Podcast
In this episode, Allen Dorsey sits down with Mark Baines to explore what the rise of AI means for the future of work, decision-making and defence. Drawing on his experience across digital systems, military strategy and the defence and security sector, Mark shares his perspective on where AI can enhance human decision-making, from analysing vast amounts of data to supporting commanders and autonomous systems. They discuss the limits of AI in complex and uncertain situations, why human judgement, ethics, trust and leadership remain critical, and how militaries are adapting their approach to technology, training and warfare.
Mark explores this topic further in his recent paper, Human advantage in the age of AI and autonomy, looking at why the future of AI is less about replacing people and more about how technology can augment human judgement and expertise.
About A little bit of Nous
At Nous, we partner with clients across a huge range of interesting, challenging, exciting and complex issues. Having good conversations is a huge part of what we do.
A little bit of Nous is a podcast miniseries that explores big topics, through good conversations. Learn something, up your dinner party game and get to know the back stories of some of the Nousers who work passionately to achieve positive influence – not just for our clients and their businesses, but also for the customers and communities they serve.
Your host Allen Dorsey is Nous’ unofficial Chief Connection Officer, who is passionate about bringing the best minds at Nous together to discuss our biggest challenges. Allen is a curious conversationalist, always ready to share A little bit of Nous…
About Mark
Mark has worked in the UK Public sector for 30+ years and can apply his extensive experience in consulting, together with strategic thinking and senior leadership to help solve complex problems. These might often involve a combination of large-scale organisational transformation, leading edge digital technologies and complex programme delivery. Mark is well versed in Government and public sector decision making and managing risk tolerances around technical performance, cost and schedules. Mark has a strong focus on programme outcomes and achieving impact and value for money for clients and citizens.
Allen Dorsey
In this week's episode, I sit down with Mark Baines, who brings a deep experience across digital systems, military strategy, and the Defence and security sector to explore what happens as AI becomes increasingly embedded in how organizations operate and make decisions, and where the human remains indispensable. We explore why human judgment, ethics, trust, and leadership remain critical when decisions need to be made in complex and uncertain environments. Grab your coffee and let's jump in. Joining us from our London office, we've got Nous principal, Mark Baines, who has a really long professional background in digital systems and military strategic, operational, and tactical levels. And he's currently working across the defence and security sector with a specific expertise in international relations and global conflict. And in his spare time, you are an avid reader. Would you say that's true, Mark?
Mark Baines
That's definitely true. Yeah. And I think it's a pleasure, and it's something that I think informs our everyday work. And I think it also gives different perspectives when dealing with some of our Nous clients and some of our stakeholders, just to understand the bigger picture, I think, or have an opinion around the bigger picture, is important in life for professional people.
Allen Dorsey
Yeah, absolutely. So we really can't get away, it feels like, from talking about AI, and really every sector in every conversation. And you really wrote an interesting paper about how AI won't ever completely get rid of the need for the human element of work. Can you expand a bit on that?
Mark Baines
Yeah. Well, the thing about AI is, generally speaking, folks out there listening aren't close to the detail of the technology, and you look at it and you look at the media coverage, it's quite a scary narrative, isn't it, Alan, where this technology's there and it seems to be destroying jobs, if that makes sense, and it seems to be changing the workplace and potentially affecting everyone everywhere all at once, if I can put it that way. But if you peel away a few layers from that and have a look at it in more detail, there's a more stable, and I think important, reassuring message underneath: that the human element is absolutely essential, the human element is irreplaceable. And why is that the case? And that's the case because AI simply can't do everything. It has some great strengths, and those will probably get better in some areas for sure, but it also has some things that it can't do, some itches it can't scratch.
Allen Dorsey
It's interesting, you mentioned the things — we've kind of got two things going. There's things that it can do and things that it can't do. Perhaps maybe we start with things that it can do. Where do you think it's useful?
Mark Baines
Yeah, so I think AI is useful and will become even more useful at two levels in our lives in this attention economy that we live in. First of all, it will become ubiquitous in the sense that it's on all your devices, and it's offering help and assisting in a fairly general way to most humans in most activities in their life when they're on their devices, in their little digital part of their day, if that makes sense. So there will be that commoditized element of AI is kind of everywhere, and to a degree that might fade into the fabric of everyday life, would be my view going forward. But then where AI will become really important is on really niche topics, where it can produce absolutely fabulous insight or connect different parts of information to generate, for example, scientific research on genome analysis, or analyzing many, many images to assist medical surgery and cancer diagnoses, and this kind of thing. That really deep and specialist use of generative AI and agentic AI will be impactful and will improve the quality of life of lots of people around the world in the next decade.
Allen Dorsey
And you also mentioned that specialist expertise, and I know you're working on another paper exploring this idea more specifically in the context of defence and the military. Any initial thoughts?
Mark Baines
Yeah, so if you can look at lots and lots of data — a large volume of data at a high speed, faster than a human — that's fantastic. That's really helpful, of course, but you need the human for the judgment. You need the human to apply the ethics, and you need the human to understand the broader situational awareness picture. So a combination of those things is absolutely essential. It's kind of non-negotiable and won't ever go away. So the human judgment is required in all circumstances, but the machines and the algorithms and the agentic agents that might be developed to do specialist things will help to see through some parts of the fog of war and the complexity of everyday life, but they won't give a crystal-clear picture on all occasions.
Allen Dorsey
And so to be able to look through that kind of fog to make those decisions, you mentioned in your upcoming piece a lot of different attributes that we think our commanders should be developing. What do you think, within those attributes, will be the most important ones that we should be focused on as we're adapting these new AI technologies?
Mark Baines
I think it's the ability to fuse information when there is no complete picture. So this is bridging the gap between what AI can do and what AI can't do. I think AI is very good under particular circumstances, but it's generally not very good at dealing with a messy situation, Alan. So dealing with the gray areas in terms of decision making, where there is an incomplete picture, or a picture that is not recognizable to the AI because it can't analyze that and match it to something in the data that it was trained on — that is the key aspect where human judgment is required, whether to engage a target, for example, or whether to take a decision to maneuver some people around in a piece of geography.
Allen Dorsey
And so it sounds like there's definitely emerging applications, and so I wonder — where are we seeing AI being applied in warfare and conflict in the present, and how has that been changing people's calculus?
Mark Baines
Common to the commercial field, and I think the military field, is that the judgment of experienced humans is in short supply, and therefore it's kind of rare. So it's an expensive resource, if that makes sense, and you need more of it. And if you think of that in terms of, for example, intelligence analysts, there are never really enough intelligence analysts with enough eyes and brains looking at a problem to meet the demand. So this is where AI and agentic AI can be applied to best effect. It's looking at large volumes of data, which can be, to a human, potentially incredibly boring — staring at a screen for hours on end. Humans get tired, humans get hungry, humans do make mistakes looking at those sorts of images, but a computer and an agentic AI agent can effectively become quite a digital twin, but almost an additional digital analyst alongside the human. So what it does is it enhances the human ability to look at vast amounts of data at a higher speed for a longer period of time, and that can then inform decisions taken by other humans or commanders, but it doesn't replace them, Alan.
Allen Dorsey
And so then what do you think we can expect to see moving forward?
Mark Baines
I think you're going to see more AI embedded into decision-making processes. They're not taking the decisions, but they're involved in the assimilation of data and presenting choices and options. I think that's going to become more sophisticated in terms of what courses of action are open to a commander, that kind of thing. I think we're going to see more use of AI as essentially an essential tool for commanders at all levels. But I think most importantly at the operational level, where there are lots of different sources of information that can be delivered over the high-bandwidth links that are required to process that amount of data. And then secondly, you're going to see AI embedded, I would say, at the edge — which is at the edge of operational level down into the tactical level — where AI processing is available to drones and robotic devices, so that they can make smaller, lower-level choices on which route to take in an autonomous fashion, if that makes sense, or which obstacle to avoid, or, for example, which minefield to steer around if it's an autonomous robot vehicle. AI will be able to help in those circumstances quite effectively. And we're seeing those sorts of examples in Ukraine right now.
Allen Dorsey
And so if we're looking at the future, we're looking at all these different applications and uses — how do you manage all of these so that it's safe? It's a strange thing to say about war, and really ethical. How do you ensure that self-flying drones, for example, with AI-chosen targets, aren't targeting, say, a school?
Mark Baines
This is really important, right? If people are thinking of these devices in a movie way, where they're moving around a conflict area and causing damage without any kind of control — that's definitely, first of all, that will be almost illegal. I think it will be unethical, and it will be hugely dangerous in terms of distinguishing between friendly forces and legitimate enemy targets. So, just to put it more simply, there has to be control mechanisms in place when you have autonomous vehicles that may be informed or empowered by AI. And those control mechanisms generally will be either geographic — so some boundaries will be applied to the area in which they're operating — and also temporal. So there will be a time constraint: they're allowed to operate for this amount of time in this specific area in order to look for the targets that they might wish to find. And then if they find them, there should be some sort of confirmation that that's a valid target, which would involve a human. That allows there to be some traceability of that decision, Alan. So that gives accountability, doesn't it, to people, certainly in the modern age, should there be a subsequent legal challenge, for example, if that makes sense.
Allen Dorsey
And it's certainly a lot to consider when we're looking at the next generation as people join the military. So what kind of changes will this have on militaries with regard to things like recruitment and training?
Mark Baines
And this allows me to circle back really to the importance of the human, in the sense that what does the human bring to the table when you have this new world of autonomous machines and AI computing power helping them move around and cause effects? And the things that the human brings to the table aren't going away that I mentioned earlier. And the first one is judgment — the ability to look at this complex, changing situation, which might not be complete, and essentially decide what needs to happen. So the human has the ability to take decisions with an incomplete picture, to join the dots, if that makes sense. What the human also will need is the ability to win and hold the trust of the other people in the organization around them. So that aspect of trust and leadership and integrity is absolutely fundamental to those humans. That applies in the commercial sphere, and it also applies in the military sphere. So that ability to exercise leadership and command with trust and integrity in a complex digital field is absolutely essential.
Allen Dorsey
How should we be thinking about how we start to develop those skills?
Mark Baines
Well, I think most leading military training organizations and leadership institutes and so on — probably ranging from West Point to Sandhurst, to others around the world, to the Friends Academy — will be thinking about how they train their young junior leaders and commanders to deal with controlling these machines alongside humans. That's not new in the military field, is it? At the end of the day, there have been some kind of machines around for more than a century or so, but to control very sophisticated, semi-autonomous machines is a new challenge. And then to do that at a larger scale, across hundreds of square kilometers, that's a whole new skill level in itself. So I think the training needs to incorporate the ability to plan and the ability to operate in areas of time and space in a coordinated fashion that delivers successful mission outcomes. So that sounded a little bit like word bingo, didn't it? But that's actually — because it's easy to say, it's also very difficult to do. So the training for that needs to be fairly realistic, in the sense of potentially using simulations of that sort of situation, and then building up to small-level training involving those sorts of robotic machines and so on, and then expanding that in terms of scale and time and complexity to include a whole bunch of different layers. So generally the areas where this sort of activity takes place is managed in layers, based on altitude or position on the ground.
Allen Dorsey
And do you see this awareness happening already in militaries, or are we seeing people making these necessary changes?
Mark Baines
I think yes is the answer to that. I think the NATO countries for sure, and various Western militaries, and globally, are all looking at these opportunities and recent developments absolutely with a clear-eyed view to thinking through, gosh, what are the implications of the new combinations of technology? So it's not one single technology, Alan, it's not one thing, "here's some AI." It's actually, here's a robotic device which is actually smaller and more reliable and cheaper than it's ever been before, combined with a bit of quite advanced processing power, combined with some sensors that are smaller and lighter than they've ever been before. The combination of all those things together is allowing new things to be done, and they can be produced at scale to generate a level of mass which can have an asymmetric effect — an asymmetric advantage over really expensive, high-powered, almost exquisitely developed weapon systems that, for example, the United States might have, and also Russia. And if they lose one of those weapon systems, or they lose one of those really expensive hundred-million-dollar platforms, that's a big blow. But if you lose a $2,000 drone, it really doesn't matter. It can be replaced, is being replaced, and therefore the effect is really interesting to see how these changes are being embedded. So everyone's watching this changing environment at the moment.
Allen Dorsey
Yeah, absolutely. And it's interesting, it feels like in our discussion we've really run the gamut as far as perspective. And so I wonder — are you excited about AI outside of the military context, more in a general sense, or are you cautious, are you skeptical? What's your gut feeling about the future?
Mark Baines
People like to say they're excited about AI. I think it's the latest, and potentially the highest-impact, evolution in the digital world. So I think it's something that we can't ignore. It's something that we should embrace. It's something that we should get used to and comfortable with. But I don't think this is the end of the world as we know it, Alan, because, as I hinted at earlier on, you've got to connect these technologies together in a meaningful and useful and reliable, resilient, and ethical way. So it's a fantastic and interesting evolution of technology. It won't be the last evolution, but it's an important one. It's an important one in general life because of the improvements it will make for human analysis and some changes in the workplace, but it is not a replacement for an individual human. What it will do is change the nature of the work that people are doing, potentially remove some of their more repetitive and lower-level analytical work, but therefore enhancing and making life more interesting for the humans working on the more creative and novel parts of their job role.
Allen Dorsey
Yeah, it's interesting. It feels like that space, and what we need to be thinking about — that's sort of where you come in. And so I wonder, what do you love about your work as a consultant and working in this particular space?
Mark Baines
Yeah, I really like it because just having, sometimes, the ability to look through the window and see what's happening in the outside world, and then draw from that, infer from that worldview, what's relevant for the workplace for our clients — what would they do differently on Monday, or how would their operating model look different in a year's time? These are more relevant questions, aren't they, more specific. And I think that's a worthy challenge, and it's about really adding value, I think, and creating impact when people may be overwhelmed with the level of change going on around them and slightly perplexed by it, and not knowing where to invest, for example, and not knowing what to do with their workforce, and helping navigate those decisions — it's a great privilege, to be honest, and it's also very context specific, and I think this is really key. So for any organization, people often need a little bit of advice, a little bit of shaping of their decisions and their choices, to allow them to navigate around some of their, for example, investments — how do they explain those investments to their board? How do they justify their direction of travel on their strategy in this complex, volatile, and ambiguous world?
Allen Dorsey
It's certainly a lot to consider, and you've certainly given me a lot to think about with this incredible discussion. Thanks so much, Mark, for your time. And of course, before we go — I mentioned at the beginning you're an avid reader, so I am just sure that you have some wonderful book recommendations. Am I right?
Mark Baines
Well, there's a range of books that everyone can see in their local bookstore, but I would just call out a couple for you right now. One, I think, is important just in terms of the human advantage that I mentioned, and that will be — it's just a slim book called Hello World, and it's by a well-known science writer and broadcaster called Hannah Fry. And the subheading on this book is Being Human in the Age of Algorithms. And I think that's apposite and helpful for listeners who are not technologists themselves. It's an easy read, and it's a really insightful little book. And then there's another book called The Sirens' Call, which I think is useful in terms of the background to how we live our lives now — this is by Chris Hayes, and this is really alluding to attention. So attention will be — human attention will become, or is already, our most valuable resource, and therefore lots of technologies are really competing for human attention, aren't they, including for people's children and their families and so on. So that's a helpful framing of all of this discussion, in a sense. AI can try and draw your attention, but it can also try and maybe free up some human attention for the most important topics and decisions.
Allen Dorsey
Yes, a good companion read to this discussion. Wonderful. Well, thank you so much again, Mark, for coming along today. It's always a pleasure getting to chat with you, so hope you have a great day.
Mark Baines
Thank you very much. It's been a pleasure.
Allen Dorsey
Thanks for tuning into this week's episode. Nous Group is an international management consultancy, and for the past 25 years, we've helped public, private, and not-for-profit clients deliver better outcomes for their customers, communities, and society. You can explore more of what we do at www.nousgroup.com, or follow us on our socials. We'll catch you next time.