in the future, the AI is going to be so clever that you don't have to engineer the problem. Like it has all the knowledge and if it doesn't, it's going to go out and find it. So I'm actually, I think on this podcast, I think we should coin a new term in the spirit of asking beautiful questions that maybe should be question architecture. Welcome to the Changing Conversations podcast, the podcast that brings you perspectives on how to develop a new concept. to harness the power of conversations to make change happen. Today's episode is about the types of conversation and the way that we are in conversation when it comes to artificial intelligence in our organizations. So we'd like to explore the conversations that already feel broken when it comes to AI, or at least feel like small conversations compared to the bigger potential that we, we could be talking about. So that's what today is about. And we're not going to go into a technical space of the tools and the chat GPT and all that sort of stuff, but trying to hold it in a bigger space. So as usual, I have Sti with me. Hi Sti. Hi. And we are also joined by Adam Hill. And I probably said your name wrong. So why don't you introduce yourself? I was close enough. My name is Adam Hill. Or Adam Hill. I think it's okay in English. That's good. Great. So Adam is an advisor to organizations who want to take advantage of AI and to use it to take their business to the next level. And you're a colleague of ours and part of a larger team that works within this space. So welcome, Adam. And as we go into this talk, we're kind of each coming from different perspectives. Of course, Adam, you are coming from the AI space here. Sti, you're kind of bringing in the strategic and board level conversation here, as I see it, at least. And I'm coming from a background of leading in uncertainty. And I also have permission to ask all the stupid questions because I really do not feel familiar in this space, which I think is something that we're going to touch upon because I'm not the only one. But Adam, you've been watching this space for some time and you've been having some opinions about the conversations you see happening. So tell us what you see and what you'd maybe love to be seeing. I think the really interesting engagement and why we, at least I wanted to have this talk, was that I think I've seen more and more businesses ask the same questions. Like when we engage with clients, a lot of the clients, they initially ask, so like, what are the good use cases? What are the business pains we can alleviate? And it's sort of, as a consultant, that seems pretty natural. We all often look for like, what are the benefits we can create and what are the business case? But it leads to a lot of looking for efficiency and looking for cost reductions and things like that. And that is not to take anything away from cost reductions, but I actually think some of the most striking things we've seen in AI has been people asking somewhat bigger questions. Like, what could we do? Instead of asking, like, what problems could we solve? But asking things that we might not really think or imagine. And some low practical things are, and this is going to be a super big cliche in any kind of business space. But to say, like, what can we do for our customers? Instead of what can we do for ourselves as a business? What needs do our customers have? What needs do our suppliers and partners have? How can we be a better partner? Already that is a slightly bigger question, I think, and I hope people can ask even bigger questions. Because the technology can actually carry that today to a large extent. But a lot of people don't seem to realize. And changing that conversation is hard. What do you see, Steve, when it comes to this topic out in the world? I think it plays into conversations in general in the sense that for many years, the challenge, the real challenge within shaping courageous conversations is to inspire people to ask questions. Most conversations are filled with exchanging viewpoints or trying to convince the other part that you're right. And I think the new thing in AI is that you don't get anywhere unless you ask beautiful questions. So maybe it's the return of the questions for the general conversation in the world. I really, really hope so. But so if you are saying that organizations, if leaders should be asking different, bigger questions, you know, how are we, how would that look? Where would that take place? And what will it take to have that conversation? I think it starts by people realizing what I am proposing, what is it that I'm actually requesting. And I had a really profound experience with one client that when we talked to them and when they realized what I was trying to get at, they almost shouted out like, you want me to rethink my business model? Like, is that what you're asking? And I paused and I said, yeah, I actually, I actually think you should. And I think they were in an industry where things are changing around them to an extent where what they're doing isn't going to work. It isn't going to be feasible. So yeah, I think you do need to rethink your business model. And that wasn't a conversation they were ready to have at the time. So today, like I'm, I'm still a young consultant in many ways. I don't know if it was the right thing to tell them or if I overstepped or anything, but it was what I felt was right. And those are the conversations I find myself having because they have this insight and I meet business leaders in this space. And I think it's super interesting to just reflect a little bit on sort of what is our role as advisors in that space? What role does business have and how do you have that conversation? But can you, can you, can you hold that conversation? Can you lift the conversation to where you want it? If you don't, if it's not a territory that you master yourself is, if there's no depth in your own insight, how, how, isn't it going to be shallow? If you raise that, what's the biggest worthwhile problem we could be part of solving kind of AI wise. And it's just evident that, that, that that's just something you throw out there, but there is no, you don't embody any of it yourself. I'm, I'm just think, I'm just thinking about this feeling that I have that many of us don't feel worthy of raising these questions yet because we don't, we don't own them yet. And we, we still feel uncomfortable. Totally. And I, and I think that was what happened because my, my script ended right there. I said, yes, I think you do need to rethink your business model. And then nothing. Yeah, exactly. Like that's, that's where it ended. But I think one thing to put into perspective. So recently you did an event where you were talking to an expert called Benedict Evans and you guys were talking about how, how old this story is. Like, as in, uh, I'm not going to embarrass myself by saying whether it was one, two, three years old, but, you know, tell, tell me, Adam, how long has this been something that we're taking into consideration? Because this feeling of being uncomfortable, this might have something to do with it. So what is the relevant history of these questions we're asking now? I would love to say that it's, ah, this goes way back. And there's a lot of people who, who likes the deep history of AI. I think the truth is that, and that's part of the reason why things are so difficult is that we are coming up on the second birthday of ChatGPT. Yeah. Which were really everybody's introduction to it. Yeah. It was launched in the beginning of November, two years ago. Um, and while there is a lot of cool history before that there, that is really a before and after moment, because that was when everyone from the people working out in the organization to the C level got their hands on a completely new kind of AI technology. And that also means that we have had two years to think about it and two years to form opinions. Uh, and I've learned as I learned that changing things in business takes time. Um, so this is also, it's, it's not an, a cry from impatience on my side. I don't think there's any need for that. I actually think the world is moving plenty fast right now. Maybe, maybe a little crazy fast sometimes. Um, but it is, it is a two year time horizon. So it's okay that the conversations are changing slowly, but I think people should be aware that they are changing. But also this feeling that we don't know, like, I don't know what that really is. Or like, you know, that you were speaking to Steve, where senior leaders and pretty much everyone for that matter are having to make choices and to navigate something that we're all just trying to figure out what is. And what, and how that actually goes against some ways that we historically have been having conversations where the ones who are at the top are perceived somehow to have the answers to things. Yes. And so actually that puts those people in a maybe even extra uncomfortable position. I will just make the hypothesis at least, you know, so what, what will it take from those individuals who maybe are feeling that I'm being called upon to make choices and to, to, to figure this out. But I don't really know, like, I don't, am I using the right words? Am I, you know, asking the right questions? What will it take of those leaders to, to, to do the right thing? What is the right thing? Yeah, I think it's difficult because you're right. There is this sentiment that the higher up, the more I should know. But I don't think that's the truth. And I was with the CEO of a very big Danish corporation last week. And he, he said to me that his big idol right now was Socrates and the whole idea about what is it that I don't know? And what is it that we don't know? And what is it that we don't even know we don't know? And I think it was actually, it was actually a beautiful moment. And it was a very rare moment that you hear a CEO having the courage to take that perspective. And I think maybe we need to go to that space to open AI up for real, the way you, you want us to open it up, Adam, this thing about that I can lead with curiosity. So I don't need to know where the conversation should go, but I should trust the conversation and the organization. And I should just lead with beautiful questions that inspire everyone to imagine. I think that's a really cool and good idea and a good way to put it. Because I think leaders, nobody knows where this AI train is going. And nobody, a lot of people have different hypotheses, but I think there is a lot of uncertainty out there for good reason. I consider myself an expert in the field and the best I can do is maybe just say no some about the questions we can even ask of the technology. And that's being said, I think there is a role for leaders in leading in that uncertainty. And part of that is listening to the organization or just maybe making decisions on incomplete information. I don't know what that looks like. And maybe because I'm more of a technologist and a leader myself. But the role of leadership in that world, I think is super interesting and super hard. Because I don't think leadership is exactly the same as it were five years ago. And if you are being forced to make decisions in AI right now, you are at least leading into greater uncertainty than you probably ever have done before. At least within my lifetime. And how you handle that and manage that as a leader and what is the responsibility of leaders in that space. And when are leaders allowed to just throw up their hands and say, I don't know. I don't know what that looks like. No, and maybe exactly. I mean, it's fine to say, I don't know. And especially if you can feel empowered from that position rather than disempowered. Because the I don't know holds an invitation to the rest of the world to play along. So maybe what you need to do, especially in this space, and maybe we can practice this in this space, is that what you can do is you can ask big questions to the organization. You can inspire them to imagine. And you can create a space for imagination for a period of time. And take away that short-termism and that, you can say, stressful feeling of we can't think about the future because there's too much to fix today. So that you can do. And I'm just, a question to you, Adam, in that space. Can you ask these questions? And at the same time, view AI as something that's chauffeur-driven? Meaning, I'm not really touching it. I read about it, but other people in my organization take care of it. Can you ask beautiful questions in this space without really getting into it yourself? I don't know, speaking. And I don't think anybody really knows yet. I think it has to be like that in many organizations. Because leaders can't know the details of what's happening in AI because few or maybe nobody does. I think there is something to being a leader that empowers. I really like the idea that you guys are putting forth that as a leader, it's okay to say, I don't know, and to ask the interesting questions of your organization. And I think a lot of leaders would find that the organization responds really nicely to it. And we know that ideas for what to do with AI comes pretty easily. Maybe the conversations are a little small. It's a lot about, again, the business pains. The organization, if not prompted to it, quickly stares into its own inner workings. And that's what organizations care about day to day. But I think a good leader could prompt a bigger conversation. I think what then needs to happen, but again, I'm stepping into other people's realms here. This is where you guys are the experts. But I think there's something to leaders then empowering these ideas. Because if you are just a regular employee somewhere who actually has a profound idea, it's super hard to get anywhere. Like getting that support from leadership and how leaders pick out the good ideas is maybe the hardest thing that I don't really have the answer to. Like I sometimes propose maybe you just throw dice, pick random. Like I don't know. And leaders, they don't respond well to that, I can tell you. No, but to get anything going in an organization, sometimes you need to ask questions, inspiring questions. But very often you also need to challenge the organization with insight. So you need to bring a novel insight to the table that kind of disrupts the way we do things, the way we think. To kind of say, wow, we need to think differently about this. Or how should we think about this? And if you don't have that insight because it's too abstract for you or you're too fearful even looking at it because you feel insignificant and unworthy, then it's difficult. Or the consequences of what you might find. Like I totally get it from that leader that said like, do you really want me to rethink my business model? Like this was in an industry that if, as I left the building, I thought, yeah, this is really, really big. This is really, really going to change. If you don't, then somebody else will. Exactly. And then you'll be thinking about it, but maybe too late. Exactly. And I totally get, as a leader, why that not be what you wished for. Do you think they did? After you said, I think you did. Did anything happen, Adam? I don't know. I really don't know. Maybe you give them a call. Yeah, that's a good point. But I, so what's interesting here is that, so we've gone through a series of elements in this conversation where one of the really nice things is that even at a tool level, one of the things that you need to learn how to do is prompt engineering, which as far as I'm aware of is coming up with good questions. So, so maybe that is the first thing that can be applied on multiple levels here is that, you know, it's about learning to ask good questions to your colleagues, to your organizations. I don't know the answer. What do you see? How are you using AI out there in the fringes? What could we try out? Or what might our customers need now that we can do this or that? So, you know, maybe those are some examples of questions. But the other thing that comes to mind is that we're talking about, okay, then you've got your options in front of you. We could do all these different things. And then you say, okay, roll a dice. Maybe another way, if we're learning a little bit from the design thinking way of going about things is maybe don't make a big bet. But instead, what are the small experiments that you can do? What are the small things that you can test out to see what that does? That also, speaking to the fear that people feel, that's another way to think about this that maybe takes out a tiny bit of that fear is if you're just placing your bets many places. Yeah, that comes to mind. I bet it's a, once again, a both-and in the sense that if you only do that, then you're probably going to have insignificant wins that will fuel the skepticism anywhere. So it's probably a matter of creating a movement where everyone plays with this and at the same time identify the bench-breaking business model transforming things where you need to put some hard thinking into how do we get that going as well. I think one without the other. I don't know. What do you think, Adam? No, I agree. I think there is something to say for doing lots of experiments. I think businesses fall on a continuum. And as I spoke with Benedict Evans about that, some businesses are likely not going to change a whole lot. And then there are other businesses that will change completely in light of what's happening. And knowing where on the spectrum you are is pretty tricky. And I think it's about trying to start to test out and figure that out. I personally really like doing a lot of experiments and trying many things out. But I think an important thing to keep in mind is that while the experiment can be small, at least some portions of your experiments needs to have big ideas behind them. And I think that really gets to what Steve is saying. It's not because I'm asking anybody to put $100 million into some kind of huge test. It doesn't have to be like that. But at least some of the ideas need to be $100 million or $100 billion or whatever. Big number you want to put in front of it. Ideas where you know it's risky. You know it's going to take a lot more investment to get there. But I think more businesses would find pretty shocking things if they started going down some of that path and figuring out, okay, some of these assumptions they actually seem to hold. And that is pretty scary. I don't know if it's scary. Like to my mind it's exciting in many cases. But a lot of business leaders I think find it a little scary. Because you didn't necessarily sign up for this level of change when you became the CEO or the CEO or whatever. Business as usual is oftentimes more comfortable. So a lot of the questions that, you know, when we talked about prompt engineering, the good questions here, the ones that I mentioned before about, you know, I don't know, but how do you see it? I feel like those were good questions for gathering data from a very wide source, from your whole organization and listening to that. But then we've been talking about something that's a little different, which is the other side of the both end, which is when to go for that big, you know, risk-taking move, that big investment. And it just makes me curious about how that conversation looks as well. And whether we, you know, what have you seen, Steve, that maybe we could offer as some advice to people that might find themselves in that situation? I can think of two things. One is that you don't start with a big investment. You always do that as an iteration. So you start with a little thing where you explore the main, you could say, assumption around that idea. And then you move your way forward like we know it from startups and that field. I think that's the model. And the other thing I think is important is that you, if you're going to make a big bet, I think very often you have to go to understand what is actually our deepest core capability in this organization. What is it that we, what's the core differentiator? What makes us unique? What is the main explanation why we win in the market? And then start exploring that field and see if there's a way to 10x that. So, so, so you, you, you actually have to really, really find out what makes you unique. And then starting, starting to look there versus just go with something because there's a big chance that it's the same thing everyone else will be looking at. And it'll be a catch, catching up game rather than a, wow, I wish we had. And it's also difficult for others to mimic afterwards because it's, it's, it's actually tied to deep things that are ingrained in your business model and your management model. So that is very difficult for others to, to mimic. I think that makes a lot of sense. And I think that is some of the pieces to the puzzle that people need to think about when they're asking these bigger questions. I'm not sure I really have much to add. Like, I think that's why it's so interesting to be among different people, whether that is other colleagues or the clients who are thinking differently about these different kinds of things. I have, I have, I have something, I have a question for you, Adam. Go. It has to do with the word, artificial intelligence. But I could also mention prompt engineering. So I have a friend, he's a poet, and he says that one of the problems with the word, for instance, feedback, is that the word holds very little invitation. So literally, it's very, it's difficult to hold as a human. And that's why no one wants to give feedback and wants to receive feedback. Because the word in itself is very, not very invitational. And I've been thinking for myself that I have a little bit of the same sense with artificial intelligence. There's something fake about it from the get-go. Yeah, why should I trust that? Why should I trust that? Why would I make bets on artificial intelligence? And then next word you throw at me is prompt engineering. And that just... You just completely shut down. That's not very invitational. And you're not telling me that the RAG architectures or vector databases or API... It's not doing it for me, Adam. No, it's not doing it for you. No, I actually... I think it's an interesting question. And I am actually thinking... I'll touch on AI in just a second. Because I want to grab onto prompt engineering first. Because that is a strange field that came out of nowhere. And some people are already arguing that prompt engineering is going away as well. Because the model makers, they're saying, well, in the future, the AI is going to be so clever that you don't have to engineer the prompt. It has all the knowledge. And if it doesn't, it's going to go out and find it. So I'm actually... I think on this podcast, I think we should coin a new term in the spirit of asking beautiful questions that maybe should be question architecture. Yes. Because I think more and more people are looking to these AIs to say, you need to ask... It only works with information. So far, there's not a whole lot of AIs connected to robotics just yet. So you aren't using it just to do things. Most profound moments with AI is when it gives you an insight. And the most best way to do that is to ask questions. And questions are something different than just a prompt. So I think in the future, especially as these models get more and more clever, I think actually designing those questions, architecting questions. I sometimes, when I work with AI, find myself asking one question, getting the answer, and then realizing it was the wrong question. But that's interesting because we've been geeking out in the field of questions for quite some time. And one of the interesting things or one perspective on that is that if you ask, and I'm just using that word now for lack of better, but if you ask a beautiful question, one of the definitions of a beautiful question is that just by asking it, you are transformed. So you're literally, when you're in a real conversation, you're overhearing yourself say things that you didn't know you knew. And that's interesting because you're basically making an invitation to rethink the whole field of what you ask and how you ask. I actually think, and this is going to get a little into the geeky sides of whatever we want to call it, prompt engineering or question architecture, but it's one of the things that I use AI for myself. Like you can ask the AI, what is the question? Like, is this a good question? Like, do you understand this question or ask me questions? And sometimes you get really far by having that sparing. And I think at this point, the only unique thing about that is that you do it with a machine. I think this is something that humans have always done. And it's some of the best conversations that we have. And the only thing that's why AI is relevant here is because it's the first time where you can do it alone together with a machine. And that's profound, but it's also, it's not such a profound thing from a human experience thing. Like, I think those are the great conversations we always had. Just looping back one, could you ask AI the question we were talking about before in terms of about the business model and strategic level? And, you know, could that even be a starting point is to go on to, would it be chat GPT and say, how will AI influence my industry? And what are some ways in which, I don't know, like, I guess you could actually do that. You totally could. There would be one caveat to it. And that is that most models that we have access to right now is intelligent, but they're not super intelligent or anything. So whenever we try that, the answers we get out, they are, they're cool. Like they make sense, but it doesn't really impress anyone. Could be a starting point. Could be a starting point or something. That's how we use it sometimes. It is a good way to get a basic understanding of a business or something. I will add to that, that the frontier models, the next generation that we have access to, that requires a little bit more special things and whatnot to use. Things are getting different. Like I've had at least one instance now where I put in a big business model into one of these because it came from another source and I, I had a feeling that this was a little rough around the edges and I just, out of testing, wanted to see, okay, what, what does the model think of this? And what I got out was pretty profound. The big problem at that point was that it was in an industry I didn't know. And the level of feedback the model gave to the business model and the proposed changes, I couldn't evaluate. Like I didn't know. So of course it wasn't super helpful because I still don't trust the AI enough to just run with it. And I think we're far from that. But you could bring it into a conversation with all the people who, you know, it will take a conversation with a lot of people to have that type of discussion. So it again could be a starting point. But I guess the premise is that you come in then saying, hey, I asked AI, here's what it said. And that, that almost is a great starting point because people usually love to correct stuff or say stuff is wrong rather than starting from a blank sheet. So, so, so that's kind of interesting. But let's, before we close off, go to the last part of this conversation that we wanted to have about artificial intelligence. And so if we've coined then the term question architecture, what, what should, what should we do with our relationship to artificial intelligence? I think that is largely going to fix itself. I think to, to the last points on the, on the wording AI, if you look up in the dictionary is machines that does human like things when it does things that we typically associate with human cognition or human capability. And as such, it also means it's a moving target. And I think we're already seeing it a little bit that once something becomes natural to see from a machine, like this is obviously something a machine can do. We stopped calling it AI. Like at one point calculators were AI and I'm still somewhat frustrated and amazed that people aren't more impressed that they can search for images on their phone. Like your phone can find pictures of your dog and it's incredible, but people are like, of course, it's just pictures of dogs. Like who cares? And I think in a few years, in two years into this, everything has just been called AI because there's an explosion of technology. But I think in five or 10 years, we won't be talking about chatbots as AI. I'm not even sure we'll be talking about chatbots, but we'll be talking about something else because chatbots just work. And in a few years, the fact that you can talk to a machine and it replies back to you in natural language will be the most natural thing ever. And nobody's going to care about it or be impressed by it. And we'll call it something else then because AI is always the thing that doesn't work. It's always the thing that's on the edge of what's possible because it's where machines eat away at human capabilities and humans keep moving, fortunately. So does the machines. Wow, really interesting. Thank you so much for joining us for this conversation. And we're going to end it the way that we normally do, which is that each of us share one thing that we'll take out into our next conversation based on the one that we've just had. And I don't know, I think for me, I'm just I'm just quite grateful because as I said at the beginning, I don't feel that I'm an expert in this. Actually, I feel more the fear box that you sort of brought into the picture, Steve. And this made me feel just a little bit more confident in the fact that I will figure it out the way that this is relevant and maybe just asking a few more curious questions in this space. So I'll go out with that. How about you, Steve? Yeah. Yeah. I actually, I was a little bit inspired by something that you didn't say, Adam, but that I maybe thought about listening to you. And that was, I just got the feeling that you have a more intimate conversation with AI than I have. I'm still trying to be clever and where I just sense that the conversation you're having is much more vulnerable in the sense that you're asking for help and you are asking AI to help you. Whereas I think I need to get it right, to use it, to be worthy even to use it. So I think in what you said or maybe what you didn't say, there was an invitation for me to create a more intimate relationship in that conversation that I think will be more fruitful than the one I'm holding right now. I think that's super cool. And I definitely do recommend it. At least it doesn't hurt. And it's fun and interesting. How about you, Adam? I think I'm taking something away on how we're going to have these conversations. I actually think I learned a lot today on that moment back when I spoke to that business leader who said, do you really want me to redesign my business model? And I just said, yeah. Like, what is the follow-up for that? I think some of the things you said, Steve, on listening to the organization, some of the things you said on how to run experiments and keeping things small while still being backed by big ideas, is actually going to enable some better conversations. I actually think I, maybe even more than I hoped for here, is actually going to live with an ability to have the conversations I wanted to have. So thanks for that. I remember one of our cool old friends, Gary Hamel, said something like, revolutionary dreams and evolutionary steps. That is good. Down that line. That is good. He's good at those. He is. All right. Thank you so much. And thank you for listening. How did this conversation resonate with you? Get in touch via the link in the show notes to share your questions, challenges, and opinions. And until next time, take care.