Digitally Curious
Digitally Curious is a show all about the near-term future with actionable advice from a range of global experts. Order the book that showcases these episodes at curious.click/order
Who is your host, Andrew Grill? He’s the AI expert who speaks your business language. After 30+ years building tech solutions at companies like IBM and a range of high-tech startups, Andrew now helps executives navigate AI without getting lost in the complexity.
He has held senior leadership roles, including Global Managing Partner at IBM, and has collaborated with C-suite teams from organisations such as Shell, Vodafone, Dell, SAP Concur, Nike, Nestlé, and the NHS.
Andrew has delivered 700 engagements in over 35 countries on topics such as generative AI, quantum computing, digital transformation, and the future of work.
Ranked among the world’s top 20 futurist speakers and a finalist for AI Expert of the Year, in 2025, he was recognised on the AI 100 UK List as one of the country’s leading voices in responsible Artificial Intelligence.
He is the author of Digitally Curious (2024), a bestselling guide to navigating the future of AI and technology, and host of the Digitally Curious Podcast (since 2019), where he translates complex trends into actionable insights.
Andrew is a regular media commentator, featured on BBC Television & Radio, Sky News, LBC, and in publications such as the Financial Times, The Guardian, and The Economist.
Find out more about Andrew at actionablefuturist.com
Digitally Curious
S8E6 - Trust in the Age of AI with Mike Stead
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
My guest is Mike Stead, founder of the Workplace Forum, where C-suite leaders talk honestly about AI under the Chatham House rule.
Mike and I reconnected when he commented on one of my LinkedIn posts, asking whether I had written it with AI. That question opened up a conversation about trust: in what we read, in the data we feed AI, and in each other. Mike shares what senior leaders really say about AI when the door is closed, why one Fortune 50 company handed back 100,000 AI licences after six months, and why, in his words, "knowledge can be automated… accountability can't."
Key Topics Covered:
- Why AI writing all starts to sound the same, and why that makes original thinking more valuable
- Building trust through stories grounded in lived experience
- The danger of running AI on data you can't verify, and why leaders must be able to "show their working" to the board
- How the Chatham House rule lets competitors talk openly about their AI challenges
- What leaders say about AI in private that they won't say on LinkedIn
- Why saved hours may be the wrong metric: focus on what you love and automate the rest
- Using half a million words of meeting transcripts as an AI coworker
- The case for a virtual AI board member as a decision partner, not a decision maker
- Why AI will demand better humans
Mike's Three Actionable Takeaways:
- Challenge yourself. Be up for discovery, even when the future feels uncomfortable.
- Talk to other people about what works for them, how they see things changing and what they would do differently.
- Have a plan to act. Schedule the time, do a course, take the first step.
Recommended by Mike:
- Next podcast guest: Richard Mulholland
Where to Find Mike:
- LinkedIn: Mike Stead
Thanks for listening to Digitally Curious. You can buy the book that showcases these episodes at curious.click/order
Your Host is Actionable Futurist® Andrew Grill
For more on Andrew - what he speaks about and recent talks, please visit ActionableFuturist.com
Andrew's Social Channels
Andrew on LinkedIn
Andrew on YouTube
@Andrew.Grill on Instagram
Keynote speeches here
Order Digitally Curious
Welcome to Digitally Curious, the podcast that will help you navigate the future of AI and all things tech with your host, actionable futurist, Andrew Grill.
SPEAKER_01Today's guest is Mike Stead, who founded the Workplace Forum, where C-suite leaders talk honestly about AI under Chatham House rule. We recently reconnected on LinkedIn when he commented in one of my posts about AI and consultants that ended up quite viral by asking if I'd actually written the post with AI. Welcome, Mike.
SPEAKER_02Hey Andrew, good to be here.
SPEAKER_01It's good to be back with another Antipodean. And for reference, we connected probably 10, 12 years ago when we were both mucking around with this thing called mobile phones. We've both moved on since then, but it was a fun time. But take me back to that comment. The comment was, did you use AI to write this, Andrew? And I think you have a wink face. What made you ask it?
SPEAKER_02Going back and looking at it and rereading it, it wasn't that the content was wrong. And AI generally doesn't get content wrong. It was more that you see that structure and even subconsciously you start to think, have I read this article before? And you think, no, actually, it was it was the way that it was laid out, the pros of it. And so AI, I think, is getting really good at creating something that feels intelligent or that you know looks at first glance. Um, and but why it feels the same is is it because it follows, I think, what you'd call a rhetorical pattern, you know. And so weirdly, the more credible AI sounds, um the more valuable genuinely original thinking becomes. And so it's it's not that the risk itself is an AI, it's that that everything starts to sound the same. And that's that's that's what I kind of thought that I was seeing there.
SPEAKER_01I mean, we talk about AI slop. I get this question all the time on stage, and I I think I echo what you say that we're gonna get to a point where the average person in the street says, Oh, I'm not another AI post. Give me something from a human being. And we I wrote in my book, Digitally Curious, I wrote about the magnet of mediocrity that there'll be so much AI slop. We didn't call it AI slop three, four years ago when I wrote the book, but everything will look and sound the same. You'll be craving for something original, and it'll be obvious that a human wrote it, but I think we're still in the build-up to the slop at the moment. Would you agree?
SPEAKER_02Our youngest uh child, our son, he's autistic and he hates AI. He passionately hates it because of the whole idea of slop and the lack of originality and things like that. Um, you know, to the point that my wife and I are not allowed to use it in our family WhatsApp chat. And and I think if if he's seeing that as a 16-year-old, you know, you've got to think what it what are what's everyone else thinking. And I think there's only a you only get a ch the chance to offend or mislead someone once, and then you've burnt that bridge. And I think it's something that far too many people don't appreciate, you know, and so protecting your trusted voice, I think should be the number one thing that everyone should be thinking about when they're posting.
SPEAKER_01So we're gonna talk a little about trust because that's something you're zeroing in on, and that's something that's important to me. And in fact, when I got your comment, I came straight back and I said, yes, I had used AI to formulate that post, and part of it was I was using a new tool called PostIV, and that's a whole other discussion. But I I paused for a moment, I thought, what will other people think if they think I made this with AI? So I want to be transparent that I did, but we're just talking off-air about the fact that if I'm now talking on this podcast, if I'm on stage, same with you, and someone asks me a question and I don't have any notes, that's when you actually have to basically not wing it, but you have to go back to what you understand and you can't just use AI. So where does trust fit here? And we might work in your workplace group where you have Chatham House rule that is a trusted environment. But but talk to me how important trust is in business generally, and also how do you evaluate trust?
SPEAKER_02This is really interesting. I'm actually doing a bit of a course on this right now. You've got to be completely genuine. And and I think the easiest way to be genuine is to tell stories. Um, you know, have a narrative hook that is grounded in human reality around something that someone told you. The person who's listening to you is then trusting that what you are talking about is from a personal lived experience. So I could talk about the fact that a Fortune hundred-level workplace leader, you know, told me that one of her biggest challenges is not uh not getting data around the workplace. It's around understanding what to believe, what to trust. And then she has to communicate that to her board when they are making decisions around, I don't know, how many more millions of square feet of real estate you know she's she she's recommending that they lease. Now, when I talk about that, and I can go into a lot more detail about that, it's because I'm talking to it from a personally lived experience. I sat down with her, I heard her in a room with her peers talking in this way, talking passionately, asking for help, asking for other people's advice. And when you talk about that thing, hopefully this did come across as completely genuine. People go, right, I trust that Mike said this, and he's obviously lived this and he can backstop it, and he's not just made up something and he's hoping that it will land so that they'll buy something that I'm selling, or something like that. So it's got to be rooted in the human emotion and something that you can come back to. And I think humans are very, very good at subconsciously filtering out something that is just a story from something that is actually lived and real. You know, so when I say story, I mean something someone has made up fabricated just because they want to sell you something or change your mind.
SPEAKER_01And it's interesting on a couple of levels there, because I get asked all the time, you know, do I do I believe what I see online, what AI is telling me? And I'm saying, well, you've got to understand that for a number of years now, we haven't believed everything on online because some of it's made up. And then I say, look, if you're using AI within an organization, you're relying on data that's has a source of truth, you should trust it. But what you're saying there is even within an organization, they're questioning the source of the data that's inside the organization. So riddle me this. If I'm using AI on data that I don't know is accurate, is that as dangerous as just believing what we read on the internet?
SPEAKER_02If I have to Google one word answer, yes, um, it is. But obviously there's a lot of nuance in there. Um this is something that uh I remember talking to one, um, this was actually in a sustainability context, and this one sustainability leader, you know, for one of the world's largest engineering firms, um, she operates in 40 countries. Um, she is pulling in sustainability data into at least nine different data lakes right now, um, you know, around workplace um statistics, power consumption, um, scope three reporting, all all all manner of things. And she's trying to understand how she can compare and contrast data that she has she has no reason to disbelieve. You know, if she gets a a printout from a from a manufacturing facility that says we've used X kilowatt hours of energy or X, you know, whatever BTUs of gas, those are true. You know, those are those are facts. Then in between those facts and the analysis or judgment that her board is looking for, there is um you know this this this work that has to be done. Now she can go and ask an AI, analyze these numbers and come up with a conclusion, come up with trends, et cetera, et cetera. She has to be confident. She has to trust that what comes out of the AI sausage machine is actually accurate. And that is a major challenge for enterprise workplace and sustainability leaders that I talk to right now is when your CFO or COO or whoever challenges you in the boardroom on your strategy, how do you show you're working when it was your AI agent in the company that actually did all of the maths? And this is a very very, very pertinent topic for people because the sustainability or workplace numbers around utilization or power consumption or whatever, they are driving strategy decisions that are now expected to hold the same credibility as financial decisions. Now we've got 150 plus years, 200 years of enterprise finance best practice. And when someone looks at a number on a spreadsheet, unless someone has actually cooked the books, you can generally go back and find the working. We aren't there with AI, you know, and that's that's a real, real challenge for the AI industry and for people trying to implement this kind of tool is how do how can you go and reverse engineer if someone challenges one of your assumptions or one of your recommendations?
SPEAKER_01That's a really interesting point, the whole check you're working, and that's about explainability and observability, which is going to be in Australia law fairly soon, that you're gonna have to stand back and say, Well, I can explain how this black box works. But let me just step back a bit because the groups that you have organized, I find fascinating the fact that you bring people together under Chatham House rule, and you might want to explain for some people what that is and the impact of that, but also you've got thousands and thousands of discussions in there. So you can kind of reference that. You're not identifying who that person is, but you've got this, I suppose, almost off-the-record discussion. Talk to me about that, and then I want to ask you a specific question about AI and what people are saying in the room versus what they're saying publicly.
SPEAKER_02The Chatham House rule, for people who aren't aware, Chatham House is actually a place in London. I forget what its official name is, the Institute of Something or Another. Um, but for hundreds of years it's been recognized as a place where you can go as a politician, a business leader, an academic, sit in a room with a closed door, have a very frank and open and honest discussion with people, sometimes competitors, sometimes political opponents, etc. And what happens is then at the end of the conversation, when everyone leaves the room, they can talk about what was said, but not who said it or who they work for. And I'm I'm I think I'm grossly simplifying it here, but hopefully it is a simple enough rule that everyone gets the idea. It's it's that um, you know, ideas can leave the room, but you know, you know the the the context or the script can't. And that empowers people like nothing else, because then when you trust that what you are saying is going to be is not going to be misused or manipulated against you, then you're much more likely to be open and frank with your discussion. And you you get to a point where, um, for example, in my communities, some of the people in the room are competitors. Obviously, we have to be very aware of antitrust provisions. You know, the law doesn't stop at the door, so you know you're not going to do any price fixing or anything like that or market manipulation. But you get, like I've seen literally workplace leaders or sustainability leaders from direct competitors be very transparent and open about their challenges because they genuinely want to hear what everyone else thinks. And you know, most people are in a job for maybe four years in a role, you know, and that's kind of a rolling average at this kind of level of executive. So they don't they don't know who they're going to be working with two, three, four years down the track. And so there is an inherent awareness of protecting your own personal brand, you know, because by the time you get to 1400 kind of level leader, a lot of people know a lot of other people, you know. And so they're willing to share, be very open because there is that element of trust um in the room.
SPEAKER_01So let's then dig a bit deeper. Under Chatham House rule, what do C-suite leaders actually say about AI in the workplace that they wouldn't put on LinkedIn?
SPEAKER_02They are challenged as to how to use it. I've literally heard a leader of a Fortune 50 level company say they turned on AI for 100,000 staff, got free licenses, and then they gave them all back six months later because they couldn't work out. They couldn't see the value. They could see a lot of distraction, they could see a lot of risk, um, but they were really struggling to see the value. Um when when I challenged the room about, okay, so where is the real value? The number one thing has been meeting transcription. You know, but but people are genuinely challenged about where is the value in in AI and w where does it cross over from um you know being a distraction to the business or to individuals to actually being something that's genuinely useful and that you can prove the the outcome, you know, you've paid 20, 30 pounds or dollars or whatever a license for you know for a team, where is it actually showing um results?
SPEAKER_01Yeah, I'm hearing a lot of this as well, that they're not seeing the benefit, and there have been a lot of studies, there's an MIT study that is somewhat uh discredited by some in the industry that 95% of AI projects fail. But I I think there's a root cause here that if you give everyone access to a tool, and a couple of things, you don't explain how to use the tool properly, so there's an education thing, but also you're using the same way you've done things. I use the example of expenses, and I stand on stage, I say hands up, who likes doing expenses? No hands go up. But we've done this process the same way for so many years. So you then you throw AI in there. Shouldn't we then go, hang on, we have an opportunity if AI is as good as it is and we know what it can and can't do and what the limits are. Should we not be reimagining how work gets done? And that is hard because I think a lot of leaders, and probably the leaders in your rooms, are coming off years of a digital transformation. Then COVID hit, they had to reimagine how they took things online. They're exhausted from all these transformations. Now, this AI thing comes along that frankly is moving faster than any organization can keep up with. And to your point, they give the licenses back because they're not ready yet. The the word picture I look at is imagine a conveyor belt, all these shiny AI tools, agents, ChatGPT, Claude coming along the conveyor belt, but they all fall off the end because there's no one there to collect them because they're not ready. So, two questions. Is it that people don't understand what it can and can't do? And that's an education and awareness and a digital AI literacy. But secondly, are they afraid to change the way they get work done?
SPEAKER_02A lot of people are are afraid of that. Um the the message to a workforce, you know, of all of a sudden we're going to um you're going to need 50% fewer colleagues because now you've got a lot of spare time. You know, if you haven't given people the empowerment to do their work better and to move on to more meaningful work, then then you are at risk of that. And and if you haven't built that use case around, okay, if we're going to generate 50% more available hours in the day because we're going to use more efficient tools, well, what are you going to do with that with that time? You've got to have a really clear strategy for how you do this. Um, I listened to a podcast the other day, I think it was a guy from Cantor talking to someone from Microsoft, uh, and they were talking about how they've embedded agents and they've enabled the staff across the whole organization. And one really good example was that if you are an employee and you need to apply for a finance and new mortgage, you have to ask your employer quite often for a letter to prove that you are still employed there. And maybe that also includes what your salary is, so you can validate that, or at least that you are employed. And that process apparently used to take days. And and it took a significant amount of time from the HR department to do this. So they built an agent that will do that automatically, you know, in literally in seconds. It would say, you know, you don't ask this agent, etc. And so that's a great example of something that AI is absolutely very good at, and that every company should be looking at how you automate those boring, repetitive tasks that don't actually add any value, really. Um certainly they're not empowering people to be more creative or um you know future-looking or whatever. You know, so I think that's an example of where businesses can actually get real value and become more efficient. But if they don't have a clear story on what they're gonna do with the time that has then been saved, then eventually someone's gonna come along and say, well, do we still need these people? And you're gonna get rid of them. And then people get concerned about, oh, if I use AI, am I gonna put myself out of a job or am I gonna lose half of my team? I think. And that's a whole different conversation.
SPEAKER_01Well, uh, studies have shown that uh back in the industrial age that workers actually sabotaged machinery. They broke into factories and destroyed the machinery because they were concerned for the same thing. And they thought if they broke the machine, then uh that they would uh they would be able to work the way they did before. So that is human nature that we don't like change. But it's interesting what you say then, and again, I'm just reflecting on this and the discussions I have with different clients. So the metric there was extra time that we have. But if you look at that one process where getting a statement from your employer that you're employed to allow you to go for a mortgage, which is actually something outside the organization, this is a cost of the business because I've got an employee, this isn't part of their job to go for a mortgage, but they need something from their internal. That's one process. Imagine every single HR process, if we could actually reimagine that and actually save time. Now, does that mean we get more time in the day? No, we're actually exposing the wasted time that's already there. So if I have to wait three days to get a mortgage application, it doesn't actually improve my productivity at my desk. That's helping me fulfill a better life, but it's actually taking some strain off that. So I'm wondering if the metric is the wrong metric that's not saved hours. Think of it this way: think of the job that you do, the job that the listeners do. You could actually slice that into a group of tasks. And imagine slicing that and looking at the tasks like that particular one there on expense claims you don't like doing. Could you automate that? Perfect. Could you actually then focus on the things that you love doing because you need human creativity or you need human judgment? The way I phrased this, I actually did this to a group of hairdressers in January from Weller. It was 500 hairdressers, and I coined this phrase which I keep using. Focus on what you love and automate the rest. So why I employed you to be this business development manager, whatever, wasn't to spend time on a mortgage application, it was to delight customers. And if I can slice the tasks away of your job and even things ancillary to your job to get your life done, your life admin, it isn't about having more time to spend on those things. I'm just wondering whether it's the wrong metric that when we take the shackles of all these uh this work admin and life admin off there, we actually see that the things that we love doing that we're we were employed to do, we're we're even better at that. What would be your response there?
SPEAKER_02Well, I think you can look at it at from something as basic as a process to provide you a letter of employment. Um, but if you scale that up to say you're a workplace global workplace leader with a team of a hundred people and you've got people in facilities and uh you know CRE and technology and whatever, um, you know, security, for example, um, and and you say them, okay, I'm gonna empower you to use AI to find all of the boring stuff that we can automate in this business, you know, whether it's reporting or report generation or data gathering or whatever, whatever it is. And that will then, you know, I want each of you to take an AI tool and go away and find five things that you can automate, that you can, you know, cut down from you know, 10 hours a week to two hours a week or one hour a week or whatever. You know, brainstorm this workshop it, bounce ideas off each other, red team each other, all the all that kind of thing. Try and put each other's plans out of business, you know, really and and and your own. And then come back to me and say, how much time could we save? And then what are we going to do with that? Where do we want to go? What has been on our wish list for years that we want to do, that we now have the time to do, you know, and and and the mandate to do? Yeah, go and not quite put yourself out of a job, but go and try and make your job as efficient as possible, um, obviously still meeting your key metrics or whatever, and then come back and let's work on the really cool stuff, the passionate stuff, the fun stuff, you know, the the the the the the er error redefining stuff that we previously just did not have the headroom to do because we were buried under, oh God, it's Monday and I've got to get this report in. Well, whatever that report is, can you have automated it and it's in your box already created, and you just need to go through and read it and fact-check it and then send it off. You know, so I think there's huge potential here for people to be much more creative and empowered and happy in their work with the right use of AI to get rid of the boring, mundane stuff.
SPEAKER_01I totally agree with you on this similar advice I give to clients as well. And I was talking to a journalist, or a four-minute journalist, Harriet Meyer, who's actually been on the podcast, and I spoke about this, but you said here's the issue: in a newsroom where, yes, there are huge efficiencies to be made, the news doesn't stop for a process of re-engineering. So we actually don't have time to sit down and think about this. And I'm wondering whether discussions you're having at that senior level under Chatham House, that people are going, we want to do this, we aspire to this, we haven't got the headroom. We would love to stop for a couple of days and reprogram and re-prioritize things, but but the business is moving at light speed. Do we need a separate team to be doing this that don't have the day-to-day responsibilities? Uh, and but who's going to say, oh, let's go and find 50 more people or a bunch of consultants? So is it a sort of a it's an endless loop where we want to do this, we haven't got time to stop and think about how we could do it. Give you an example. Um, I uh I was asked to talk to a family-run business there in a logistics area, and uh the it was a husband and wife team, and the wife said, uh, I really want you to talk to my husband. And he was in the background and he was racing into a meeting, and he said, Oh, he hasn't got time for to talk to you about this. He's got to do a production meeting. There are 30 people they meet every week and they go through inventory statistics by hand. And I said, This is a wonderful irony. I could actually save those people, it was 15 people, 30 person hours a week by not having to do this because they could actually do it with AI. But he didn't have time to stop to hear that he could save the time because he was too busy doing it the old way, which kind of is frustrating. But how do we break this cycle?
SPEAKER_02This predates AI, right? This is you know the the endless CEO conundrum of spending time on the business versus in the business. You know, and if you aren't blocking in the time in your diary to step back, whether it's a daily thing. A weekly thing, a monthly thing, a yearly thing, you know, quarterly thing with your team. If you aren't making that time to step back and think, are we still doing the right things for the right reasons for the right people in the right way? All those questions, again, red team your own business, because if you don't, someone else you're a hell is hell's going to. So look at everything, no sacred cows, say everything is is potentially up for grabs here. Are we still completely relevant to our original mission, our cause, whatever you want to call it? You know, and and part of that should be what are we doing that we can either stop doing and outsource to AI or or refine, etc. But if you, as a business leader, aren't making the time for that, or as an individual, if you aren't making time for that in your own work life, then eventually you are going to become less and less relevant. And someone is going to come along, either as a business or an individual who is more relevant than you are, and you're going to find that you go broke or you get replaced, whether that's by technology or you know, someone else who is better at the job than uh than than than you've become. So yeah, there's there's no silver bullet there. You just got you have to make that time to to pause and review and adjust.
SPEAKER_01I've got an example that might help our listeners here because making that time is actually uh you've got to do it piecemeal. A digitally curious executive who has had the aha moment when they realise they can save hours in doing things, whether reviewing spreadsheets or meeting minutes or whatever, it means that let's say they claw back three hours in a week, what they have to do is ring fence that and say, I'm going to use those three hours for those things we just talked about. I'll give you a real-world example, and I love the fact that when when you tell stories, uh, from being in the room with someone who's actually lived this, it's it's more believable. So this is a trust story as well. I spoke to a group of lawyers, a uh a law firm, all their partners, and in the room also were their support staff. And the lady who actually is in employed to do the board minutes, she said, I love what you said there. Can I use transcription? Because we had Microsoft Teams to describe transcribe the board meetings and maybe write up the meetings faster. And I said, Yes, of course you can. And you're probably gonna save two to three hours every time you have a board meeting. She said, She almost hugged me. I didn't know it could do that. So I've just saved her three hours, let's say a month that she's got in there. Imagine if her partners could also use that to claw back some time. So maybe the answer is you've got to become digitally curious, you've got to become comfortable with the technology. Pick a task that you really don't like doing, but you could automate, and then get the hours back by doing that and slowly chip away at the problem. Maybe that's it. I'm just answering my own question in real time here.
SPEAKER_02Andrew, I agree completely. And something that I have um, you know, started to do really, really in the last year. I mean, there is no way I could have founded and grown the businesses that I have in the last year if I wasn't using Copilot. Just flat out no way. Um first of all, my my brain just jumps all over the place too much. Um, you know, and uh the idea of reviewing endless screeds of notes that I'd written in my own spidery handwriting just would not work. Um as I told you before we started recording, I've now got knocking on for half a million words of transcripts of calls, one-on-one calls, and meetings that I've had over the last year. All of this done with the awareness and participation of the individuals. Um, I treat those incredibly confidentially. Um and they become a coworker. You know, it's like I've got the best um, you know, COO in the in the world or something like that, who knows everything about every conversation I have had and can filter through and can spot the patterns and can spot the gaps or the similarities or whatever, can spot the new ideas to go for, um, ideas that perhaps aren't closed out, loops that need to need completion, that sort of thing. It's hugely powerful and empowering. And I think that, again, this comes back to trust. People have to trust that you, when they are being recorded or transcribed in a Teams or a whatever platform you're using meeting, they have to know who is going to use that data and for what purposes. And so long as you are giving them back genuine value, then most people I think will be happy with it. Now, there might be some instances where perhaps it's a it's an HR conversation or a financial conversation or something where actually you don't find. But you know, I think in the last year I can't remember anyone ever saying, no, I don't want to be transcribed, because every time I've had meetings and what they will say is, can you please send me the summary of the meeting? You know, send me the notes because I don't want to have to write. And when you free your mind from having to think and write down frantically and try and remember all of the nuance of it, when you know that it's going to be there, stored, that you can go back to, it it means that you can you're you're freer in the conversation to explore different directions and to and to come back to points and things like that. So I still take notes. I'm taking notes right now on a pad in front of me as we talk. Um, but I'm not having to remember every single thing we talked about.
SPEAKER_01You've had that aha moment around meeting transcription, and you mentioned them a couple of times as a as a really value add, and I've seen it happen in real time. A couple of examples that are going to resonate as well. So my book contains about 60 interviews from my podcasts. To be honest, I was really struggling to get this book finished. And what I then did was thought, okay, I can use transcription. Let's feed the 60, nearly 60 hours of podcast audio into Otto. Not only did I get a perfect transcript from everyone, I could then interrogate using the GPT inside that, the same way Microsoft Copilot does. So, for example, I could say, okay, where's that brilliant sound bite about quantum computing from the guy from IBM? Bang. So I then had not having to write 75,000 words from scratch, which was my fear, and that's why I've held off writing a book for a long time. I had half a million words ready to interrogate. And that was that allowed me to write the book in six weeks because I was delayed, delayed, and the publishers went, we need the book manuscript. And I, when people say, How long did it take to write? I go, Well, the short version is six weeks because I used AI to surface the content. The other thing that's really interesting. So just before we came on air, I had five minutes left after a previous call. It was a what we call a pre-booking call with a future client. Uh, I had the transcript of the notes. I needed to send my agent an update on where we're at. So I literally said, I use Claude, and I I love Claude because it connects to everything else I use. So Claude connects to Otter and to my CRM and everything else. I literally said to Claude, grab a summary of the meeting in a format I can send to Dominica as a summary of where we're at. I changed two words, I sent it, then I clicked to go online with you. I had five minutes left, something that would have taken probably 15, 20 minutes, and it just saves so much time. I tell that story and you tell your stories. I want people on this podcast to hear that and go, I didn't know how I could do that. I need to do that. If that's going to save me hours a week, I'm up for that, but I've got to try it. And do you think people are hesitant because they haven't done it themselves or they don't know that it can be done?
SPEAKER_02If you haven't tried it, it's hard to communicate just how empowering it is. So I'm on a number of boards. Every quarter we get a 50-page summary of the latest board meetings. You know, here's what was discussed here are all of the executives who came in, et cetera, et cetera. Um, it's 50 pages, my God. And a lot of it's pretty dense. The idea that I could look at that one, remember what was in the previous one from three months ago and draw conclusions is just laughable. It would take me all day with a lot of sticky notes and whiteboards and drawing bits of string between different things. I can put those two documents together in the same kind of containerized AI agent space, whatever you want to call it, and say, identify the key trends, the key changes, the pressures, you know, that kind of thing. And it comes back in like 30 seconds and it gives me this amazingly polished, you know, document that I can then go inside or go and challenge. But it it gives me reference points. You know, I can say, okay, so point five, which was about growth of the organization, where were the proof points? You know, where was the statement in the in the Q2, you know, board summary versus the Q1? And it will go and work that out for me. And so again, because it doesn't have access to the general internet, it's highly unlikely to hallucinate because it's not trying to look at anything else. I'm asking it to compare and contrast two documents, two PDFs. You know, and that is incredibly powerful. And if you aren't using this in your role as a business leader, then you are missing out on a huge amount of insight, you know, that will make you a lot more relevant to any conversation that you go into.
SPEAKER_01Yeah, I advise boards as well, and I get pushback. One on the recording thing, that sometimes they have discussions that are difficult, they don't want recorded, and that will get over that. But let me ask you a couple of qualifying questions. Don't tell me the organization. How long have you been a board member and how long's the organization been going for?
SPEAKER_02Um, it's been going for a long time, but my engagement's um a couple of years.
SPEAKER_01Okay, so this is brilliant. Same thing. I'm on a board in London, it's a charity, I've been on it for nearly three years, but the organization's been going for 70, so similar parallel. I talk to people at board level, I say, what about if you created a virtual AI board member? Now, this is not someone who's going to replace a board seat. You just gave an example there of going through documents. The wonderful thing about AI is it moves and works at speed and scale. You could read through every document. You could read through every document ever before your board time and you would have probably the same design, but you can't. So, what I proposed is if you've got a difficult decision in that containerized zone when it's inside the organization, feed in all the past board minutes and the financials and all that sort of stuff. As a two or three-year-old board member, you don't know what happened 10 years ago that might be the way to solve this problem. So, what's your view on the decision partner? Not a board replacement, but we basically say, let's call it um Mike. So, Mike, Mike AI, you've been through all of the documents since zero, the day zero. We have this difficult decision to make. What do you think? Now, would you trust the person with three years board experience or the the thing that had every single development in the in the company? And is this valid and viable? And what's the blocker from installing this into boards and senior leadership as a decision partner?
SPEAKER_02Wow, this there's a lot to unpack here, Andrew. So I would, as always, trust but validate. You know, and so if you're having a discussion around, say, finance and you've got a person with three years experience saying we should do this, and someone with you know with the with the awareness of say 50 years of documents, that might there might be something involved with a say a property lease that happened 10 years ago or an investment strategy from 20 years ago, or something like that, where it can bring his historic context that perhaps a more a newer employee board member, whoever, isn't, isn't aware of. Um, you know, so so that's absolutely you know something that and and I think you you've obviously got, I mean, probably the most obvious product I can think of here is Microsoft Facilitator, where you can have facilitator in a Teams meeting and you can give it an agenda and you can you can be asking it and talk literally talking to it like a real person in the meeting and say, have we covered this? What should we be focused on, et cetera, et cetera. Uh and it can scurry away and find documents or bring up the the sales presentation from Q1 or whatever. So if there's a discussion backwards and forwards about are these numbers valid, who's hit their targets, you can actually say, well, hey, facilitator, can you go and do this, please? Yeah, and this kind of idea of um uh I was in a workshop back in February, um, and we were kicking around the idea of a pan business agent, you know, and the very quickly we got to the point of no, you wouldn't have one agent across a business. You would have multiple agents who would know how to talk to each other. You'd have a finance one, an HR one, an operations one, a project one, et cetera. And they would be empowered to talk to each other or not, as as the case may be. So, for example, the HR one would be allowed to tell maybe high-level things, but wouldn't be able to divulge someone's personal information to the finance one or whatever. You know, so you'd have to ring fence it. But back to the idea of a virtual board member, yes, absolutely. The key thing here is that it sits there as someone, and probably the best analogy I can think of here is in I think it's Star Trek. It's the character Data, you know, who is uh an android. Okay. He obviously has access to everything it knows about the ship, all of the history of the organization, all that kind of stuff, military tactics, everything. So it can make recommendations, but it's ultimately Captain, not Captain Kirk, whoever, Captain Picard, whoever it was, I'm not a Star Trek fan, um, who makes the final call. And they are making that call with a more rounded view of the information, you know, or or the context or whatever. They've got input from the other humans on the bridge, you know, the communications officer, the warfare officer, whoever, and then there's data, there's the AI agent who is saying, well, this has been tried 10 times before against XYZ, you know, alien force, and it's never worked, or this has always worked, or whatever. So it could potentially prevent a tactical misstep that could have easily been avoided. But ultimately, it has to be the leadership in the room, the humans that trust and understand what they know and equally as much what they don't know, and they then make the final call.
SPEAKER_01So this is where I say I frame this as a decision partner, it's not making a decision, but it's helping you rapidly uncover things and doing that sort of what-if analysis. We we all know playing with Logitech 123 and Excel the what if. But what I like using AI for is to plug in all the information I know and then ask it to justify the things it's telling me, and it you it directs my thinking. So, not just in a LinkedIn post, that's probably the most basic example of it. But if I've got a difficult business decision to make, I put all the information in there and I ask it to what am I missing? I had a situation last week where I fed in all the information, I said, what's likely to happen next? What haven't I thought about? And it brought a couple of scenarios that I hadn't actually considered. Now I have to go through and work out whether those scenarios are valid and and they should be explored, but it helps me uncover it because it's got access to more information. So I I'm not sure if you answer this, but what's stopping us from doing this? And it in your Chatham House discussions, if I was in there talking about this now, would I have people laughing at me or embracing me?
SPEAKER_02I think I think they're absolutely embracing you. And they are all um hearing that other people have the same challenges, I think is quite often the number one value of being in a room like this. Um I've heard multiple people describe it as the best group therapy they never paid for. Um, you know, so ultimately a leader in that position is paid to make decisions when the data is incomplete and the consequences are real. Okay. That's still a human job. You know, no one, I don't think we've worked out yet how a company or a customer would sue an AI agent if it got something wrong. Okay. Microsoft's not taking on that responsibility, Anthropic's not taking on that responsibility. Every single AI thing you ever get has a little tagline at the bottom saying this could be completely wrong. What I do and what I'm sure you do is when you have information, you can ring fence it to make sure that it's not going outside and hallucinating. You want to keep it confidential, be 100% sure it's not leaking onto the internet, all that kind of thing. Ultimately, what it really boils down to is knowledge can be automated. How you process it can be automated. Accountability can't. And that's what you've got to get to grips with as a leader.
SPEAKER_01I get asked all the time, how do I trust what I see on AI? What do you do to validate internally and externally that what you're seeing is trustworthy or or not?
SPEAKER_02If I'm reading a thought piece on a LinkedIn or you know, a newspaper article or whatever, you know, I can if it's got facts in it, I can it's very easy to fact-check generally, ultimately you go back to, I don't know, the World Economic Forum or or some government report or or whatever. Where it's random opinion, that obviously is then very subjective. And you can then use AI. So you've got fact-checking, and then you've got, um, I'm not sure what the fact checking isn't the right word here, but kind of is this the right vibe? Is this person's assumptions leaning in the right direction? And so, you know, kind of fact-checking is kind of binary, yes, no, it's right or it's wrong. The assumptions that people came to as a result of facts, either correct or incorrect, is then subjective. You know, and so again, it comes back to this ultimately human thing of you've got to trust that you've that your own personal processes, that the the the rule that you are putting over all of this information in front of you is going to lead you to a path that you are prepared to stand up and defend. And you that's really where your own personal brand trust and credibility is on the line. You know, so you have to do the homework. If you simply put something into AI and get it back and go, fine, here you go, boss or board, this is what I think. If you haven't done the diligence behind that, then you quite rightly deserve to be called out. So you've got to do the homework.
SPEAKER_01Yeah, one thing I say, AI won't take a job, someone who knows how to use will take a job. I go one step further and I say that AI will demand better humans. Not nicer humans, but if I have really interesting information that's surface in front of me, and I don't use my superpowers to understand, interrogate, use my training and and and empathy experience to understand what that means, that's I think gonna show up that that here you go, here's here's three different scenarios. You've got to be a better human to understand that rather than just take it at face value. And I think that's the real power of AI, that it's gonna demand humans really stretch their thinking and lean on what they know, what they've with what they've seen, and and actually ask the community.
SPEAKER_02Imagine you know, a thousand years ago, and obviously there were these amazing um libraries across the world of Islam, and uh you know they they that they they had these huge repositories, thousands, thousand tens of thousands of documents and hundreds and hundreds of years of learning. You imagine walking into a library like that as the leader of a tribe and you've come out of the desert and you've never seen any of this stuff before. You don't immediately have the kudos or the or the the the the knowledge of that entire room. You know, but there are librarians who can help you. So you could say, go and find out this about the crops, crop cycles, or find out this about politics, or find out this about how to build a building that doesn't fall over. You know, but it is still up to you, as say that tribal leader, to take that information that you've learned and the and the outcomes, the decisions that you made back to your tribe to say, right, this is how we're going to build our fields and our buildings and this is how we're going to defend ourselves, that sort of thing. I I I hope the analogy holds up. Um, you know, so AI in some ways is that library and the agents in the library, the librarians who are individual subject matters, how you use that that that that um opportunity is is up to you.
SPEAKER_01Sometimes I suggest to people they put AI in the org chart only because then it forces people to go, we have this resource available to us, it's there 24-7, we should use it. I've really enjoyed this discussion, it's made me think, and I'm glad we reconnected over a LinkedIn post. But what I like to do with all my guests is run them through the quickfire round, we find out a bit more about you. So if you're ready and up for it, let's go. Window or aisle?
SPEAKER_02Oh, window, every time.
SPEAKER_01Your biggest hope for this year and next?
SPEAKER_02That I will have built the foundations that leave my family and my community better than I found it.
SPEAKER_01I wish AI could do all of my Accounts. The first thing I asked ChatGPT was How do I use you? The app you use most on your phone.
SPEAKER_02Overcast podcasting.
SPEAKER_01The best advice you've ever received.
SPEAKER_02Don't give your money to a jerk.
SPEAKER_01What are you reading at the moment?
SPEAKER_02It's a book about um the importance of geography and and what borders mean, that sort of thing.
SPEAKER_01How do you stay digitally curious?
SPEAKER_02By challenging my own viewpoints, by asking myself, where am I in a in a rut where of where I'm self-reinforcing my own views and what's going to get me out of it? Who should I talk to who is going to disagree with me?
SPEAKER_01Who should I invite next onto the podcast?
SPEAKER_02I think you should invite Richard Mulholland.
SPEAKER_01How do you want to be remembered?
SPEAKER_02As someone who helped others to trust each other.
SPEAKER_01So, what three actionable things should our audience do today to understand better how we can use AI responsibly?
SPEAKER_02Challenge yourself. Are you up for discovery? Are you up for being scared about what the future brings? Um talk to other people about what works for them, um, how they see things changing, evolving, what they're doing, uh, what they would do differently. And then finally, have a plan to act, schedule time, do a course, take that first step.
SPEAKER_01Mike, fascinating discussion. How can we find out more about you and your work?
SPEAKER_02Uh LinkedIn. So if you search uh MikesteadNZ is I think my username on LinkedIn. Um and uh yeah, pretty much all my thinking is there these days.
SPEAKER_01Mike, thanks for a challenging discussion. Made me think, and I don't think this is the last time we speak.
SPEAKER_02Looking forward to the next time, Andrew. Thank you.
SPEAKER_01Thanks, Mike.
SPEAKER_00Thank you for listening to Digitally Curio. You can find out more about Andrew and how he helps corporates navigate the future of AI and beyond with keynote speeches and Texweek Workshops at actionablefeatures.com. Until next time, we invite you to stay DigitallyCurious.com.