Leadership in the Age of AI with Bupa's Clare Simmons

Kate Bevan: Hello and welcome to this episode of the AI Interrogator, Infosys podcast on enterprise AI. I'm Kate Bevan of the Infosys Knowledge Institute, and my guest today is Clare Simmons, who's the group head of Vendor Management at Bupa. That actually disguises how incredibly interesting and thoughtful she is on AI and leadership, and that's what I want to get into today.

So Clare, thank you very much for joining us. It's lovely to have you. Tell me a little bit about what you're doing at the moment and how you got there.

Clare Simmons: I'm doing work primarily around strategic vendors and how we want to work with our third parties. And how this links into the AI piece is I started to think about what the impact AI will have on the strategic workforce as a whole, specifically starting off with external supplies, the classic being is this like the, I guess five to ten years ago results of process engineering where we all started to see 15 to 30% efficiencies in our managed services. And I was thinking, I wonder if this will be a similar type thing and how would we manage that? How would we contract for it? And so that's how this started.

Kate Bevan: Are most of the vendors you are dealing with now, are they AI vendors or what's the proportion of AI vendors are you dealing with now?

Clare Simmons: Oh, you're still seeing the classic suppliers, all the ones everybody's worked with ever since they've been working in the industry. They're all still very large and they're all still omnipresent, I would describe it as, but now they're increasingly coming to you saying, "I've got this amazing AI thing I'd like to talk to you about."

Kate Bevan: So how do you evaluate what kind of AI services you might want to buy and might want to implement? Because, I mean, this feels to me if they're coming saying we've got all these AI services. That feels a bit like the tail wagging the dog. Isn't the starting point, what does my business need?

Clare Simmons: Yeah, and ever thus, it has been, I think. That's the challenge, really. It should be the other way around. Then, personally, I'm quite concerned this could turn into the proverbial sledgehammer to crack a nut. Do we need AI to help do some parts of the finance process? Maybe not, maybe just improve the process. That's my concern primarily. But yeah, you are finding they'll come along and say, I don't know, "We could use some AI to improve your testing. Speed it up by X percent, make it more efficient by Y percent." That type of conversation is pretty standard at the moment.

Kate Bevan: So thinking about bringing AI into companies, and actually probably just more generally onto any kind of big digital transformation, what's the top level challenge for leaders in all of this? Is it technology? Is it staffing? Is it data? What is it?

Clare Simmons: That's, I think, the really interesting thing about this, is the variety of challenges. So you've got people challenges, almost certainly, quite a lot of people challenges around the classic fear of change, and this is going to be a real game changer for us all. I know it's boring and we've all heard that phrase before, but I think it will be. I do think as well, you've got this issue of when do you use AI and when not. I used to have a boss who used the phrase, you've got many different tools in your toolbox, use the right one at the right time. And I am a bit fearful that AI is turning into the big heavy hammer that you can use for almost anything and that's not appropriate. There's definitely places where it's best suited and then there's other areas where maybe just leave the process as it is. And I think you've got to work out, is my business suitable for this or my team?

So for example, if you are managing huge quantities of data and you think that there is gold within that, and probably AI is going to really help you get there. The human brain can only cope with a certain number of data points and this is where I use the super skill that AI has to its best of its ability. If you fall into that category, I think it's amazing and I think identify that and really go for it and get excited about it and really emphasize those benefits because I think if you try and do it everywhere, almost certainly your organization will get change fatigue and there'll be a huge amount of fear about it around job security, et cetera.

Kate Bevan: And do you think those fears are valid. I mean, certainly, I've been around a long time and I remember going into newspapers in the mid '80s when there was a big transition from the classic technology of hot metal into what we call new technology in those days, and it certainly meant some jobs went, but it also meant opportunities and retraining and re-skilling and jobs changing. I'm seeing the same paradigm with AI. Do you think that's fair?

Clare Simmons: Without doubt. Absolutely. I think the thing is, yes, jobs will change, but what you and I did 10 years ago, 15, 20 years ago has all changed as well. And I think the same paradigm applies here. I think the best thing you can do is to look at this when the proverbial glass is half full and say, what's the opportunity for me and possibly what are my gaps? That's partly why I really started to look into this and get interested in it because I thought to myself, is this going to be a game changer? The space I've traditionally worked in is a lot of ADM type outsourcing. Well, actually, how will that look in 10 to 15 years time? Quite different. And if I don't get ahead of that now, then, now, I'm going to be behind the game. That's partly why I really started looking into it.

So I absolutely agree. I think there will be some jobs which will be massively impacted by this and then there'll be jobs you and I don't know the names of right today and that's exciting.

Kate Bevan: Yeah, I'm not even sure data scientist existed as a job title even just a few years ago. So again, so much change there. Maybe it did and I just wasn't paying attention. So as a leader then, how do you prepare the people in your business for this? What are the challenges that business leaders are looking at in terms of managing their teams?

Clare Simmons: I think this is one of the things we've got to start to look at and look at it from so many different angles. There's one of head on, encourage people to start looking at the tools and what it can do for them. I had this conversation with somebody in my team six months ago and I said to them, "I think Copilot will be revolutionary for you. I'd like you to start using it." "Oh, I don't want to. It is going to take away a lot of my work." This person is thriving now. Now they've embraced it and seen how it can take away some of those... Copilot is brilliant for taking away some of those mundane tasks all have to be done, but we all find very, very years irritating. But I think it's that do it in a way that is fairly embracing of the challenge the person will be going through at the other end, right? If you do it in a very direct, hammer type way, I think it's going to be really, really tough.

Then, I think there's going to be differences from our finance perspective. This is another subscription-based cost model that our FinOps teams are going to have to embrace. That's going to be really hard. And are we expecting everybody to have this in their business budgets? We will all have a section of AI costs and those types of things. Don't know. I think we need to have a look at that because I think if you don't place those costs in the owning cost center, then you could find the IT costs really balloon, although they have no control over who's running an agent 24 by seven. And then I think there are some interesting things around training, recruitment, those types of things that need to start looking at because you're going to have, in essence, a hybrid workforce. So embrace that.

Kate Bevan: Yeah. I mean, one of the things we talked about when we first chatted about doing this podcast was the need for different kinds of people, different kinds of worker. How do you hire for that? What kind of different workforce are we going to need?

Clare Simmons: Yeah. Well, I think you're going to have a hybrid workforce in terms of me and my team as it were. I think I'd be looking at, okay, well, how am I going to manage with the fact that really will I need junior testers? Will I need junior developers? Is it a problem we're going to skip that phase? And I've been thinking about this and you could say, yeah, it's going to be a problem we're going to skip that phase because nobody's going to go through that experience cycle, but do they need to? I'm not sure they do. Actually, what they need to be able to do is to ask the right questions to the bot so they can put in the right prompts so they get the right answers, and they need to be able to sense whether that agent either, A, needs a refresh of its data, B, is hallucinating, C, it's an answer that we were expecting to go in the right direction. The skill is going to be in asking those prompts and how you base those prompts and whether you test understanding those prompts are going in the right direction.

Kate Bevan: So we're going to need to train for that, aren't we?

Clare Simmons: Yeah, absolutely. Yeah. I mean, that's a skill that even if you look at stuff like Copilot, you ask it some random questions, you get some random answers. And that's a whole new skill that I think that we're all going to learn. But then I think what were you putting into Google 10 to 15 years ago as a search and what can you put in now, I mean, you can put in almost anything now, whereas 10 to 15 years ago, it had to be quite a well-structured search, if you recall, and I think the same will happen.

Kate Bevan: So we're in a way going backwards because you're going to have to learn to structure those prompts to Copilot or to whatever agent you're using in the way we had to learn to structure stuff for Google. That feels like stepping back a bit if you're going to have to relearn skills that have become implicit if you like.

Clare Simmons: Absolutely, and I think that's where it's going to be interesting how much this really takes off throughout businesses where actually people get frustrated with that and therefore the pick-up is a bit slower than we're all expecting because people aren't going to learn those prompts. They don't have the time to... You do see it today. You see that, "Really? I have to learn prompts?" You can almost see people thinking it and I do wonder whether actually we're going to end up in this trough of disillusionment for a year, 18 months whilst those elements get ironed out and then it becomes much easier to use the tools.

Kate Bevan: So if we're looking forward, we know what leaders look like now. What's a leader going to look like in 15 or 20 years time? What are they going to be managing? What are they going to be leading?

Clare Simmons: I think one thing that's going to be happening is I do wonder whether we look back on a 40-hour a week and say, wow, can't believe we did a 40-hour week. I think that might die. And the reason why I think that is because all of those recovery tasks that you have during your day, which allow your brain to just recover will have been taken away. But I don't think you can do seven, eight, 10 hours of highly intense work. So I suspect the hours that people work will have changed. I don't think that means the output will have changed. I think the output will probably actually go up a bit, but it's not going to, you are going to have seven hours of deeply intense thoughts. So I think that's one thing that will change for leaders.

I think the other thing is this piece around budgeting and things like that. I think you're not going to have this very simple budgeting of I've got 10 people, all of my roles are fails and therefore I've got my budget. With things like AI, you're going to get much more variability, so you're going to have to be able to manage that.

As well, the challenge, which we haven't, we've touched on a little bit during this conversation, but not so much. They're going to need to think about if they're using AI in a huge amount. How I manage the risks? So how am I curating that agent? You could call it what you do for your staff today, what you do for your agent. So performance management, you're going to need a performance manager, AI agent. It'll look different.

Kate Bevan: That sounds dystopian to me.

Clare Simmons: But what I mean by that is, is it starting to have bias in which case that needs to be fixed? Does it need to, the equivalent of L&D for an agent might be expanding the data set or refreshing the data set. All of those things, I think you will have to do. You cannot offload that or outsource it to your tech function because the reality is your tech function won't know it's impossible. They might be able to say to you, "Looks like your data set is getting a little bit tired around the edges," but it won't really know the quality of the output that you're receiving.

Kate Bevan: Yeah, actually that's quite interesting because one of the things we have been seeing in our research is the most effective AI comes when the C-suite is involved. And it's not just, I mean, the CMO driving it or the CTO driving it. It's right across the board. I suppose that's only going to increase as time goes on.

Clare Simmons: Without a doubt, and I think they've got to be really tight with it because the reality is it's going to be a huge driver of what their business delivers. And if your AI agents are not delivering what you want them to in accordance with your values and your targets and your goals, then you've lost control of probably over 50% of your workforce in effect.

Kate Bevan: That's quite easy because it feels like there are a lot of really traditional leadership skills still at work there, but we're going to need a whole bunch of different skills as well from our leaders. Have we got a pipeline ready for that? Are these leaders of 15, 20 years hence in the pipeline now?

Clare Simmons: And I think that's where now I would say anybody who wants to be a leader in 15 to 20 years time, curiosity, curiosity, curiosity. I cannot emphasize enough how important it is to never think you've stopped learning. You need to keep learning. You need to keep refreshing. You need to keep looking at everything in a curious way in saying, what could the impact of this be in five, 10 years time and what should I be doing to understand that better?

Kate Bevan: And also, I suppose the curiosity to understand what the technology is and what it isn't. I have conversations with people all the time when they say, "Oh look, but the AI turned out this absolute garbage." And I said to him, "Well, it's not a knowledge engine. You should understand that if you're going to work with it. It's autocomplete engine. It's a very clever autocomplete engine." So it's a real mix of skills that are we developing them.

Clare Simmons: That's one thing I think I would thoroughly encourage people not to say is, well, it churned out this awful answer. Humans do things which they don't mean to every single day. And so not really an acceptable answer. You've got to look at that and say, why? Why did it do it wrong? Because we're not going to go back. That's the thing is this is almost a one-way street, and if you try and reverse, I think that's going to be really, really hard for you as a person in your career. So I think you have to look at it, and that's what I mean by curiosity. Look at it and say, why did it do that? And what can I do to resolve that for the next time and the next time and the next time? Because it's a really intelligent thing, but you still do have to feed it and water it.

Kate Bevan: That's a great way of looking at it. Yes, it's the genie's out of the bottle. So now we have to learn to manage the genie. Right? So I'm going to finish up the question I ask everybody, which is, do you think AI is going to kill us all?

Clare Simmons: No, I don't think so. I'm really hoping it's going to be a buddy for us all and it'll help us all achieve more, but definitely there's some new skills ahead for all of us.

Kate Bevan: That's a really optimistic place to leave it. Clare Simmons, thank you very much indeed for joining me today.

Clare Simmons: Thank you.

Kate Bevan: The AI Interrogator is an Infosys Knowledge Institute production in collaboration with Infosys Topaz. Be sure to follow us wherever you get your podcasts and visit us on infosys.com/IKI. The podcast was produced by Yulia De Bari and Christine Calhoun. Dode Bigley is our audio engineer.

I'm Kate Bevan of the Infosys Knowledge Institute. Keep learning, keep sharing.