conditions, contains very, very quickly. And this is where the hardware becomes really important. So drones can inspect large areas, identify anomalies, collect visual or thermal data, support emission monitoring, um, and really reduce the need for manual inspection across very, very difficult terrain. You've got to remember a lot of our customers, you know, they have assets in the middle of the Texan desert, they have assets that are offshore, they have assets that are in mountain ranges. You know, all of these things that go where the resources Ultimately, so you need to be able to, to, to create a kind of ecosystem environment where you can acknowledge the fact that um, a lot of these assets are disparate, poor connection, which is why we can run at the edge, for example, you know, very remote. Uh, and they're also kind of legacy as well. So a lot of these, you know, assets, for example, they're 10, 20, 30 years old. So you're missing a lot of sensors. You know, telecoms company at the moment are going through a massive transformation in updating a lot of their underground pipelines, you know, because 90% of Internet connection still comes from pipelines that are on the bottom of the ocean. Right. But these are all being replaced at the moment with smart pipes because they don't have any data on what these pipes, the performance of them, the integrity of them. So there's a, there's a huge, um, kind of undertaking that's currently underway with a lot of these big companies and their assets and trying to, to evolve them and modernize. Ultimately, um, the sensors that attach these zones can create a more persistent view of asset condition, pressure, uh, flow, vibration and kind of a lot of other indicators. Um, and so for us at Intelligent Core, the future is not software versus hardware. It's the integration of AI sensors, drones and operational workflows. Um, and this is what creates the kind of much more complete intelligence layer. Um, so the hardware is not a separate product. The hardware then feeds the intelligence system. Um, you know, the hardware improves the data, the software interprets the data and then the workflow turns that interpretation into action. So, so that's a loop and to take your point on, on um, potholes for example, um, which, living in London, you know, I'm sure we're both victims of, of potholes and you know, burst water pipes and all these sorts of things. So you know, imagine a world where, you know, Thames Water for example, or the local council could know, oh, actually there's going to be a, you know, potential pothole or there's be, um, you know, a potential leak in one of these pipes, you know, before it happens and they can address it, you know, which means there's going to be much less cost, much, um, more efficiency in terms of their operations and that, that's where the hardware really becomes um, invaluable.

Speaker B: Yeah, and it's very topical because I'm actually looking out on a burst water pipe to my left up the street. And there was a burst water pipe down the right hand side of the street um, a week ago, which they fixed which is now, I presume, caused the one up on the other side of the road. But in terms of the, you know, I think people will be familiar with the concept of uh, you know, smart homes where you have lots of, uh, devices all plugged into one sort of system. And now as we started getting some of those more um, H vac, uh, so the ability to keep all the, to monitor and optimize the environment and the air and the heat and all of those extra technologies, we're starting to see those fitting into the smart home as well. Is what you're seeing for in industry, the big industry, is it sort of like a larger version of those, those smarter homes where things will, um, you know, actually I suppose, turn on and off different component parts to optimize and make it really efficient. Is that, does that analogy work?

Speaker C: Um, it's not what I've heard before. Um, but, but that doesn't mean it doesn't work. I think, look, I think there are definitely parallels you can draw between you know, creating smart homes, um, and all of the, the technology that can be plugged into that, um, and then kind of big operations. It's just the challenges are infinitely bigger and infinitely more complex. So instead of, you know, turning, creating a smart home, you're creating a, you know, a smart street where every house has, you know, different plumbing, different electricity, different energy demands. Perhaps, you know, some have tv, some don't. So, you know, perhaps some are connected to the mains, perhaps some use solar. Uh, you know, all of these different elements, you know, play into the um, the kind of approach, I guess what you're, I guess the, the adoption ultimately. Right? So how do you, how do you get people to trust, how do you get it, to adopt it and, and then ultimately how do you roll it out? And I think the, the key thing there and the same, I guess, a similar approach to how we take like smart home approaches. It's not a replacement for, you know, light bulbs or something. A smart bulb isn't a replacement for an existing bulb in the sense that, you know, you don't need them anymore. It's more a tool that's going to help them see more, decide faster and reduce unnecessary work. Um, and I think what, you know, what the front line of all these operators need to understand, um, needs to understand what the system is seeing, right. And why it's making that recommendation. Um, you know, why do we need smart, you know, why do we need a smart home? Why do we need sensors? Because it will reduce our overall cost. Right? It will reduce our, you know, our cost to us individually. And the same thing really applies, applies to these businesses. And I know we talked a bit at the beginning around it's more than just a cost cutting exercise. Um, but they need to understand why it's making that recommendation and how that recommendation connects to reality. So, um, if the system helps identify a problem earlier, uh, avoids unnecessary risk or usage, uh, reduces burden or improves prioritization, that's where the trust starts to come in. Um, you know, trust is only earned through performance. So you can't just tell people that AI is accurate. You can't tell people that smart homes are more effective. Like I know, you know, there's a lot of discussion around, you know, like energy smart meters from, you know, recently around how accurate they are, you know, and you know, some reports about them overcharging people, um, for, you know, predicted usage, for example. So the same thing still applies, you know, to AI accuracy. Um, there's uh, a terminology called AI hallucination, for example, um, that is kind of one of the biggest stumbling blocks. How do you make sure that if there's a gap in the data, the AI is filling that in with accurate recommendations. So you need to prove the value of it through pilots. Um, and so that's particularly important in safety critical environments. So in heavy industry, do not scale by forcing belief, you scale by earning confidence. And that means transparency. It means human in the loop design, it means clear metrics and operational relevance.

Speaker B: Yeah, um, it comes back to trust. Uh, you mentioned that a couple of times. I think that is fundamental to us as an industry. We have to recognize that humans um, need to trust, um, the technology. And at the moment there is a real lack of trust. We all have a responsibility to help build it through transparency, through education. Um, just a quick question around, uh, cities and we've got on to talking around things like potholes and uh, the utilities, how that's connected. Are you seeing that some of the systems thinking about, um, how to optimize is being brought into, uh, cities and regions, whether that be in the UK or worldwide. And you're involved in some projects like that?

Speaker C: Yes, uh, we are. It's predominantly from a utility point of view for us effectively. Um, so that's looking at energy consumption forecasts, grid performance, uh, utility performance, um, you know, the smart city conversations are generally focused a lot around, as you touched on a moment ago, connected devices, um, mobility platforms, cameras, uh, kind of citizen or social services. Um, that's where a lot of the conversations are happening at the moment. Because, because they're the least complex, let's say that. Um, but the opportunity for smart cities is, is ah, a lot, lot deeper than that. You know, uh, can city infrastructure or can the city predict infrastructure failure before it happens? You know, we take our uh, you know, burst water pipes example. Can it optimize energy, water waste, transport dynamically? You're seeing a lot in San Francisco M at the moment with you know, driverless cars and driverless taxis, all these sorts of things. You know, um, smart trains. There's a lot going on particularly in, in Silicon Valley with this sort of stuff. Um, and China is a whole different ball game. Um, you know they are miles ahead of us, light years ahead of us in terms of how they are deploying AI across their smart cities. Um, you know, so there's, there's a lot of those conversations happening. Um, but they're not quite yet at the level where we're going to see meaningful change I think at scale at least for the next five to ten years. Um, particularly in what, I guess the global environment that we're in at the moment, which is very volatile. You know, there's a lot of cost pressure. The government spending and borrowing is at an all time high at the moment. So you know, there needs to be kind of long term thinking around it and you know, ultimately how Intelligent Core can help and how we're speaking to a lot of these customers, helping complex physical systems um, become more predictive, more autonomous and more resilient. Um, and that can apply to a whole kind of swathe of different areas, um, that you know, touch on this idea of a smart city.

Speaker B: Yeah, no, I think that's really interesting particularly that uh, efficiency and resilience points. Um, and, and it's a conversation I'd love to uh, pick up again as we start to look at this. Across the uk we're looking at how um, data centers can be um, integrated with energy, with telecommunications within um, regional clusters and how all that infrastructure then comes together. So there's definitely a further conversation to be had there and bring you into that as well.

Speaker C: I mean ultimately just on that point Tim, I mean smart cities effectively, they're really dense networks of infrastructure assets, right? And they're very, very kind of energy hungry. Um, you know, China for example, they're now building data centers that uh, they just drop into the ocean and that's how they're called. You know, so you know there's, there's a lot of, there's a lot of kind of innovation that can happen. For example you take oil and gas, you know, they, they flare off gas, you know, daily, vast amounts of gas for safety reasons. Imagine if they took that gas and used it to power data center on site near a well. You know you're so there's a lot of opportunity around what our current infrastructure is that if viewed in a slightly different way can deliver real meaningful change.

Speaker B: Yeah and funnily enough I was talking to an Argentinian um, business uh, that does exactly that. They take the excess flare and they use it uh, for secondary purposes, data sciences which they can build up. Fascinating. Well, um, yeah, I'm going to have to start to wrap things up but before we do I wanted to just get your thoughts for what are the things that you're most excited about as you look six months ahead? Um, what's coming down the line in your area that you think is really exciting?

Speaker C: Well, I think for us intended core, what we're really excited about is our ah, hardware, uh, products. So um, we are very quickly driving a lot of exciting developments within our ah, drones. So we already have kind of long range drones that can fly for four hours straight. We're continuing to innovate and drive research and development on those. We're developing a drone that uh, we, we anticipate can fly you know, thousands of kilometers, not just hundreds. You know, we're looking at our sensor packages as well and continually to innovate and drive development within how we collect data, what forms of data we collect. And um, and so we're you know, intelligent core. We're really excited for, for the impact that our hardware is going to have. I think more generally if you look at an industry AI, I think what I'm most excited about is the kind of, you know, with great power comes great responsibility. Right. And so you know, AI has huge opportunity to do very meaningful good to a lot of people in the world. I think that fall beyond what we would consider developed nations. So I'm really excited to see how AI can really help support sustainable growth for businesses. And when I sustain mean sustainable, I mean it in every sense of the word, both environmentally and commercially, um, but also how it can start to um. I think what's really important is that you know, from a government point of view they, they use AI not as a chance to you know, cut cost but as a chance to lift people up ultimately and, and create better um, living standards for a lot of people where previously they haven't invested into it because of cost, uh, requirements or investment or you know, whatever it is. So I'm really excited for the future in terms of what it can bring to society, how it can lift people potentially out of poverty. But there's a lot of onus on regulators, on governments to ensure that the frameworks are there to ensure that, um, ensure that um, AI can be deployed effectively. Ah, sorry about that ringtone there. Um, effectively and responsibly.

Speaker B: Yeah, thanks. Thanks Tom. I think that is the ah, key thing. There is a lot of opportunity there for AI to drive so much, whether it be product efficiency or as you say, actually that social progress, but it needs to be done within the right guardrails. Um, so how can we find out more about the work you're doing at ah, Intelligent Core and stay in touch?

Speaker C: Uh, well, I mean we're LinkedIn, um, so all you need to do is search for us on LinkedIn, um, and drop us a message where you can get in touch or go to our website intelligentcore IO and there's a contact form there. We can get in touch, learn a bit more about what we're doing, how we do it and how we can potentially um, help your business or your organization. We're speaking to a lot of strategic partners at the moment as well. Um, we believe at Intelligent Core that AI will be as transformative for heavy industry as cloud computing was for software ultimately. And so we're very excited about the future and we also believe that, I think success will come from networks and community. And so we're very open to partnering and working with a lot of different, know, AI companies out there to really think about how we can address some of these bigger problems together.

Speaker B: Amazing. Well, um, we're looking forward to involving you in some of those conversations as well because we've got many members who are building all aspects of the AI, uh, tech stack. Um, and I think there's a, there's a lot of partnerships and collaborations there as well as some of the bigger players, the data centers and utility companies and the uh, local governments who are looking to actually use these sources. So plenty more to discuss, but no, it's been fascinating. Thank you so much for spending some time with us. And we've looked at uh, quite a few different ways, um, in which um, you're helping already to make uh, heavy industry in the oil and gas, um, uh, industry, uh, better able to use the power of AI technology but also some of those fundamental challenges. And I love that list. I still got it up here that you gave earlier on of you know, the lack of operational ownership, the misaligned incentives, data fragmentation, workflow failure, lack of trust. Those are all really key insights. Uh, and I think they apply not just in, in this sector, but also right across all organizations going through that transformation. So it was great to have your, your, um, your insights there as well. Um, so, Tom, thank you very much for joining us today.

Speaker C: Tim, it's been a real pleasure. Thanks for your time this morning.

Speaker A: Don't miss the next episode. Subscribe now. UKAI is the trade association for businesses across the uk, tech and non tech, large and small. AI is our business. Find out more at UKAI Co uk.