CRSP08: State of AI 2025 pt.3: AI Unplugged - from data to sovereign intelligence with Johanna Hutchinson, BAE Systems

In this last episode of the special AI mini-series, we now explore the human side of transformation, where technology meets purpose and people remain at the center. From future jobs and critical thinking to working with C-level leaders, how human intervention and high-quality data drive success in an AI-powered world.

This week Dave, Esmee , Rob sit down with Johanna Hutchinson, CDO at BAE systems about why data matters, the rise of Sovereign AI, and the skills shaping the intelligence age.

 

TLDR

00:55 Introduction of Johanna Hutchinson

02:09 Explaining the State of AI mini-series with Craig

06:01 Conversation with Johanna

34:20 Weaving today’s data tapestries with AI

40:20 Going to a rave

 

Guest

Johanna Hutchinson: https://www.linkedin.com/in/johanna-hutchinson-95b95568/

 

Hosts

Dave Chapman: https://www.linkedin.com/in/chapmandr/

Esmee van de Giessen: https://www.linkedin.com/in/esmeevandegiessen/

Rob Kernahan: https://www.linkedin.com/in/rob-kernahan/

with co-host Craig Suckling: https://www.linkedin.com/in/craigsuckling/

Production

Marcel van der Burg: https://www.linkedin.com/in/marcel-vd-burg/

Dave Chapman: https://www.linkedin.com/in/chapmandr/

 

Sound

Ben Corbett: https://www.linkedin.com/in/ben-corbett-3b6a11135/

Louis Corbett:  https://www.linkedin.com/in/louis-corbett-087250264/

 

'Cloud Realities' is an original podcast from Capgemini

Creators and Guests

Host

Dave Chapman

Chief Cloud Evangelist with nearly 30 years of global experience in strategic development, transformation, program delivery, and operations, I bring a wealth of expertise to the world of cloud innovation. In addition to my professional expertise, I’m the creator and main host of the Cloud Realities podcast, where we explore the transformative power of cloud technology.

Host

Esmee van de Giessen

Principal Consultant Enterprise Transformation and Cloud Realities podcast host, bridges gaps to drive impactful change. With expertise in agile, value delivery, culture, and user adoption, she empowers teams and leaders to ensure technology enhances agility, resilience, and sustainable growth across ecosystems.

Host

Rob Kernahan

VP Chief Architect for Cloud and Cloud Realities podcast host, drives digital transformation by combining deep technical expertise with exceptional client engagement. Passionate about high-performance cultures, he leverages cloud and modern operating models to create low-friction, high-velocity environments that fuel business growth and empower people to thrive.

Producer

Marcel van der Burg

VP Global Marketing and producer of the Cloud Realities podcast, is a strategic marketing leader with 33+ years of experience. He drives global cloud marketing strategies, leveraging creativity, multi-channel expertise, and problem-solving to deliver impactful business growth in complex environments.

What is Cloud Realities?

Exploring the practical and exciting alternate realities that can be unleashed through cloud driven transformation and cloud native living and working.

Each episode, our hosts Dave, Esmee & Rob talk to Cloud leaders and practitioners to understand how previously untapped business value can be released, how to deal with the challenges and risks that come with bold ventures and how does human experience factor into all of this?

They cover Intelligent Industry, Customer Experience, Sustainability, AI, Data and Insight, Cyber, Cost, Leadership, Talent and, of course, Tech.

Together, Dave, Esmee & Rob have over 80 years of cloud and transformation experience and act as our guides though a new reality each week.

Web - https://www.capgemini.com/insights/research-library/cloud-realities-podcast/
Email - [e-mail]

CRSP09: State of AI 2025 pt.4: AI Unplugged, from Data to Sovereign Intelligence with Johanna Hutchinson
[00:00:00] I, I categorically refuse to enter into such a pedestrian debate. Robert, that's the best you can do about flying cars. If we got practically, I wouldn't trust flying cars on the school, Ron. Absolutely ridiculous. We'll take about 30 seconds and eight kids would be squished.
Yeah, but would you win your kids up and then save on parking challenges? Oh, there you go. Actually, now they're thinking outsideof action. That's kind of thinking, that's gonna move us forward as a society.[00:00:30]
Welcome to Cloud Realities, an original podcast from Capgemini. And this week a conversation show as part of our state of AI miniseries. And in this episode, we are going to be talking about the leverage of very large data sets in complex industries. I'm Dave Chapman I'm Esmee van de Giessen and I’m Rob Kernahan.
And I'm delighted to say joining us for the conversation we have got our guest host for the. [00:01:00] Series Craig Suckling. Craig, you good? Yeah, Greg, fantastic to be here. Again, love being part of this. Nice to see you mate. Nice to see you. And I am, uh, doubly delighted to say that we have got Johanna Hutchinson, who's the Chief Data Officer at BAE Systems joining us for the conversation today as well.
Johanna, how are you? I'm good, thank you. Thank you for the invite.
Uh, it is absolutely our pleasure. Wonderful to see you. Thank you. And the logistics around actually making this mini series happen has been, let's say, [00:01:30] a little challenging over the course of the last, uh, few weeks for a hundred different reasons, not least, our brand new technical platform wherein we finally managed to get this whole Gus group of individuals together after wrestling diaries for a number of weeks.
And the platform disappoints us. As if AI is actually trying to hear your insights. Greg, what do you think?
Yeah, there is a little bit of a conspiracy, uh, theory forming there. Um, but we've got round it. We have been able to break the hallucinations. We have [00:02:00] extracted the bias from the model we bow through.We're battling through, I dunno, Dave still is. So, there's still a fair bit of bias in the system.
Craig, remind us what we're going on about in the miniseries.
Sure, thanks. So super excited to be continuing with this AI mini series. Um, the, the series Cs are speaking to the most innovative, the boldest leaders across industry in data and ai diving deep on the practicalities and the realities of how they are driving [00:02:30] change in their organization and their industry, but also looking to the future.
Thinking about what this looks like as we think about the way AI is gonna impact how we work, how we live, and how we run our organizations. So really excited to see this continuing. Really happy to have Johanna as part of the series.
Very good. Yeah. And, and, and a bookend to the series. I think this is probably gonna be the last one we do in this particular run of it at least, is indeed Craig, what's the key insight that you've taken away from some of the conversations we've had, um, over the last [00:03:00] couple of weeks?
Well, well, for me, I think it's been the variety, and this is, I think, representative of how foundational AI is as a technology. 'cause every single person we've spoken to has a different view, has a different set of challenges and a different way of approaching and thinking about the value it's gonna unlock and the impact it's gonna unlock, unlock in, in their area.
But there's, there's also been a lot of commonality. The commonality on keeping humans in the loop, staying authentic, me making sure we balance how we deliver on value with trust and security as well. So variety. [00:03:30] But a lot in common as well is my main takeaway. I,
I, I like it a lot. Mine, mine is similar. Uh, the challenges of scaling, the challenges of capability.
Um. Are are prevalent in, um, all of the industries that we've talked to. Uh, and also they're not dissimilar in a lot of ways to some of the other technologies that we've deployed over the course of, you know, the last 20, 30 years. Maybe a differentiating factor, however, in the scale of impact of [00:04:00] AI in our lives compared to some of the others, perhaps Roberto, any big ones for you?
I'm gonna go for the excitement is. We're definitely there. Technology is actually providing results. Industries up. Starting to implement it. It's just the, the huge potential that exists. It's like there's been this massive lift in the last two years of the art of the possible, as we've been exploring it through the industries, there's really exciting things going on and what that's gonna be.
So like we've only really had this lift [00:04:30] for the last 18 months, two years. Mm. What's five years gonna bring? If you look at cloud, it's been around for say 15. It took much longer to permeate through. Yeah, right. And now the implementation pace is so much faster. And that's the bit I'm like. So much potential.
What's actually happening today is already really exciting. What's gonna be the world in another 2, 3, 4, 5 years?
Yeah. Even if, even if you see AI as not the, Hey, we're on a path to a GI, and we're on a path to artificial consciousness and we're on a path. Even if you just see [00:05:00] it as like. We've got a, we've got some strong tools, like better tools than we've had for absolutely ages.
Um, particularly in automation and particularly in interface, and particularly in getting to the root cause very quickly of a particular thing or aggregating information for you. Even just at that definition, the impact is huge, isn't it? Oh yeah. And especially the potential for me, what I see is that the potential of humans are that that question is [00:05:30] rising more and more.
If we use AI as a tool, what? What do we want? Uh, it to do for us, right? Who do we want to be as humans and society? And I love that question. And it's rising and rising because we now, we're not afraid of ai, I think, or at least a lot less than a year or two years ago, but now more and more about us. Uh, so I'm very curious and it feels, I don't know, it feels quite liberating, I think already.
Well, let's dive in today's conversation 'cause I think we're gonna touch on a number of those themes [00:06:00] actually. So let's, let's go, let's turn to Johanna. And Johanna. Let's start by just understanding your background. So tell us a little bit about how you ended up as the CDO of BAE.
Yeah, so this is the first group role as Chief Data Officer for BAE Systems.
So it's really sort of an honor to, to hold this and take the organization into its maturity in the, um, data and AI space. So, recognize the organization's had huge long history of, of working with data analytics, [00:06:30] AI across engineering products, but really what we're starting to see now is, um. The level of integration and of course the speed of technology development.
So it's quite a moment in time, um, and, and very apt to bring in a chief data officer. Um, so my background originally trained as a scientist, so I have then a PhD in comparative psychology. And from there. After I'd spent far too many years, um, with a, being very decadent in my twenties, [00:07:00] um, doing exactly the research that I wanted to and answering exactly the questions that I thought were interesting, I thought you were gonna go a wholly different direction there.With that, I moved into the, the civil service and. Um, spent 15 years, uh, leading data analytics capabilities across a range of different departments. Um, cumulating in, um, running analytics for the COVID response in Oh yeah. Um, what was called the joint biosecurity Center [00:07:30] in, um, the Test and trace organization.
Right. Well, there's a lot in that, so let's, let's maybe come back to that in a second. I'm just interested in your views on that generally, but over, over that. Sort of span of time in the data and analytics industry, and then of course, like now data becomes the foundation for AI itself. What are, what are your, your observations in how, um, organizations have been leveraging it?
Like we've been hearing about things like data being the new oil for [00:08:00] 15 years. Nobody's really done that. Obviously in the world of ai, that actually becomes a. A very likely scenario, but what's that journey been like from your seat?
It's really interesting, isn't it? Because at the heart of some of, um, some of the most basic AI capabilities, so the large language models, the machine learning elements, these are methods that.
We've been using for quite some time. The generative, um, element of it [00:08:30] enables the interaction with the, the vast majority. So no longer do you need to be able to do your advanced coding, um, to understand and extract the output from there. So that's been the, the leveler and, and a creator of, of much hype.
Um. And really the essence for me now is how we stabilize this into our businesses. How we stabilize this into our societies. Um, but also how we bring everybody on the journey in a considered [00:09:00] way. How do we understand the risks and the benefits? Mm-hmm. Um, where and how do we applaud them? Deploy the AI capability.
'cause a lot of this will go into backend systems, from those point of views as well as. Being a new capability that we will see and that which will adapt and which will evolve and which will continue to, to generate new speculation and discussion for quite some time. I should imagine. I would imagine.
And, and let's talk briefly about COVID and, uh, [00:09:30] test and trace organizations. So for those who, um. Kind of work, you know, kind of maybe need their memories jogging or didn't follow this closely. I think test and Trace was about understanding the spread. Was it? Mm-hmm. Um, so maybe tell us a little bit about that and then the data challenges.
'cause I would imagine that's a pretty high volume of data. Oh, absolutely. So test and trace phenomenally interesting to, to work in as an operation. Mm mm Moved from zero [00:10:00] to the size of Asda in about six weeks. Wow. Wow. Can you imagine that that is it absolute, it's hard to even get your hand head around it.
And you know, led by D Harding at the start of that and she was an absolutely formidable leader, enabling us to bring together people from across society. So a lot of people from consultancy, a lot of the civil servants, and a lot of contingent labor people as well. Um, and driving forward effective operations to actually [00:10:30] stand up.
The capabilities that were needed to enable testing, to enable tracing the operations, the laboratory space. Can you remember all the labs that were needed? Oh, yeah, yeah, I do. And then, of course, to deliver the outcomes of that in a world of changing policies. And I think, you know, just, just recently listening to the, the COVID inquiry, there's a lot of discussion about those changing policies from a data perspective.
Of course, what that means is your data flows are changing quite regularly. Right, right. Um, and being able to interpret that and [00:11:00] understand the impact of those changing policies is one of the, the main roles that we took. And of course, in government, we are there to support the decision making.
Um, and so providing that evidence base and ensuring that that evidence base was as good as we can get it, ensuring that we had the interpretation of what that data meant, and that we presented it in a way that people could consume it easily. So we had the COVID-19 dashboard and became sort of. Key levers in that time.
So a lot [00:11:30] of very, very complex modeling, um, and capability that was built to enable that modeling in the background, but fundamentally, um, you know, ensuring that heat maps and turn the curve mantras, um, and these types of, of concepts came through and what understood, um. Across society and presumably from a tool set perspective.
So I think that build, you were talking about probably predated chat GP T three by what, about [00:12:00] 2, 3, 2 years. Three years. That's right. Something along those lines. Yeah. Um, so what technologies did you use? And if it, you know, let's hope it doesn't, but if it did happen again, the, the like gen ai, what, what kind of difference do you think that technology would make, particularly in terms of making sense of the data?
So one of the challenges that we had was being able to give an up-to-date picture as to what happening, what was happening in the country at any given time. [00:12:30] So it's that pace, isn't it? That pace of insight. So the minister's being asked to make a decision and make a decision at 12 o'clock today.
Right. But actually, if you're running an operations system, the testing data's three, four days old. Yeah. Yeah, it's just been through the lab. It's come out of the lab. It's come into the system. You've reconciled the positive test with whereabouts in the country that person is. You've reconciled then what that means for prevalence in that area.But it's three days out. Oh. [00:13:00] So these are, these are areas that, um, that AI and the modeling techniques, um, could support us on more to be able to understand those trends, to be able to bring together multiple data flows, a lot of, of automation and, and capability there to be able to understand the impact of what we're talking about in any given way.
The challenge then was, of course, we didn't, um. I didn't know much about COVID. Right. Novel, wasn't it? How it behaved. Yeah.
Yeah. And so we didn't know how it behaved as a, [00:13:30] as a virus. We knew a bit about Coronaviruses in general, and we, he looked at the flu data, um, and there was a lot of modeling going on in the background from, um, our historic data on those types of things to try to understand and predict.
And then of course, do you remember we saw all the different variants come through? Yeah, alpha Omicron, and again, you know, they were behaving in, in different ways. So those prediction capabilities, those prediction elements are certainly areas where we can see, [00:14:00] um, AI strengthen and support.
What'd be interesting is your view is after that experience, o obviously it was new and everybody was finding their feet and all, you know, globally, there was a lot of confusion about the effect.
Do you do, would, would, you would hope that the learnings. And how to deal with it, how to model it, have been fed back in. So if something like that comes back, I mean, although the response was massive and you talk about, you know, to the size of Azure in six weeks, it was very impressive [00:14:30] response and it was all new.
Hopefully that feedback loop's been created and we've probably got a basis that's a bit more. Was that work? What, what was your view on that work and the sort of get ready, if it comes back that we've got a better start point?
Yeah. So I mean, the thing to say about, um, the colleagues in the UK Health Security Agency is they are absolute experts.Yeah. And many, you know, academic experts in their particular disease area, so Right. We are not short of public health experts who [00:15:00] understand, um, any different disease that comes through society. And of course we have international monitoring, so we're looking at where diseases are popping up around the world 'cause.
You know, we're just a small island. It's rarely here. Um, and modeling the impact as to when, how, and if it came into the country and what that, and what would that need. So there's no shortage of, of those experts. The capability that we built was much more about integration and the operational elements of it, um, and how you bring that [00:15:30] together at pace and scale to ensure that actually you can mobilize faster.
And after the pandemic, when Boris announced the living with COVID policies, the work of our teams then was to sunset some of that capability. And so, you know, it's, it's packaged up, ready to go if it's needed again. So, as there's a lot of talk from a policy perspective, there are key elements of pandemic response and preparedness and modeling capability that are there and, you know, for us to use as [00:16:00] required. Touch wood, hopefully not anytime soon.
Yeah. JJ Johanna, one of the things that strikes me of that period was the, the need to act very quickly and make decisions fast in a environment where there's a lot of noise and uncertainties and. I think, you know, coming back to Dave's earlier questions on, on like the pace and the advancements happening in ai, how much do you think, you know, if, if, if an event like that was to happen again, that we would be able to use more AI to do so many [00:16:30] more scenario analysis, so many models and simulations faster to be able to derive a more informed view to make decisions in, in a more rapid way? Do, do you think that would change things in the future?
Um, I'm conflicted in two ways, Craig. Let me tell you why. Um, so one of the, the strengths of the response was the SAGE committees. So we had our, our expert users sat on, um, on Sage, who would do a lot of the [00:17:00] modeling in advance and bring that modeling in, um, and enablers.
Um, and the doing modeling very, very quickly as well to enable us to look at different scenarios and, and different impacts of policies that were being suggested and, and coming through as well. Right? The first thing I should say, um, about models is their estimates. So, of course, you know, they're never 100% robust.
Um, and the second point I'd say about the, the modeling is the more modeling that you have, often the more conflicted people are as to what the outcome will [00:17:30] be. And so for me, this comes back to people are at the heart of these things as well. Right. Um, the more information that you have, often the harder it is to form a consensus.
And there were consensus statements published on a regular basis from our scientific community. Um, and, you know, and then to make a decision on the back of those. So going into the AI elements, yes, you could do more modeling, could do more scenarios, you could speed up [00:18:00] that process. Would you get to the right answer though, Craig?
What's the right answer? And there's an interesting paradox. You, you've drawn out, there's like, essentially the decision still rests with humans and like more, more information doesn't necessarily really mean more clarity in making a decision. Exactly. But can AI help you to get to that point of this is the best decision you can make considering thousands of different factors and simulations possibly, right?
I think so. It's very interesting. Yeah, because the, [00:18:30] the big challenge for government decision makers, and particularly when they were communicating it to the public was, I, I'm guessing what you can say within what context, you know, kinda what, what up to date data, so I guess it could. Create a more informed picture, assuming of course, you know, um, the data's in good order and all of those sorts of things.
It could probably get you to those things a bit faster. Interesting. Interesting. You, you could speculate all day in terms of the differences that would make, but let's, let's, let's move on. Um, [00:19:00] to, um, your role now, uh, Johanna in, in the world of BAE. So paint sketches a little picture of the history of BAE, obviously a a, a very large and prestigious organization, do massively complex technical things.
Give us a sense of that and how the world of data is, is maturing within.
Yeah, so I mean, BA Systems is, um, an enormous global company. Um, I think many, many people in the the uk um, think it's [00:19:30] a predominantly UK business. It is a UK business, but we have a footprint in over 40 different, um, countries as well, and a growing footprint elsewhere as well.
So it's, it's a really interesting evolution of a business. The organization, um, splits into to different elements. So we have, um. A large air capability, um, a large maritime and land, um, sector, and also a digital intelligence. Um. Sector as well, much [00:20:00] more focused on sort of the, um, cyber and digital capabilities generally working with governments. So, um, our customers are governments of the world and working hand in hand internationally and on some of those contracts. Um, and also of course on some of the, the big platforms. So the submarines, the planes, the ships, and in collaboration, close collaboration with, with other, um, defense companies as well.
So this must [00:20:30] push the, this, just the context within which you're working there, you know? Not just the physical equipment that you are, uh, manufacturing and installing systems in, but the world of cyber and digital that you were talking about. This must infer, or correct me if I'm wrong, that your thinking on data, how it's leveraged and the emerging world of AI must have to be pretty close to the cutting edge.
Absolutely, [00:21:00] absolutely. So many of the products that we have brought to market this year, we'll continue to bring to market, have highend AI capability baked into them. So high end engineering elements, talking about, um, working on the edge, um, talking about mission system com combinations that go in there, autonomous.
Vehicles, um, from drones to land rovers and tanks, um, looking at swarming capabilities, thinking about [00:21:30] what our, um, customers are, defense customers are now talking about is multi-domain integration. Um, so how do you bring, same as what we're just talking about with COVID. How do you bring multiple streams of data together at pace and to enable a decision to be made?
I think just, just as a slight, just as a slight aside, before we go into, uh, that, and we'll come back to your question in a second, Rob, I just wanna point out to Rob, uh, the notion of swarming capabilities, and I knew this was gonna come back. Yeah, just, just be, [00:22:00] before we started to record today, Rob couldn't get his head around if you had a flying car, how you might drop your kids off at school with other parents who also it'd be chaos.
You thought it would be chaos. But the, the cars can talk to each other and swarm in the way that has just been described by Ja. The solution is there, Roberto. Now it's human behavior. You see, they'd go, the swarm would be working and then a human would think, I know better than the swarm. Hit the button, take control, and then that. Would that be true? That would be the event that would cause utter bedlam, I think I was saying. [00:22:30]
A, a swarm of vehicles or a school pickup is more chaotic than on a battlefield. Yeah, I think that's what Rob said. I'm not, I'm not sure I'm signed up. I don't think, I think you've dramatically misquoted me there actually, in taking me outta context, Rob.
Go on. But actually that, that multi-domain operations, uh, via multi-domain integration, absolutely key for say, an individual ministry. But we've also got the problem of. Other countries we need to integrate with as well. So there's this huge complexity [00:23:00] of I'm already dealing with some of the most secure information in the world that I need to protect 'cause it's mission sensitive.
I've already got to integrate my five domains from subsurface to space and suddenly there's a couple other nation states I need to talk to as well. That information. Data flows and then that might be over a chaotic communications network. I mean, the, the problem space that that defines and the, the way you have to engineer it, it's incredible.
Cha challenge, I'd be interested in your view on, on how you might approach that as a sort [00:23:30] of, I'm not sure if I'll be able to connect, I dunno where things are, I dunno who I might need to talk to in 10 minutes. I mean, that, that is just chaos personified from a communications perspective and data perspective It is. Each, if you think about its flows of data. So each flow of data works in a different way and comes and has different challenges. So, um, for example, we start thinking about satellite data, huge amounts of of information. Um, is there any value in it? Interesting. So a lot of the satellite, um, [00:24:00] technology and capability now is differentiating the useful data up in the satellite and moving the data that's required back or the insight that's required back at pace.
So that speeds up. Your, um. Your ability to get that information down, that's very different to being under the water. If you are under the water. That's an environment that's very, very difficult to detect, um, to collect some sensor information and to make sense of it. That's like, you know. So those [00:24:30] two environments for me are the most complex.
Um, and some of the areas where there's a huge amount of research, um, and engineering intelligence going into, to understanding and developing capability to help us through those. 'cause, you know, on, in the underwater environment, your intelligence is as good as your sensors in effect. Um, and thinking through how you get sensors to work in that very, very different medium.
That, of course is changing dramatically with, um, you know, wherever you are [00:25:00] at, at what level you are in the sea, um, and what what is going on around you. Um, so. Each element of data flow, if we're starting to think in that multi-domain integration has got to be, um, developed independently understood, um, from its own value, um, from the insights that come from it.
And then if we're starting to think about how you speed that up developed and deployed in a way that gets to where you need to be. Then you have to start [00:25:30] thinking about the technology stacks and the capabilities that you have and how you're working in any given environment. So, you know, it's, it's fine for us today, isn't it?
We're we're here recording this podcast and we're all linked in through technology into one environment. Just about seamlessly. Just about, yeah, right on the edge. Some technical hiccups at the Be beginning. Um, whereas Roberta, we were talking about, um, some of our, our drone swarms. Um, well, you know, your drones will come on and off. Um, [00:26:00] the, the computer edge as we call it, um, to download, upload information, take their mission, um. Packages, um, and then to, to deploy. So you're not expecting, um, a standard flow of information. It's quite intermittent. Mm. Um, independent sometimes upon what work, um, those strains will be doing in different environments.
Joan, just, just hearing what you're saying there, like it, it sounds like firstly, hugely impressive and [00:26:30] hugely complex, and it feels like this is a real bringing together of not just data ai, but physics and really top end engineering. So h how, how, how tightly coupled to these kind of like expert teams need to be.
It sounds like these have to be very embedded in the operations. Very decentralized, very integrated. So it's not. AI data is an afterthought. It's AI data is really embedded and integrated into the physics and the engineering. Is that, is that an accurate thing to say? [00:27:00] No, it is. Um, and of course, you know, bringing the, the best of, of all of our different institutions together in a lot of things as well, so.
Thinking about startup companies, thinking about working with, with SMEs, with different expertise, thinking about different technologies that are emerging all of the time, how they're coming to market, how you bring those together, and, and then how you integrate those into sort of those wider systems. So, you know, what I'm really talking about here is, is the cutting edge challenge of the moment.
[00:27:30] Uh, I don't think anybody has got this solved. This is where we're heading. This is the direction of travel that the customer is requesting and therefore the industry from a defense perspective. And of course, defense is often working in a arms race, so we are trying to beat the technology of the adversaries in effect.
Certainly from a cyber perspective, um, are working hand in glove, um, to bring the new capabilities to market, to understand the new technologies, [00:28:00] to have that own advantage from our own business perspective. Um, but also, you know, for that, that global, um, and wider good to understand how this capability works in these defense, um, and national security environments.
So what quality does that require from data experts to be in that ecosystem? Trying to connect with everyone, understanding data. Yeah, we talk about it a lot in terms of [00:28:30] intimacy. So we have, in the same way we talked about public health, where we have our health experts of different diseases. We have the same, um, in our industry as well.
And we have engineers of specific caliber with specific expertise in, in each area. And that might be, you know, the aerodynamics of how a plane flies, um, or the mission control systems that exist within, um, within some of our technologies as well. So it's that ability to be able to bring [00:29:00] together. Deep knowledge, expertise with broader people who have more experience in maybe a coding methodology or a technology stack, or, um, and the ability to be able to platform those people and enable them with some of the common tooling that they need, um, whilst also recognizing that they're differentiating tooling and capability, um, around that as well. It's not a, it's not an easy landscape [00:29:30] for our IT professionals, I can tell you that.
No. Especially when the cyber security elements have to come into it as well.
Right. Exactly. That, and let's just stay with people, uh, as we maybe bring today's conversation too, a little bit of a close and, and just double down on the point you made there.
So as we're sort of managing talent pipeline and building capability for the future, uh, particularly an AI. Impacted future. What do you think that looks like, both in terms of [00:30:00] what you are building for complex industries and the way you describe it, but also perhaps a thought on, uh, you know, like wider society.
I think we're at a, um, junction now where we're just starting to see a significant difference between our younger workforce coming in, um, and some of our, um, more experienced workforce in expectations of the workplace, um, and what and how they will expect to work. And that is quite a challenge I think, for [00:30:30] any business to be able to flex in the market and to be able to ensure.
You know, you'll gain in your business, but be able to enable safely. In our world, it's all about safety. Um, to be able to enable safely the use of new technology, to be able to develop an efficient and effective workforce. And of course, to be able to do that in a way that the considerations on regulation, considerations on [00:31:00] governance, um, are adhere to.
Some of our contracts that we hold to, you know, build the submarines for example, are it takes 10 years to build a submarine. So these contracts don't change readily. They are. You know, pretty robust and, and in there, and we deliver for a customer who has an expectation against those contracts whilst looking at a very fast paced, changing technology environment and different expectations of the workforce to be able to come together.
So how do we. Keep everybody safe. [00:31:30] How do we build a knowledge base in those that haven't, you know, aren't native digitally native in these environments? Um, whilst also satisfying the needs of some of our junior colleagues who will be able to uptake in this technology much faster and keep pace with it at a faster pace than the rest of us as well.
I put myself in that category there shouldn't, should I, I think we're all in that category, sadly. I think Esme might have a bit of an objection to what you just said there, Dave, but Yeah, maybe true. Maybe [00:32:00] true. Um, but the, and then in terms of looking at the sort of wider societal digital talent pipeline for years, of course has been diversity issues in that pipeline.
Um. And it seems that we, yeah. I'm not being overly critical here necessarily on the education system, though maybe slightly. When you look at it, you wonder if, um, it's moving fast enough and keeping pace with technology evolution and the consumption of technology. Um, I wonder if you, I [00:32:30] think that feels very acute at the moment, given the impact we think AI is going to have.
Um, I wondered if you had any thoughts on, on the, on the macro pipeline and you know, how we deal with that, maybe from the educational base?
Yeah, I think we need people who, um, are curious. We need people who are curious and we need people who can engage. So fundamental at the heart of all that we've spoken about today and, and all that that I do from a, a day-to-day perspective, we still have huge workforces of [00:33:00] people who have to keep up to pace with this, who have to understand, who have to work in a, in a certain way and within our guardrails.
Um, that is one of the, the major challenges. Um. That all of us have, um, regardless of the dynamics of that workforce, so. How do we, how do we manage that? How do we manage to be able to build people's curiosity, but also to focus more [00:33:30] on squiggly careers? Right. We don't, we don't expect necessarily to come in and do the same job all of our lives.
Um, so you know, we have, um, we're a big employee of apprentices. We have a huge apprenticeship program. Some of those apprentices will come in on welding programs. So what will welding look like in 20, 30, 40 years time? What will the robotics process mean there? How will they change their job? How will we bring them along?
Um, whilst currently we have a [00:34:00] huge demand for, for excellent welders. It's a technical skill and, and that emergence.
Yeah. So I came across a story about the bajo tapestry, uh, and it's over 70 meters long, and tells the story of the Norman Conquest through thousands of scenes. [00:34:30] Scenes are stitched, and historians actually call it one of the earliest forms of data visualization. Every threat carries context, bias, and perspective.
And that actually struck me, 'cause our data today is like a modern taper street, right? We weave millions of points together into a picture, but the way we stitch it and how stories are. You know, being drawn through it completely shaped what people see as truth. It's the same with the dashboard that we mentioned in [00:35:00] COVID or, you know, we're storytellers, I think especially also in your, um, in, in your field. Um, Johanna. Um, so, but how do you make sure that in the weaving process of today's data, uh, especially also now with AI involved, how do you just not only create efficiency, but make sure that meaning, context and trust is part of that picture? How aware are you of the fact that you're creating stories all day? I think that's a fantastic analogy. And of course the [00:35:30] interesting thing about those tapestries, about the cave paintings and, and all that we see there mm-hmm. Is they're not a true depiction. They're somebody's perception. And that's, that's one of our major, um, challenges with data is that actually the interpretation of those could go in multiple different ways.
Um, and so the nice thing about data is actually. In its historical form, and we have, um, huge amount of, of data storage, historic data, um, in our organization. You can [00:36:00] actually reanalyze it from different perspectives and ask different questions of it and gain different insights. And, and this is where, for me, it comes back to being intelligent users, um, and curious enough to ask those different questions and to understand the narrative and the so what from that?
Fundamentally, the world is a hugely complex place. It will not get, it will not get less complex. And so the ability to be able to [00:36:30] hold two conflicting views from the same data and both of them being correct is the area that actually we as humans need to grapple with, understand, um, and determine how we take that forwards. It's, it's like the constant frustration of. The data profession to a certain extent, it's like the world just will not simplify in a way that my architecture represented accurately.
Think may have heard that many times this week, I think. I think there's a very good point about what the bio [00:37:00] tapestry was, and it has a particular level of fidelity.
Which is, it's a recording of information done post-event through somebody's eyes. I think what we're getting better at now is keeping hold of the raw data. And preserving that so that we can reanalyze it, put a new perception over the top so you actually record what happened as opposed to post-event and things like this.
Mm. So what, what we're getting better at is taking the human outta the loop as we, as we [00:37:30] capture. And then I suppose, and then there's that point about how we represent that data is the perceptive point. We stick over the top. So I think we're, now, because we're in, we've moved out the information agent into the intelligence age, I think we've got a better chance of going back and maybe reworking it if we think we've. Got it wrong. Whereas like the Bay Tapestry, you can't go back and redo the battle again. It is tricky though because you, yeah. Do you really wanna interfere rewriting history?
Well, many do, don't they? It's a popular thing these days, isn't it? I was gonna add to that as well [00:38:00] because, and it goes back to your point, Johanna, on the cultural change.
As we start to see more AI diffused into our people in our society, we become more reliant on AI for the answers we receive, but also AI is creating so much more data now on our behalf. And is that gonna rewrite the future tapestry? Yeah. Where we see it starting to skew in ways we don't expect because AI is creating more of that data set in the future. And what about our synthetic data? What's the validity of that?
Well, I heard the amount, the amount [00:38:30] of to keep models going. They have to create huge amounts of synthetic data. Where's it going? What's happening? And then if the AI is used to create real data sets, does the synthetic data actually permeate itself into something we consider to be a, a proper data set?
And vice versa, is it all just gonna get munged together and mixed up? And we're gonna get into a right model. I do wonder about that. 'cause I think, well, hang on a minute. Who's controlling? What's is the right metadata Being able to attach so we know it's not being, yeah, there's a whole, uh, providence of data I think we probably still [00:39:00] haven't quite got a handle on yet.
It all depends on the question you want to ask. Yeah, I think so. And also, if the question is full with biases as well, you know, we, we must be so aware of what question we're asking to what data mm-hmm. Uh, is, is that a skill that you also see with your data officers, even, even the youngsters. But they, I, I love the youngsters for calling out those questions.
Like, you know, why are we doing the way we're doing things? Yeah. Uh, how good are we in asking the right [00:39:30] questions to data?
And this is the role of teams. So, you know, we get some great analysis that comes through some, some brilliant prototypes and and abilities, but it's that people being able to sit around and go, so what, um, how could it be better?
And why, why, um, did you think about this? Um, you know, there's no, there's no one person that can do. A lot of this build and technology and capability and taking us forward at this pace on our own, it is about being curious. It is about leaning back on professions. Mm-hmm. [00:40:00] And that exist within and outside of our organization and, and challenging ourselves to, to get to the best possible place.
Well, on that note, let's, uh, let's move on, but let quick, uh, tip of the hat there. Esme, that was a cracker. Good. That was a good one actually. I loved the very nice and cheap. Yeah. Yeah, yeah, yeah. You've done well there.
Now, um, Johanna, we end every episode of this podcast by asking our guests what they're excited about doing next.
And that could be, you're gonna go and see a tapestry at the weekend, or that could be something in your professional [00:40:30] life, or it could be a little bit of both. So Johanna, what are you excited about doing next? Well, I do have a ticket for a rave tomorrow. Oh, wow. T do say more. Do say more.
Well, I think it's, um, I think it's one of these new age daytime raves that enables me still to be in bed at a reasonable time. Oh, brilliant. That's the shift in the shift in demographic of the rave. Oh, yeah, yeah. The nineties, the age. It just keeps scrolling, doesn't it? Yeah. Yeah. Who's [00:41:00] DJing work is ex. I have no idea. I haven't looked up the details. Um, but certainly, you know, work is exceptionally busy. There's a huge amount of new capability coming through.
Almost feels like, you know, there's something new across the desk every day. Um, and really thinking through how we, we drive and deliver some of that new tooling, um, out into the market is, um. You know, certainly what keeps me entertained, interested, and engaged on a daily basis. Well, well, we wish [00:41:30] you, uh, endless luck with that as well as making it to bed at a decent hour after a daytime rave.Uh, and thank you so much for spending a bit of time with us today, an endlessly, endlessly fascinating subject. So thanks Johanna, Brilliant. Thank you.
If you would like to discuss any of the issues on this week's show and how they might impact you and your business, please get in touch with us at [e-mail].
We're all on LinkedIn and on Substack, so we'd love to hear from you. Feel free to connect and DM if you have questions for the show to tackle. And of course, please rate and subscribe to our [00:42:00] podcast. It really helps us improve the show. A huge thanks to Johanna, our guest host Craig, our sound and editing visit, Ben and Louis, our producer Marcel, and of course to all our listeners.
See you in another reality next [00:42:30] week.