Exploring the Challenges of Introducing AI into the Boardroom

“Technology-augmented boards—let’s assume that for the next 20-30 years, they will still be human boards. We’re not going to see super-intelligent AI taking over. Boards are still relevant in that respect; the level of decision-making they’ll be doing will be elevated. They’ll certainly be using technology to support their decision-making.”

– Paul Smith

What you will learn

• Understanding the role and dynamics of boards in decision-making
• Leveraging sub-committees and deep democracy in board decision-making
• Exploring generative AI as a creative tool in boardroom decision-making
• Using generative AI to introduce counter-views and support critical thinking
• Introducing AI into the boardroom and addressing the challenges of adoption
• Enhancing board member performance and accountability with technology and self-improvement techniques

Episode Resources

Deep Democracy

Iceberg Analogy

Generative AI

ChatGPT

Transcript

Ross Dawson: Well, it’s wonderful to have you on the show.

Paul Smith: Thanks, Ross. Thanks for having me.

Ross: So, you help boards amplify their collective cognition? I gather that’s part of what you do.

Paul: That’s a part of what I do. Yes. So there’s working with boards directly, to help them make better decisions on behalf of all stakeholders. But also, my business Future Directors is developing and has gone into the market as a SAS platform to help boards, manage their board business, create more data insights, and to educate and build capacity along the way, as well. And that’s all about accessibility. So it’s taking the human need for the human to be part of that journey, that consultant or the train to be part of that journey.

Ross: Let’s frame that as cognition. So we have individual cognition, so taking information, making sense of it, hopefully making some decisions. A board is a particular set of individuals, yes, whatever it is eight, 10, 12,  more whatever it is, and, and I think it’s very useful to frame the cognition of a board in terms of the alignment again, how it is they find relevant information, make sense of that, and to make decisions. So what are some of the approaches which can help a set of individuals that end up around a table, to make better sense of the world and move towards better decisions? 

Paul: Yeah, like such a great question. And, you know, to think of the board as a collective unit  is so important as a collective decision making unit – that’s what they’re there to do. They’re there to guide and steward a company, organization or institution forward. Most boards range from a few people through to, as you said, much larger numbers, some boards are 20 plus. The optimum sweet spot is in those high single figures to make sure you’ve got enough cognitive variance. I think the other thing to say to give context to people listening to this around the boardroom is that most boards are not together all the time, they meet periodically, that’s the nature of board, they might meet once a month, or once a quarter, or whatever it happens to be in and they are charged with making decisions at the higher end of a business. So the governance end of the business strategic side of things, the risk management, long-term decision making, as opposed to the operational day to day. 

So really, there’s two parts to this, which are really important. One is the information they receive. Most boards are responsive to the information they receive from management, or executive, depending what you call it, and the conduit for that is the CEO. Their responsibility is to ensure that they’re getting the right level of information in order for them to make those decisions. But most or more tend to delegate that responsibility outwards, at the best board seek out their own information as well, both individually and collectively to supplement not only the information they’re receiving from the internal teams, but also their own arguments and opinions when it comes to debating and discussing a particular decision. 

The second part of that is the culture of the board itself. What is the decision-making culture? Many boards are quite autocratic, or what maybe hippo which is the you know, the the highest paid, loudest person type of thing, right. But the most effective boards, understand the balance for ensuring that you hear as many voices as possible, but make sure they’re relevant voices. So it’s not a case of everybody has a say, but everybody has a chance to say if they’ve got value to add. So information coming in, but also what is the culture of the board to actually help them to make the most I wouldn’t say best decision because you can make decisions based on information at hand and we live in this VUCA world. So it’s probably the most robust and resilient decision possible.

Ross: Yes, as you pointed out, the CEO or Executive are a source or filter of the information about the organization, but of course, decisions at a board level have been made within the context of the business environment, yes, social environment – everything which is technological environment, all of the external world. So some of that again, can be supported by was provided by the organization itself. But there is then a responsibility, of course, for the boards to understand the context; to be current. And I suppose that can be done both individually as in just – let’s just go out and be informed. But there’s also I suppose, are there ways or what a useful way in which boards can collectively find external information or input which can shape their thinking? 

Paul: Yeah, and you know, boards are still very much in the traditional camp of using advisors, consultants – they have various structures. For example, most companies and organizations of a certain size will have what’s called sub-committees. So these will be specialist groups made up of board members, employees, and increasingly third-party specialists, who you delegate thinking to decisions to, they play a sort of semi-advisory operational, but for the board, it’s not like day to day company operations, but board-operations that you will delegate certain parts of responsibility to, to bring up those recommendations boards. 

So delegation of responsibility is one of those tools that boards do because they do meet so infrequently, they can’t do everything. They’re everything. So they delegate a lot of this stuff outwardly, but it’s still when you use that word responsibility, it’s still their responsibility legally, as well as I suppose morally, to seek out what they should do so and then the board world, whilst you have these mechanisms, it’s still very much independent, individually-led because the majority of board the directors are what’s called non-executive, so not employees of the business. And they may be independent, which means they have no financial stake as well, they may get paid or might be unpaid. But essentially, they are independent. So a lot of the work they have to go out and do individually. 

So the collective side is bringing in external specialisms either directly to work with the board to upskill them or to teach them something or to delegate that through to a committee or the executive. However, the more technically savvy boards out there, technologically savvy boards out there, and unfortunately, boards are laggards when it comes to technology. And there’s a number of reasons we could get into that, but I don’t wanna become ageist. As you know, they are starting to utilize technology to support not just information gathering, but also their creative thinking and decision-making thinking. But there’s also some other techniques, human-based techniques, which are steeped in neuroscience and behavioral science, which can be used to elevate the cognitive ability of the group as well.

Well, there’s one in particular, which I like called deep democracy, which is what I teach. So it’s really used for when a group, in this case the board, is stuck. And that’s because a lot of the wisdom of the group or the views are hidden below the surface. I’m sure you understand, you’re aware of the iceberg analogy, where, what is known, is only a small proportion of what’s really there. So hidden agendas, outside influences, whatever it happens to be. So what deep democracy does it works with groups who are babies may be stuck making decision, I’ll give you an example of one I worked on last year, which was the board couldn’t decide whether to fire the CEO based on their performance, and they were stuck because there was support as the CEO, there’s people who said we have to move on, we need some fresh blood. And they were stuck on a decision. And it was a majority and a minority, but they couldn’t reach a general consensus of how to move forward. And so I was brought in to support this decision-making and the way that deep democracy works is like you run like a debate, but you get to debate yourself, that’s all designed to raise the EQ of the room by having you play your own devil’s advocate. Because what usually happens in any decision-making structure is some people be loud and opinionated, others will be quiet, some people will be entrenched in their view with an ideology or a dogma and other people will be variants, other people, so everybody plays a role. And the more the group stays together, the more they roleplay. So what this does, it tries to take that sense of I have a position out of it. So you start with a statement. And then every single person has the equal amount of time to argue for and against that statement. And you have to come up with reasons which are true for you. So even if you’re on one side to start with, you actually have to think of arguments against yourself. And what that does is it sort of semi tricks the brain into going ‘Oh, actually, that’s what that person said, it sort of makes sense. Because I have to I can’t, I’m not filtering it through what I think about that person. There’s no filter, it’s just me now.’ So what it does is you go round, and it’s all very contrived and staged. And it was a bit uncomfortable to start with. But everybody gets an equal amount of time to be heard. And you keep going until they’ve exhausted all the views. So all the iceberg is out as much as possible. And then you see how that’s impacted people’s viewpoints. 

And this is not about creating consensus in terms everybody agrees. It’s making sure that we have enough buy-in to a decision because everybody’s been heard. And so even if you get to a point where there’s a majority-minority, then you go to the minority Have a decision based on the rules of the group and you say, ‘Look, we’re really sorry that your position isn’t the one is up, what would it take for you to come on this journey. So you almost come up with an agreement, that again, it’s not 100%. But it’s enough to move people forward.’ Now, where that doesn’t work is where there’s real big personality problems, and they just cannot agree, not because of the facts presented to them just because of the ideology. So you might say, on the US political side of things, you know, there’s no way some of them can agree just on ideology grounds. But within most boardrooms, their job is to try and make the best decision they can. So once they can get over the hurdle of having to play a role or a different type of role. It’s really, really powerful.

Ross: That sounds fantastic, actually. And I think it’s really, there’s ways in which you could apply that conceivably to your own individual thinking. You go through that process yourself and think through sort of what are the issues at stake and to argue them to yourselves. I mean, it’s a little hard to do it for yourself.

Paul: Couples therapy. I like I’ve I haven’t done this, because I’m not a therapist. But I’ve told people who have taught this to boardrooms, leadership groups, large groups. At a conference, for example, some have come up to me afterwards, I’m going to use this with my partner.

Ross: Absolutely, yeah. That sounds really powerful. I really like that it is a tool to amplify collective cognition, sense-making. And, you know, there’s obviously an analog there with red-teaming, where you have some kind of decision and you get a red team to argue that case? And, yes, maybe a few variations on that. Are there any other similar techniques or approaches or structures you would highlight?

Paul: There’s not ones that I would particularly use directly, because I think they’re very much consensus building and in the boardroom, I tend to try and steer people away from absolute consensus, because it can be a race to the bottom. But increasingly, from a technology perspective, there’s generative AI that is a really powerful creative decision-making tool, as you and I both know, and hopefully more and more people are earning. It’s not a fact based tool. It’s not a fact-checking based tool, as yet. But as a creative supporter, as another voice in the room to stress test your ideas with the right bit of information. That’s what I’m starting to work with some boards on now as how we can dip our toes into the water of utilizing generative AI inside the boardroom to support our decision-making. And that’s like, for example, giving it a huge amount of context and decision, and then seeing what it comes up with. And it’s just another voice in the room, but not one that has any ego attached to it.

Ross: So for example, you might in a particular context, get an opinion from the generative AI or give it to offer another perspective?

Paul: You could say, you know, you could ask it well, ‘This is the thing we’re dealing with right now, here’s the context.’ With Generative AI, the more context you give it, the better. So the context around your operating environment, that decision needs to be made, what inputs are coming in? And you can ask it for, what approach would you take to help us make this decision? What information would you seek out, and then it might present you with things you hadn’t thought of? If you think about the collective intelligence of any group, there’s always going to be gaps. So what AI can do is try and help you fill those gaps. And then what I always love to say is like what I always like, in an iterative process with Gen AI, is I always like to say, What am I what have you not even thought of yet? What have you not told us yet? Like, what’s your craziest idea, and just see what comes out. And again, because it’s done instantly, almost instantaneously. It’s not burning up time, it’s actually just providing you with for one assurance, if you’ve done everything it sort of comes up with, but if there’s something that you think, ‘Oh, that’s a good idea,’ then you can delve deeper into that either with AI or by yourself through one of your other avenues.

Ross: So, for example, I’d say there’s a few differences. I mean, I’m interested in that at this point. So one that I think is really interesting is what other information might be useful. I suppose that framing that and another could be around, you know, option generation and others could be around, you know, challenges to ideas. I mean, are there any particular specific points in that overall deliberation and decision process that you think of particularly fruitful?

Paul: Yeah, I think anything which presents a counter-view. So if, for example, a group is agreeing with each other, you almost unanimously too easily then using that as to create a counter-view, why wouldn’t we do this? Or, you know, why shouldn’t we make this decision to come up with reasons against, it’s similar, again, to deep democracy where you have to argue against yourself, but you’re using technology, which has got the collective knowledge of what’s on the internet up to a certain point to go by. And that’s why I think it’s such a great critical thinking tool to support boards.

Ross: That’s, I suppose another interesting frame around that is, as you suggested earlier, you want cognitive diversity in a board. That is a bit of a problem, if you don’t? 

Paul: Once you cut the borders, don’t. That’s for sure.

Ross: But if you do have, whatever the diversity you do have, you can always complement that, you know, there are always additional perspectives, you know, when you’ve got any limited number of people, then you can always bring that to bear. So, and I suppose it is whatever views are expressed, just be able to add others or another or other viewpoints that can be thrown into the mix.

Paul: Yeah, and look, you know, it’s not foolproof, none of this deep democracy is a technique is not foolproof, because it completely depends on how it’s facilitated, who’s in the room, their personalities, their starting point. And AI is not foolproof, but it’s busy, it’s completely down to the who’s inputting the prompts. And making sure that you know how you’re interpreting the response is as unfiltered as possible. So you still got, what is the number one problem that we deal with in the boardroom, which is the people problem, you’ve still got the issue of humanity in itself. But at the end of the day, all you’re trying to do is augment our own abilities. And the best groups, the best decision making groups who will be the ones who have that level of self awareness, or self governance, who are aware of their biases are aware of their limitations, and they openly seek ways to limit the impact of those.

Ross: Yeah, you can’t. There’s only so much you can do with dysfunctional people. 

Paul: Yes, that’s exactly true. And boards can be very dysfunctional.

Ross: I am fully aware of that. Yes, I’ve got yeah, I’ve got some tails on the board. 

Paul: If you haven’t got some tails, I’d be worried.

Ross: I mean, I know that would be incredibly diverse. But I mean, how have any directors responded to introducing Generative AI into the process?

Paul: Usually to start with wide-eyed deer in the headlights to start with, like, they hadn’t even realized this was a thing. I remember about six months ago, so sort of when ChatGPT was really becoming a house, you know, household name as it is. I was doing a talk at a governance conference in the States. And I did a workshop on the use of AI in the boardroom. At the time, in my space, most people were talking about boards introducing AI policies for their company. It wasn’t about AI in the boardroom. It was about okay, how do we manage or govern the usage of AI from employees all the way through to say, school students, and that is the responsibility of the board. And if I may go down a little side avenue, still to this day, so few boards have actually introduced AI policies for their companies. So people may be using AI without any guidelines at all of how that’s being used. So if you’re listening, and you sit in the board, and you haven’t got an AI policy right now, get on to it, and featured Axios can help if we’ve got a template. 

But I was doing this talk. And you know, I was talking to people who are in the governance profession. So they were consultants working with boards, they were board directors, they were board chairs across a wide spectrum, several hundred of them. And you know, in the room, only a handful have even played with it yet. And so when I was showing what it could do, and as I said, this was sort of six months ago, so obviously things have moved on since then, as they have as they do in this world. It was just wow, I didn’t even know it could do that. And I was doing some basic stuff to start with. I had to tailor it back because when I sort of prep for this, they were going yeah, just assume we’re just all absolute beginners. So just when their response was phenomenal in terms of this is amazing. But it was almost overwhelming for them as well because they’re dealing with so much other stuff. At the same time, the role of a board and board director has become so complex now with different stakeholders. I’m different demands on your time. And they do have limited time. Whilst this is essentially, for me a time saving device to a certain degree in a critical thinking device, it’s another thing for them to learn. And boards are still trying to catch up with individuals trying to learn something. So it’s a case of just small iterative changes. The problem with AI is it’s moving on so fast, anything they learn now is almost defunct later on. So I focused most of their attention on how to work with it from a prompting perspective. Because I think that’s the stuff that’s not changing is how we prompt it, how we give it context. And if you think about what a board’s job is, it’s to think critically, we’re trained for our brain. So let’s use that bit to give it the right prompts to select about. 

I think the biggest concern from most of them was privacy concerns. But I just said don’t pump any proprietary data in there. It’s not that it’s and also they thought they could just use it like Google to search for facts. And so we were steering it, so a lot of it was education, but then it was moving forward. But again, ever since then, the response has been amazing. But it’s also been ‘Wow, this is too far for me. I’m still here, I need that to go there. I need to go a long way.’ And I think that’s because and I will say this now, the average age of a board around the world is about 60. So you don’t have people who have grown up with technology in the room. So when technology comes along, this is why board portal technology software is such a slow uptake. You also don’t have anybody who’s leading it. Who’s the one who’s in charge of this technology? Is it the chair? Is it the company secretary, is it the executive, so no one is taking real responsibility for it. And that’s the case with all technology inside the boardroom. So until we see, I think a generational shift happens, or an educational shift or a mindset shift happens, boards are going to be laggards when it comes to any type of technology. And it’s not a case of ramming it down their throat. And so if you don’t catch up, you’ll be left behind. In their mind, it’s when we still do our job. Our job is to oversee not to do the doing. So there’s a cognitive what’s the word I’m looking for? There’s a disconnect, cognitive disconnect between the power these tools have and actually starting to use them. 

Ross: Yeah, well, I mean, as we’ve already seen, for a long time, there’ll continue to be a divergence between those who are early adopters, and laggards. So related topic is, you know, I distinguish between what I describe as Analytic AI and Generative AI. Okay, I’ve been, you know, a lot of it related to machine learning. And a lot of the, you know, pre Generative AI techniques, which are obviously, extraordinarily valuable, both in implementation, but also in decision-making in the sense of being able to pick out emergent trends, particularly within internal organizational data, possibly external trends, and so on. So one of the key issues there, of course, is how has that data presented, so you’ve got to harass your business intelligence, people that come up, and hopefully the nice charts or on printouts, whatever? And so just I mean, it’s a massive topic, but just briefly, I mean, are there any insights into how we can essentially take the power of big data analyzed appropriately, which in a way, actually does usefully inform group decisions?

Paul: The simple answer is yes. If you’ve got the right people, the bright business intelligence people who know how to use these tools effectively, you’re quite right, though, like, right at the beginning of our episode, I was talking about how this information is presented. You’re quite right. And that’s the other part of the boardroom, because you have these diverse boardrooms. And although not everybody absorbs information the same way. Some people are very visual, some visual audio, so a return. And so but usually information is presented as a one size fits all approach. So I think when it comes to sort of, I suppose the boardroom, we’re not expecting them to use these tools, but we’re them to interpret the data. 

But this is where I think the delegation comes through, is that you are relying on them to give you the answer or the scenarios that they’ve done the work to get you to a point of going, right, our job is to stress test what you’ve told us because our collective wisdom, in terms of new ideas, or different cycles, we should be able to see all the potential pitfalls or things that will stop this from happening, regardless of where the information has been brought from, you know, or how it’s been developed, rather than to create the strategy in its first sense. So their job is to pull it apart as much as they possibly can. And then the response of the business analyst is to go, ‘Okay, this is my response to your query.’ That’s how it works best. That’s how the boardroom works best. 

Now they can use their own tools to help pull things apart. But that’s really where they add value, if that makes sense. So it’s very much that she did that Human Critical Thinking element, rather than utilizing that stuff. That’s definitely an operational level using those sort of Analytical AI tools. I would never see a boardroom at this point, going near those — be too much. What I would love to see happen with this technology as a way in which this information can be presented in ways which are semi personalized to the board members. I’ve seen such diverse groups, some who have ADHD, and the way they absorb information is absolutely diametrically opposed to somebody else. And yeah, as I said, everything’s presented in the same way, I would love to see it whereby some sort of, you know, inputs are put into allow board, especially using technology with more and more board papers, for example, which is the information packs being presented from through technology, as opposed to the old school printed PACs, which I still some do that like printed out folders of information for each board pay people, I think having those personalized in some way would be a real step up for the intelligence the collective, what we call BQ board intelligence of the group.

Ross : As in thriving on overload, a lot of where I look at is how cognitive styles and how we take in information and, you know, big believer that we you know, we can certainly not one size fits all for how we take in information that can be very much tailored. So your organization is called Future Directors. And so let’s go into the future. And looking at the potential, what is the potential of the future?Let’s, let’s imagine, what a future board could be, amplified by technology. I mean, let’s, you know, it’s a matter of finding the right, maybe finding how it is we find and assess and bring together the right people, what is the what, how do they interact? How do we get an absolutely better future board than we have today?

Paul: Look, I think, technology augmented boards, let’s assume that the next 20-30 years will still be human boards. We’re not going to see sort of, you know, super intelligent AI taking over, let’s assume that and boards are still relevant in that respect, I think the level, the level of decision-making, they’ll be doing will be elevated, they’ll certainly be using technology to support their decision making, as we’ve talked about. And that’s only going to get easier. And I think what we’ll start seeing though, is boards meeting not on schedule, but by design. Because the world is moving too fast, the technology will be presenting information so quickly, boards will get together to make critical decisions and have time to absorb that information. 

I actually think the power of technology is going to be in the monitoring, and the learning element of how decisions were made, how people interacted with each other. And I think this is where AI will be able to play a role of observing in some way, let’s say observing, and playing back how the group worked, how the decision was made to stress testing, the After-Effects and actually measuring the performance of a board. I take you back six years to the same conference that I did six months ago in the States when I first presented the future of the boardroom. And this is before any of this gen AI stuff was coming out and I was talking about how AI may if we’re all plugged in . I went onstage wearing one of those muse headbands that monitors brainwave patterns. And I was working with a group at the time and I was demonstrating how my brain waves are working and how the majority of us are in beta, which is that fight flight freeze response. And we’re masking it with coffee and all that sort of side of things. But we want to be in alpha focus and theta knows once from a creative, intuitive element. And you can actually start to measure the cognitive ability of the individual any one time so if they’re really tired, they might be operating at a lower low level. 

And I was supposing that board members will be plugged in and be monitored. If you are not operating at a certain level, you will be excluded from decision-making. Literally in that meeting. So you’d have to learn breathing techniques, meditation, all the things to actually raise your cognitive level but some people come into the room. Take it, take a spirit example. But hungry imagine coming into the room, feeling so hungry that you’re just unable to focus. And yet you’re expected to contribute to multibillion dollar decisions or decisions affecting hundreds of thousands of lives, but you could be distracted. So I think this is where technology scares the bejesus out of everybody in the room, by the way, which is most of the fun. And everybody thought it was like Terminator, you know, gone crazy type of stuff. 

But they could get the idea of what I was trying to get across is their duty, their legal duty is to turn up at their very best because they’re being recruited for their brain, they’re no different to an athlete on race day, or game day, you know, body mind has to be in tune. So why is it any different for the boardroom, and you get found out, you don’t qualify, you don’t win the game, why not different for the boardroom. And they use lighting technology more and more and more, and coaches more and more and more, to ensure that on game day or race day, they are at their peak? Why not this, what I was trying to get across is why not the same for a boardroom that is tasked and legally responsible for that level of decision-making. 

And I think the stick and the carrot for this sort of thing will be stakeholders. So the owners, the shareholders, employees, the more because boards become transparent. And it’s only really in the last 10 years through shareholder activism that we’ve seen, people take the board on, and actually start to see them as something that they can have a look inside of as opposed to this, you know, top-floor behind closed door secret decision-making group. The more we see transparency, the more they’re held accountable by external internal stakeholders, the more they will have to do to prove that they’re doing everything they possibly can to work and work at their best. Now, I’m not saying this is all good. I’m just saying that’s where it’s heading. I think technology is going to both support them, but also work to stretch them as well.

Ross: Well, that’s, I can hope that you know, some of those, I think some few really important elements there. And the thing is, yeah, I always wonder at how dysfunctional a lot of boards are, or not all of them. But there’s a lot of expectations, transparency. But if there is this focus, and I think what you just were saying really brings this out on the fact that they do need to have outstanding cognition individually and collectively. You know, these are critical decisions and critical roles. So, if we start talking in the language of amplifying vision of boards or whatever language we use around, I think that’s a fantastic step forward from where we are in any case. So where can people find out more about your work?

Paul: So futuredirectors.com, they can see what we do as a business. And then if you just Google or future Paul Smith, or go into LinkedIn and do future Paul Smith, my name Paul Smith is quite common. So add the word future in there and you will find me and you can connect with me. Follow me to see some of my work as well as the work of future directors. Fantastic.

Ross: Fantastic. Thank you so much for your time and your insights. Paul.

Paul: Great to be here.