Geoff Woods returns to Beyond the Prompt to discuss the updated edition of The AI-Driven Leader and what has changed since his first appearance. The conversation explores why leaders need to use AI themselves rather than delegate it, how to think beyond AI "use cases," and why the biggest opportunity isn't adopting new tools but improving the way leaders think, make decisions, and shape strategy.
Geoff Woods returns to Beyond the Prompt to discuss the updated edition of The AI-Driven Leader and what has changed over the past 18 months. Rather than focusing on the latest AI models, Geoff argues that leaders need to use AI themselves before asking others to, using it to think more clearly, shape strategy, and make better decisions.
The conversation explores why many organizations confuse access with adoption, why strategy should come before use cases, and how AI can change the way leaders approach everything from business strategy to organizational design. Along the way, Henrik and Jeremy challenge Geoff's ideas on authorship, judgment, and whether understanding AI changes what leaders believe is possible.
Key Takeaways:
The AI-Driven Leader: aileadership.com
Geoff's LinkedIn: linkedin/geoff-woods
00:00 Are You Qualified to Lead on AI?
00:35 Meet Geoff Woods
00:54 The AI Slop Dilemma
05:48 Putting Your Stamp of Approval
09:19 What Changed in 18 Months
12:13 Access Isn't Adoption
15:50 Why Leaders Can't Delegate AI
20:09 Strategy Before Use Cases
22:02 BarkBox's AI Strategy
26:23 Reinventing Strategy with AI
33:11 Compressing Months into Hours
37:22 Human Skills as Superpowers
42:46 The Debrief
📜 Read the transcript for this episode: are-you-qualified-to-challenge-your-team-on-ai-with-geoff-woods-author-of-the-ai-driven-leader/transcript
[00:00:00] Geoff Woods: I wanna create such a profound, transformative experience that they deeply internalize that this is not IT's job, this is my job. I do not have to become an AI expert, but I do have to become an AI-driven leader, meaning I personally use the technology every single day to elevate the impact that I make in the organization, and as a result, I'm qualified to now cast a vision for the future.
I am qualified to define strategy for how I'm gonna win, and I am qualified to lead the team from practice, not theory.
I'm Geoff Woods. I'm the author of the number one bestseller, The AI-Driven Leader. I was on the show 18 months ago when my book first came out, and I'm back again as my updated book is now being released, and we are talking about what has changed over the last 18 months and why I still believe AI is here to enhance you, not replace you.
[00:00:54] Henrik Werdelin: Can I start with a question? There's this architect called Bjarke Ingels, who... Well, I, I used to serve on his board. He's a good friend of mine, and we were at this cool conference last week, uh, called Brilliant Minds. And I've been working on this manifesto because I think the world needs basically more positive versions of what could a future look like when- Mm
AI kinda, like, roll through. And so I show him this manifesto, and he takes off his glasses and he looks at it and he goes like, "Did you r- write this yourself?" And I replied, "Of course not, 'cause I don't write anything fully myself anymore." And then he looks at me and he goes, "Well, if you can't be bothered taking the time to write it, I can't be bothered to take the time to read it."
Right? Which then kind of obviously started... I mean, it wasn't, like, meant in a harsh way, but it's just, you know, like, I... You know, we started having conversations. A- a- and obviously I've spent a lot of time on this document, right? You know, I've spent endless amount of time. I've had, like, different modules that I brainstormed with Claude on, and I created an Obsidian folder with all the background research, and then together when Fable was out, had Fable help me write some of it, right?
So this is, this is good writing. And so, uh, it started this, I think, very interesting conversation for me is when length of copy is no longer the, uh, the proof points of effort that you put into it. What is? How do I know if what you send me is AI slop that you just kind of did one prompt and you just send me whatever random long text that you got back, or that you spent, like, six to eight hours actually really working through this document? And I don't have a good answer for it, so I'm posing it to you guys.
[00:02:28] Geoff Woods: Do you need one?
[00:02:30] Henrik Werdelin: Ah. You know what? I, I, I think I do because I also get AI slop, and I find it to be annoying when- Yeah ... somebody sends me something and I get halfway through and it's clearly it's just kind of like their first shot. Yeah.
A- and I also want to be respectful. If I send something to you, Jeff, then I, I would want you to feel that I value your time, and- Correct ... if you for a second felt that I didn't, I would feel embarrassed by that.
[00:02:57] Geoff Woods: I have thought about this, and again, I... The role of the human is to be the thought leader. AI is your thought partner. As thought leader, you're in the driver's seat of the car. As thought partner, they're in the passenger seat. I write every word, but I use AI to write every word. And so for me, it's I can tell when somebody has just one-shotted it and sends it across. You're right. That does not leave a good impression. The question that I find myself asking is, did they as thought leader review this and approve this?
Did they put their stamp of approval on this to say, "This is my voice. This is the message I'm trying to deliver. Do I believe this is the appropriate length?" Did they apply judgment and discernment to the message? When that happens, I love it.
[00:03:45] Henrik Werdelin: Mm.
[00:03:45] Geoff Woods: When that's void, that's when I judge them.
[00:03:47] Jeremy Utley: Part of the challenge is this bias. I mean, yes, I agree, of course, everything you said, yes. Um, Bjarke's problem is his bias And I don't think Jeff has w- no one, none of us have a chance to give a well-reasoned, um, you know... That, that's, that implies that he's willing to engage. There's something interesting, Henrik, there, which is he sees the document.
It sounds like he didn't read it, right? He doesn't read it, and instead he asks you whether you wrote it, and then he decides whether he's gonna read it, which is, that's a pre-consumption hesitation, bias, et cetera. I wonder-
[00:04:28] Henrik Werdelin: I, I think it was 'cause it was, it was a long document. What, this wasn't, like, a few paragraphs. This is something like clearly-
[00:04:33] Jeremy Utley: Oh, it was the length that he was reacting to.
[00:04:34] Henrik Werdelin: I think it was the length he was surprised. And I think he also, uh, I think he's expressed abso- I mean, like if, A, I, I didn't answer his question. If I said, "I spent a lot of time on this," he would have read it immediately, right?
Like, and so it, this was not, like, a contentious conversation. This was just kinda, like, one that kind of became, I think, just interesting because of, like, the way- Yeah ... it played out. And I also have another friend, when I call him, uh, he, I said, "Am I calling at a bad time?" And he says, "Yes." And I go, "Should I call back later?"
And he replies, "You don't have to." So, like, we, there is the, the, the nice bender there. Um, but I think what, I think what he was just reflecting on is that he now often get these long documents that somebody have one-shotted. And because that you kind of wasted your time when you're halfway through it- Yeah
and figure out it's... I think he, we didn't talk about this, but I, I imagine what he's kinda looking for is some kind of qualifier that this is something that is now kind of not worth his time, that sounds so, uh, pretentious, but, like, something that he has to really spend real focus on. And I think- Yeah
actually that's a fair thing. Like, if you send me something and it was, like, a one-shot thing, I'll also be like, "Hey, man, I just spent 20 minutes reading this," and, like, clearly it was just the first thing that Claude spit out.
[00:05:47] Geoff Woods: I agree. The message I deliver with my team, 'cause I do still get some documents sent to me that I can tell are AI generated or feel off, and I will just call that person and I'll say, "Walk me through what your process was from the output to you refining it to, to putting your stamp of approval on it."
And I'm looking for the holes there. And I'll even go to sections and say, "Do you agree with this? Do you stand by this wording? Do you stand by this thinking?" And then they say, "No." And I go, "Okay, this is a coaching moment. Never again send me anything unless you are saying, 'This is my stamp of approval.' So this is me making the investment in you as your leader.
This is now a standard, 'cause moving forward you will be wasting my time. I won't tolerate that."
[00:06:36] Henrik Werdelin: We'll all end up being po- politicians. It'll be like, "I send you something," and it'll have to end with, "My name is Henrik Nordlund, and I approve this piece of copy." And I,
[00:06:42] Geoff Woods: I approve this message
[00:06:44] Jeremy Utley: But it's-- I think it's actually, it's an interesting point. There's the managing of the AI, which is the, call it, uh, employee's responsibility. But then as leaders, there's the management of people who've now been given a new managerial job, you know? And you actually, you still have to take responsibility for your team's work. And in a way, what, what I hear you saying, Jeff, is, "I'm gonna teach you how to be a good manager, team member," for lack of a better word, right?
And I think the, the bummer if you're a CEO or you're a leader of an organization is what you want is for these people to know how to be good managers. And what you're realizing probably, what a lot of people are realizing is these folks have never had new manager training like we used to do when all of a sudden Jeff was given a team, we invested in his ability to get exceptional work out of them. Now we hand Jeff Copilot and are disappointed he's not a good manager, right?
[00:07:34] Geoff Woods: Correct. It's a training gap. Yeah.
[00:07:36] Henrik Werdelin: I do think, though, I just wanna make sure that we mention this also. Jeremy, I think, was asked some-- a question a little bit like this at one point, and I, I think you said something to the extent of, "I almost find it to be kind of, like, criminal liability if people have not used AI."
Like, in the same way that if you go to, in, for an operation, the doctor say, "You know, I didn't use a microscope 'cause I wanted to eye it with my own natural eye," you know- Right ... you would be very uncomfortable with that, and so.
[00:08:01] Jeremy Utley: The interesting thing there for you and BR, Kenrick, is to say, uh, to stand by your statement, of course.
Like the most interesting part of that conversation to me is your willingness to say, "Of course," because most people in that moment go, "No, no, no, it's, it's, it's all me. That was-- It's all my handcrafted, artisanal cognitive labor there. Be-- Or trust me," right? And I like that you took a principled stand. Then the conversation, I think, should be, "Why did I say of course?"
Mm. Let's talk about that. Why do I now think it's a, it's an abdication of responsibility to do it without augmentation? And that's like, that to me is a really interesting place to take that conversation.
[00:08:41] Henrik Werdelin: Yeah, and, uh, I mean, we spent a long time at this, but, like, I then wrote a LinkedIn post about it today, and the commentary is kind of interesting 'cause I also think that there are some people that use writing to formulate their own thought.
Now, I do that in voice more than I do that in text. It's better for me to talk it out than it is to write it down. And so having AI then to be able to take the conversations that I have, and then after having many of them, distill that and put it into something that is thoughtful and short, is something where I have, you know, get a lot of help from AI.
But for other people, I think the way they get to that point is by sitting and actually putting the fingers on the keyboard, so maybe it's also just a little bit individual.
[00:09:19] Jeremy Utley: Okay, Geoff we have to know, switching from, from Henrik's, um, fascinating Bjork moment, we have to know what's changed. Okay, we had you on the show, what is it, 18 months ago?
I just thought that'd be a fun place to start with you, talking about, uh, your expertise, what you've learned in the rooms that you've been in. What's changed in the last... Has it been 18 months? Is that approximately it, two years? I don't even know.
[00:09:41] Geoff Woods: Uh, well, the book came out, AI-Driven Leader came out September of '24, so we're-
[00:09:48] Jeremy Utley: Nearly two
[00:09:49] Geoff Woods: years, okay maybe a little under 18.
[00:09:50] Jeremy Utley: Yeah.
[00:09:51] Geoff Woods: Yeah. So here's what's interesting. If I look at the things that have changed, I, I kinda put them into two categories. There's what's changed that's been timely, like, well, this model came out, and now you can upload a PDF. Now it can generate a PowerPoint. And there's what's timeless.
For me, I think most of the changes that were timely were distractions. Now, I had this theory when I was writing The AI-Driven Leader that most of the news, the tools would be distractions 'cause for me, I don't believe the goal of any company is to, is to adopt AI. I think they wanna harness this technology to create a better business and better lives.
So that's my bar when, when a new model comes out. Is this going to help me build a better business or a better lives or a new functionality? Is this going to help me build a better business or a better lives? Like, if I build this GPT, this project, this agent, is it going to help me build a better business or a better life?
That's a very high bar. Most things don't qualify for that. So there are so many things that have shifted in the technology. What has not changed is- Using the-- I didn't realize how profoundly impactful it would be if you just started using AI to elevate how you think as a human.
[00:11:06] Jeremy Utley: Mm-hmm. That hasn't changed, you're saying?
[00:11:08] Geoff Woods: That has not changed, and I think where the market is overall, I mean, if you go back 18 months, every leader knew AI w- AI was the future, but they had not started, and they felt like they were falling behind. If I look at where the market is today, they know it's the future, they think they've started, but they're confusing access with real adoption-
[00:11:29] Jeremy Utley: Yeah
and they still think they're falling behind. Yeah, that's, that's, that's a really, that's a really cool characterization, Jeff. I totally resonate with that. 18 months ago, they knew it was some- it was on the list of things I need to do. May- maybe a different way of saying it that's kind of punchy is now it's on the list of things I've already done.
Right? And
[00:11:45] Geoff Woods: the- But
[00:11:45] Jeremy Utley: it's a
[00:11:45] Geoff Woods: false sense of confidence.
[00:11:47] Jeremy Utley: Exactly, exactly, 'cause anybody who's really deep in AI goes, "I'm not done with the internet." Right? Does anybody think of internet and go, "Yeah. Oh, yeah, yeah, I'm done with that"? You go, "No, it's now an integral part of daily life." And I think the leaders who, you know, maybe 18 months ago thought, "Uh, I need to do that," maybe now they're, you're saying now they're in danger of saying, "Oh, yeah, I've done that.
I rolled out Copilot licenses," right? Or something like that, right?
[00:12:11] Geoff Woods: Correct. Correct.
[00:12:13] Henrik Werdelin: Geoff, how do you then tackle with the feeling that you're so ahead of this, and I'm sure your own organization has the autonomous agents, and they do all these different things. And then meanwhile, we go into organizations, I'm sure you do that, and then you get the sense that they're just getting, like, Copilot rolled out.
And it almost seems that we are now getting fascinated by this next level game that you can play with AI, where people are still playing the first version. Do you think that people that are advanced are kind of almost losing kinda, like, the majority of people that would like to go on this journey 'cause we're already thinking about autonomous agents, or, or how do you compute that?
[00:12:56] Geoff Woods: Not necessarily. I mean, I remember I, I was studying adoption curves when I was writing The AI-Driven Leader, and it's like, you know, you got three and a half percent are your innovators. Then you got your early adopters, early majority, late majority, laggards. I knew that the adoption was gonna follow that curve, and that's just proving to be true.
Um, I'll still-- Well, the-- We do a lot of, uh, sessions for executive teams of companies, and more often than I would like to say, when I say, "All right, open up Chat or Copilot or Claude," there's a C-level executive in the room who hasn't logged in before.
[00:13:31] Jeremy Utley: Who doesn't have it on their device, maybe, yeah.
[00:13:32] Geoff Woods: Yeah, and it's, and it's their first time.
And for me, I know my job, and this is just a timeless leadership principle, your job as a leader is to meet people where they're at. Not where you think they ought to be. So if where they're at is, "I need to help them log into ChatGPT," great, let's do it. I'm gonna help them do it remarkably well. But now let- let's, let's get them on the fast track to using it.
So that is still where the market is. I also question, Henrik, how much of it is really gonna matter over time.
[00:13:59] Henrik Werdelin: Hmm. Say more?
[00:14:00] Geoff Woods: Well, imagine if a company today said, "We have a competitive advantage 'cause we use Microsoft Word or because we use the internet." I think most of AI is gonna become a source of competitive parity.
Right now
[00:14:13] Jeremy Utley: you just can't- They'll be competitively disadvantaged if they don't use Word. Correct. Yeah. A company would never brag, "We're, we're an anti-Word shop," right? Yeah,
[00:14:19] Geoff Woods: exactly. Exactly. Exactly. So there will come a day where just AI's table stakes. It's like electricity. It's like internet. And at that point, it's competitive parity.
Real competitive advantage doesn't come from AI. It comes from using AI to enhance your internal capabilities that are valuable, rare, or costly to imitate. Those are the raw ingredients of a competitive advantage. So even when I look at companies that are slower to adopt right now, I do question how much of that matters.
Now, specific industries, it does, because if you're in an industry like the accounting industry, I was with a, a firm, one of the... A big one, and they're all playing this game of chicken. Like, they all know whoever gets rid of the billable hour first, everybody else is gonna follow. And there- The whole, the whole house of cards falls
there's a real opportunity.
[00:15:10] Jeremy Utley: Yeah.
[00:15:10] Geoff Woods: There's a real opportunity to gain market share, and I'm on stage in front of their 450 partners and, and I, I said, "I'm gonna be a velvet hammer. I feel like you guys are all on the Titanic. You can see the iceberg and you're celebrating the meal that you just had." That doesn't seem too velvet.
[00:15:28] Jeremy Utley: There's no vel- Okay. There's no velvet
[00:15:31] Geoff Woods: there. They loved it, by the way, but it's like they know they have to do something, yet they're still debating. It's like, dude, you're saying you wanna win, but you're not thinking and acting like you wanna win. Like, you have a legit window and time to move.
[00:15:45] Jeremy Utley: Right.
[00:15:45] Geoff Woods: Move.
So that's why, Henrik, like, de- I think most companies, depending on adoption, how it
[00:15:50] Jeremy Utley: matters.
[00:15:50] Geoff Woods: What
[00:15:50] Jeremy Utley: do you think's going on there? Wh- why, I mean... And again, this speaks to kind of timeless, whether it's leadership, innovation, et cetera. Why the reluctance to move? Do you, do you have any reflections on that? And again, this is kind of apart from AI, but why is the default, "Let, let's stand on the sidelines and wait and see"?
Is it a question of where folks are in their kind of, you know, careers?
[00:16:12] Geoff Woods: I see three buckets. One is they don't know how to, so it's a genuine they know they need to move, but nobody has shown them how to and made it simple enough that they can start. Two is you've got industry or personality types like- accounting, legal, like they're risk-averse by nature.
The construction, they just tend, they tend to be laggards when it comes to technology adoption throughout time. And then there's strategy. I think Apple had a strategy. They've never won by being the first adopter. They've won by letting everybody else spend billions in R&D, and then they just step in and do it better.
And instead of spending billions, now they're gonna get paid billions to use somebody else's AI model.
[00:17:02] Jeremy Utley: I'm hesitant to give too many organizations the benefit of the doubt that, oh, y'all are the Apple here. You
[00:17:09] Geoff Woods: know? Yeah, and by the way, if you think you're the Apple, let's assume you're not.
[00:17:12] Jeremy Utley: Yeah.
[00:17:13] Geoff Woods: I think
[00:17:13] Jeremy Utley: that's a safer,
[00:17:14] Geoff Woods: that's a- That's just a, that's just a convenient excuse.
[00:17:16] Jeremy Utley: Okay, so you get in these executive sessions, like you, you get an executive team saying, Jeff, I mean, you just mentioned that, where I'm with executives and I, I, I love that phrase, meet people where they're at. So if it's login, great. What are you hoping? I mean, if, if you don't have to deal with login issues, and you get a team of executives in a room for whatever, 90 minutes-
[00:17:35] Geoff Woods: Yeah
[00:17:35] Jeremy Utley: what are you hoping to accomplish? In your ideal world, what would you do to help unleash possibility there?
[00:17:43] Geoff Woods: Galvanize them to become AI-driven leaders. For me, I wanna create such a l- a, a profound, transformative experience that they deeply internalize and commit to this idea that this is not IT's job, this is my job.
I do not have to become an AI expert, but I do have to become an AI-driven leader, meaning I personally use the technology- Mm ... every single day to elevate the impact that I make in the organization, and as a result, I can now... I'm qualified to now cast a vision for the future. I am qualified to define strategy for how I'm gonna win, and I am qualified to lead the team from practice, not theory.
Because I, I'm seeing a huge gap, Jeremy, and I know you see it too, which is every executive team now has said, "We're gonna become an AI-driven organization," but they're talking the talk but not walking the walk. They then delegate
[00:18:40] Jeremy Utley: it and say, "That's everybody else." That means tell other people to do something.
Yeah, and you're saying, what it sounds like you're saying is, get folks to stop telling others what to do, and get them to start elevating their own game themselves. Do it yourself.
[00:18:52] Geoff Woods: Do it yourself. Yeah. That's, that's hypocrisy, not leadership.
[00:18:55] Henrik Werdelin: I realize that I'm just the slow person in the room now, but like, I finally just got what you've been talking about here, and I think that is very profound.
I mean, like that, I-- let me just mansplain
[00:19:07] Jeremy Utley: it back. What do you think I'm talking about, Henrik?
[00:19:08] Geoff Woods: Yeah, no, but I think when you talk about the AI-driven leader, AI-driven company, I have been like, oh, you know, like, but then you also say that it's not really about AI, but-
[00:19:16] Henrik Werdelin: Everybody on this podcast has gotten, so let me just mansplain it to everybody who listens.
But what I- No, this
[00:19:21] Jeremy Utley: is good. It's good. Roll it back ... you're just, you're just saying, and I think
[00:19:22] Geoff Woods: this
[00:19:23] Henrik Werdelin: is very profound and very true, that this has nothing to do about making sure that your organizations use AI per se. It's about you using the tool that's available you, to figure out what is the next step function.
What does your company want to be when it gets old or older? And a lot of people just haven't asked that question to themself for a long time because we've been going through the trauma of just having to deal with COVID and then this new thing kind of like landed. And the only way to really be a, having an Iron Man suit on and move at the speed that you need as a professional business person today is to understand these tools profoundly so you could use them for yourself to- Correct
build an organization of the future that- Correct ... is probably gonna use AI, but that is not the, the point of the whole thing. I get it. So I- I'll go home now. Thank you ...
[00:20:09] Geoff Woods: and, and I realized this myself- Well said ... when I was an operator. Like before AI, when I was, I was a chief growth officer for a public company out of India, and when I saw AI, I started using it myself, and I realized very quickly I was using it for all the things that didn't matter I was using it to write better emails.
I didn't get to that level by the quality of the emails I wrote. I got to that level because of my ability to think strategically, to solve problems, to communicate, to collaborate, to create. And I found myself wondering, how can I stop asking AI questions and start getting it to ask me questions instead to elevate my thinking and my leadership?
Uh, the chairman saw this. He puts me in charge of AI for a company. We have 100,000 employees. This is a big fricking company. He says, "Go figure out AI." Partner with Google, learn a lot from them on the tech, how to identify use cases, train tuned models. We're driving it through the org, and I'm realizing, and Jeremy, this is when we met, and I'm realizing it is not technology first, strategy second.
It is strategy first, technology second. Everybody has... A- and they're still asking, "What's the right use case?" I think it's the wrong question because you're, it's like you're holding AI as a hammer, and you're looking for a nail, versus what are the biggest problems that we have that if we could solve them, it would unlock a new level of growth for us?
Now, how do we solve that and recognize that AI is a tool to help you solve those problems? So my realization is focusing on what's timely, like how do we use AI in our sales process, or how do we use AI in customer service? I don't think that's the right question, versus what is timeless? What is timeless when it comes to building and growing a great business, and how do we use AI to supercharge those processes?
That's where all my focus
[00:22:02] Henrik Werdelin: has been. Can I give you an example, which I think- Please ... might k- So I'm still involved with Barkbox. You know, like, it's a $400 million turnover, uh, treats and toy business for dogs. And the way that we've seen our strategy and how the world have evolved in that space is that it started by you bought your dog toys in the supermarket, then you went to Petco, and then you went to Bark because we kind of, like, we niched down, right?
And so one of the things we're working on right now is how do we niche even further down to a breed level? Because if you have a Goldendoodle, then you probably just wanna be served by somebody who just do Goldendoodles, and it also work with, like, the conversion funnels of, like, influencers being more powerful, so there's a more elegant thread between somebody who talks about your product now.
Now, there's 360 breeds in the US, and if you want to do mixed breeds, you have to do 360 times 360. That's 116,000 different permutations of this thing, right? There is no way you could solve that with humans. I cannot make a business where I have 160,000 different storefronts, except with AI. With AI, I can make this thing, right?
And so I think for me, as you're saying, that is an articulation of, like, the strategy is niching- the, uh, the way that you can get that and the reason why you need to understand AI is because you didn't know that that was even possible to do if you didn't understand
[00:23:19] Jeremy Utley: AI. You couldn't implement that. Yeah, and that to me is where I would put, I don't know if the right word is pushback, Jeff, but I'm scrolling up on my notes because you said something earlier and I, I wanted to kind of think about it.
Uh, the idea is And it gets at what Henrik's saying. The, the strategy we can now imagine of niching down, it was previously outside of the realm of possibility. So there's something about a familiarity with the capabilities of AI that's actually required to ratchet up what you imagine. And so, oh, oh, I remember what you were saying.
Timeless-- Sorry, I, I forgot for a second. Brain fart, but small language model here. What you were saying about timeless changes versus timely changes, and you said, Jeff, that, "I don't think model developments really matter that much," or s-some version of that. And where I would push back respectfully, and, and in the spirit of kind of, you know, collaboration, is they do matter when they fundamentally change what the kind of strategic possibilities are.
And so- Yes ... I don't think that someone can be oblivious to what the best-in-class models can do. But again, not because they need to be techy, but because it redefines what is the kind of strategic possibility set that we need to be- I agree ... discussing as a leadership team. Okay.
[00:24:35] Geoff Woods: And here's my own experience.
People will always ask me, "How do you stay ahead on AI?" And my answer is, "I don't try to," 'cause I think most of the latest things are distractions. I don't study the latest model release. I'll hear that it'll come out. I don't spend half a second investigating what its capabilities are. I don't. When it matters, it surfaces.
Like when Claude Cowork came out, I started to hear about it, and anytime there's a, a significant model that changes real capabilities, I hear about it, then I dive in. So I don't waste my time with the latest tool, the latest this, the latest that. The true 20% that drives 80% of the results, the cream rises to the top, then you focus on it.
How do you think,
[00:25:32] Henrik Werdelin: and maybe it's a different way of asking the same question, but I do think it is complicated for most people that are doing strategies today to make a strategy that is AI native in the way that we just talked about it with the niching down. So how do you think a business leader that, yeah, like is not as in tune with what's going on, even if it's only the 20%-
[00:25:59] Geoff Woods: Correct
[00:25:59] Henrik Werdelin: even have the ability to formulate the vision for their business without having that insight?
[00:26:07] Geoff Woods: I can tell you how we do it. Um, this is just our approach, and this is also an area of expertise for us. The last decade of my life has been helping companies shape strategy and turn it into execution, so I understand how do you do that.
But I used to be the guy that walked into the offsite with the sticky notes and the Sharpies and, you know, asking questions, and everybody's writing their answers down, and we're getting them on the wall. We're clustering them. But I realize now with AI that model is broken because one person talks at a time, the loudest voice dominates, the smartest ideas don't always speak, whether because of politics, psychological safety, or maybe they weren't even invited into the room.
And the process takes absolutely forever. People usually nod their heads, but then they kind of go and do whatever they were already doing. I found myself wondering, 'cause I keep, I keep a sticky note on my desk that says, "How can AI help me do this?" Just as a, as a trigger in my environment. I don't go looking for use cases.
What I'm doing is the use case. And I was preparing for an off-site, and I was like, "How
[00:27:12] Henrik Werdelin: can AI help me do this?"
[00:27:15] Geoff Woods: We ended up inventing and patenting a model specifically for this, where instead of sticky notes and Sharpies, we architect a prompt. So like I'm doing this for the executive team for DaVita next week.
They're giving us all their strategy data, we're ingesting it into a model that has the right levels of security and architecting a highly sophisticated prompt. And when we get in the room next week, they're just gonna open their computers and open what we call our thought partner, and they're gonna click a button Thought Partner's then going to simultaneously interview every single person in the room about the strategy and ask them the right questions.
What's the AI-driven business that will put you out of business? What would it look like if you built it first? Think big, cast that vision. What do you believe are the internal capabilities of the organization that are valuable, rare, costly to imitate? And so it's asking very strategic questions to people, and they're gonna be walking around, speech-to-text, for probably 20 to 30 minutes answering all these questions.
And what it'll do is it'll lay out almost a McKinsey-style report of every person, of their thinking. We then click a button, it will simultaneously read all of them and merge them, and then AI on the back end, 'cause they d- they don't know all the capabilities of AI, but this is where the prompt architecture on the back end can prompt their initial context to then conduct deeper analysis and research.
And then on the board, in the screen in front of everybody, it'll say, "Here's where you're aligned, here's where you're not. Here's what people are thinking but not saying. And based on that, here's my first question." And it will then, AI will co-facilitate the meeting with me, where it will listen to probably 30 minutes of debate, ingest every word, and then ask the next best question to not only shape strategy, but truly to drive them to alignment in a matter of hours.
Normally, that would take three to six months. So that's how we now do it, is we use it to unlock the collective knowledge of a group of leaders, but also harness AI on the back end to do things that the human can't do with that context we unlock, and then to co-facilitate meetings to drive them to alignment.
That's cool.
[00:29:40] Jeremy Utley: How, um, if you think about the, uh, bias that we discussed in the beginning- Yeah ... Bjarke wondering whether he could trust Henrik's memo, even though Henrik had put a ton of time into it. How do you think about that challenge? Take DaVita as an example. Just because you haven't been there yet, so, so we can, we can speculate, you don't know.
But take DaVita as an example. What makes them invested in the outcome and owning it? Because one thing I observe is folks see AI do a bunch of stuff, and yet you ask them what their favorite ideas are, and they go, "The thing that's on the sticky over here that I didn't say to the AI, that AI didn't read back to me."
You know what I mean? Yeah,
[00:30:15] Geoff Woods: yeah,
[00:30:16] Jeremy Utley: yeah. How do you ensure that sense of ownership, uh, given all of the processing that's happening in the background?
[00:30:24] Henrik Werdelin: Authorship creates ownership.
[00:30:27] Geoff Woods: What's, what's amazing, uh, we've done this enough times now, people see their words on the board. They can see their thoughts. They see their DNA in the output.
And also, um, because it's an anonymous interview and we design it to not accept surface level answers, it pushes for what they really think, things get said to AI that have never been said out loud in the company, because people don't wanna spend the political capital to say it. And when AI puts that stuff on the board, and people look at that and go, "Holy shit, that's never been said, but that is true," like the trust and the quality goes through the roof, and that's just the initial summary.
Then they start to talk about it, and then the summary gets enhanced with... And they can see the language of the room now being reflected in it. They realize they are, they are, they're authoring something with AI, and they just realize, "Oh my gosh, as a human, like our ability to have a meeting and talk one person at a time is completely inferior to if we suddenly engage AI as a thought partner."
[00:31:38] Henrik Werdelin: Have you seen any organization lean in and act in the way that you're prescribing if the CEO and all the senior management is not embedded or is kind of leaned into AI?
[00:31:52] Geoff Woods: Uh, yes. Every company we've driven this with starts that way.
[00:31:57] Henrik Werdelin: So tell me, like it starts with somebody more junior that kind of like introduced them to your organization, or how does-
[00:32:03] Geoff Woods: Nope.
Somebody, some leader reads The AI-Driven Leader, whether they're on the C-suite or it's one of their subordinates that passes it up the chain. Both happen, and they eventually say, "We want our whole executive team to think and act this way." And we get brought in to do a session on becoming an AI-driven leader, where we shift how they think.
Every person learns how to use this as a thought partner. And then sometimes we're also in that session doing what I just, what I just referenced. It's called speed to alignment, where we are shaping strategy or taking the biggest problems of the company and using AI to align them on top solutions. You always have people in the room that are naysayers, skeptics, sometimes hostages when it comes to AI.
It's just 'cause they haven't experienced their light bulb moment. And the moment that AI starts asking them questions and unlocks a profoundly deeper sense of thinking, and then they can see their answers on the board, it's a transformative moment. Truly. And how long do you think it'll take from...
[00:33:14] Henrik Werdelin: Uh, you know, this obviously is a very new way of doing this.
Basically, you know, you're walking wh- eating your dog food, as it were, like, you know- Yes ... doing this in the same way as this- We call it drinking
[00:33:24] Geoff Woods: our own champagne.
[00:33:27] Henrik Werdelin: When do you think consultancy at large is going to do like you do? Or do you think they're caught in the same issue as the accountants and the lawyers we talked about earlier, that-
[00:33:39] Geoff Woods: I don't know.
I, I-- There are certain consultant firms that have seen us do this and then ask, "Can we do it?" Um, which we have conversations with them about what that might look like. But I do believe... I mean, this literally happened, I won't say which one, but one of the big ones. Um, I was at Air Canada, and the, the CEO had made a giant declaration to the public markets on some results they needed to deliver over the next three years, and they needed to guarantee that the C-suite was aligned on this.
This consulting firm had the first hour, and they put up their classic PowerPoint deck, said, "We'll send our army in to interview everybody for the next 30 days, and then we'll come back and we'll give you another PowerPoint deck." They walked out of the room, and I said, "Everything they said they would do in 30 days, we will be done with in one hour.
Open your computers and click the button on the thought partner." An AI interviewed every person simultaneously for about 20 minutes, threw it on the board. That unlocked 45 minutes of debate. It compressed 45 minutes of conversation into five bullet points, asked the follow-up question. By the end of two hours, they were fully aligned.
Then they said, "Great, now how do we drive this to the next 25?" 'Cause these are the people that are gonna have to make it real. So we got back in the room, in an hour and a half, aligned them. Then they said, "How do we drive this to the next 400?" Same process. Um, with Herbalife, we went from 25 people to 400 people to 4,000 people, and now this is gonna become their backbone for how they drive alignment across an organization of hundreds of thousands of people.
Um, I just think this is the future overall.
[00:35:19] Jeremy Utley: I wonder, Jeff, I'm just, uh, brainstorming with you on, on your business. One thing I observed is in the last wave, there's this company called IDEO. I don't know if you're familiar with them. And they- Yeah ... they took a lot of business from the McKinseys and BCGs of the world at the time, and, you know, companies like, I don't, I, I don't know the exact book of business.
I've never been inside. But imagine Target or Walmart or someone, they're going, "Do we wanna hire McKinsey again?" And at some point in the last, call it 20 years, they're like, "We've heard about this kinda cool..." Like, they all wear cool shoes. "Let's get IDEO in here," right? You know, it's like something different.
And to me, there's this moment Where it's not cl- and the inc-incumbents w- they all developed design capabilities. By the way, IDEO now basically, you know, is, is a shadow of a former self. Again, I don't have any inside information, there's just outside observation. My question for you is, have you thought about how do...
Or maybe that's okay. I don't know. I think that is a potential outcome that may be desirable. Have you thought about how to avoid that kind of flavor of the month almost, like BCG, McKinsey, Bain, they're always... And they're gonna develop these capability. If you just fast-forward five years, they're gonna be able to do all of it.
How do you think about long-term differentiation? Or, or do you think about that?
[00:36:35] Geoff Woods: I do. Um, I find myself asking, what are our internal capabilities that are valuable, rare, and costly to imitate? What is it that makes AI leadership AI leadership, and where can we use AI to enhance it, and where might AI undermine it?
And if it does, we're not touching it, and we stay focused there. Um, I've heard so many great leaders say, "Don't spend all your time worrying about the competition, just worry about being the best version of yourself." That's where most of my focus is. Of course, I think about competition and that other stuff, but at the end of the day, all I can do is make sure that AI leadership is the best that it can be, and let's go kick ass and let's take names, and now let's solve the problem that's in front of us.
[00:37:22] Henrik Werdelin: You've done this now a lot with a lot of companies. Where have you found that AI needs re... You know, where have you found that something is deeply human and, and requires, like, just the human in the loop is where the value is, not necessarily, like, the thing. For example, when you're-- I would imagine you've probably done some kind of like, uh- Some of your models have listened to all the interviews, and then it's elevated- Oh, yeah
to th- five of my points, and then you look through the transcript and go like, "You know what? There's like this thing- Yeah,
[00:37:51] Geoff Woods: yeah ...
[00:37:51] Henrik Werdelin: that it didn't pick up."
[00:37:52] Geoff Woods: Um, I think about, you know, 80/20 rule. Most people spend most of their time doing 80% of the stuff that unfortunately only drives 20% of the results. I also think that 80% stuff is where people are acting like machines, focused on outputs instead of outcomes, sending the message, generating a report, doing the task.
Like, that's all going away. Yet there's the 20% they do that drives 80% of the results, and quite often, those are more human skills. And I'm far m- I don't really worry about using AI to take the 80% off my plate, 'cause I could spend a career doing that and not move the value of my company. I just go straight to how do we harness this on the 20% and turn our human strengths into superpowers?
So I'll give you a real-world example. Marianella is my former executive assistant. When I hired her, I told her, I gave her the job description. I said, "These are the KPIs that you have to do really well on to do your job at the highest level." Now, listen to this really carefully. I told her it was her job to hunt me down every week and get me to score her on a scale of one to 10 on every KPI, and if I did not give her a 10, she had to ask what could she do differently next week to get a 10, and she could never ask me those questions without an AI note-taker present I said, "I'm not gonna come to you ever and proactively give you feedback.
It is your job to lead with agency. It is your job to own your development. You come to me and ask me for feedback, and make sure that AI is listening and transcribing every single word." We did this for every week for four months. By end of four months, she is crushing it, so it's time to promote her. I did this differently than I would've in the past.
In the past, I would've said, "Congratulations. Here's your new title, here's your new pay, here's your new job description," which is me saying, "Fit into this box." That's done. I looked at her and I said, "I want you to go to AI and write CRIT," which is my prompt framework, context, role, interview, task. "I want you to have AI help you discover what you think your greatest human strengths are, and cast a vision for how you think you can bring 10 times to 100 times more value to the company."
She goes, "Got it." Next week, she comes back with her vision. She goes, "For me to bring 10 ti-" And you
[00:40:15] Henrik Werdelin: now, and you now report to her. Please let that be what happens.
[00:40:18] Geoff Woods: Come... Exactly. Wait, wait for this. She comes to me. She goes, "Here's my vision. For me to bring 10 times more value based on my strengths, these are the 20% priorities I'm currently doing.
I gotta double down on them. But here's the things I'm not doing that, based on my strengths, I really should be doing. They're gonna bring 10 times more value, but if I focus on those things, here's all the stuff I gotta stop. Now, I've segmented it into three groups. Here's what we gotta stop, 'cause they're distractions.
They don't matter. Here's the things I can use AI to augment or automate, and here's where we still need a person, so here's a job description for my replacement." She walks me through their 20%, their onboarding, their comp, asks for my approval. I'm blown away. And she goes, "I appreciate it. We're not done yet."
"What do you mean?" She goes, "That's my vision for 10 times more value. That's not what you asked. You asked for 10 times to 100 times." I'm freaking out now. And she goes, "My realization is I can't bring 100 times more value unless I make you 100 times more valuable." And she hands me a prompt where AI interviewed me to cast a vision for my role if I were bringing 100 times more value.
It spat out a system of rules that we would have to implement, and then she anticipated, knowing that I like to break the rules, she gave me another prompt where AI interviewed me to understand how I would try to break every rule, and created a defense process so if I ever wanna break a rule, I have to go through AI first.
[00:41:45] Jeremy Utley: Does it involve a shot caller or something where you say, "No." No,
[00:41:47] Geoff Woods: it does not. Her job description went from check my email, manage my calendar, handle my 80% tasks, to run our Latin America advisory process and run our rhythm of the business process inside our organization and make me 100 times more valuable.
[00:42:08] Henrik Werdelin: That's awesome. That's awesome. So
[00:42:10] Geoff Woods: to your question, like where I think about distinct human amplification Challenge people to use it to think. And don't put a ceiling on people's achievements. Don't pretend that you know how they can make the biggest impact. Put them in the driver's seat. Give them agency.
Like align them on company mission and goals, but get out of their freaking way and let them self-discover what their greatest strengths are and how they can make the biggest impact, and support them.
[00:42:39] Jeremy Utley: That's cool. I love that. It's beautiful. It's a, it's a, it's a wonderfully uplifting note to end on.
[00:42:44] Henrik Werdelin: You know what?
Thank you so much for coming on. Jeremy Utley, another interesting conversation with Jeff.
[00:42:50] Jeremy Utley: Henrik Werdelin, sitting in Copenhagen by the window today. Yes.
[00:42:58] Henrik Werdelin: I'm sure that people tuned into this episode to hear you sing.
[00:43:02] Jeremy Utley: Did you love that, folks? You love the banter. Said no one. They tuned, they tuned in for the banter.
Said no one ever. They tuned in for the banter.
[00:43:08] Henrik Werdelin: Um, okay, I have two things. What do you have?
[00:43:10] Jeremy Utley: Yeah. Go. No, I mean, four or five, but go ahead. Of course you do.
[00:43:15] Henrik Werdelin: I... It's not a competition, but it is. Um, I, uh... I mean, like, I was slow to get it, but I get it. I, I gave you this fair- That was great. I love that I think we've been talking about this- I love that moment
we've talked about this a few times, that, you know, the more we learn about AI, realize this is not about A- AI, it's about humans, and it's about, you know, just running your business. And I think it took me a while, but I get it now. Like, what he's articulating is the specific thing that you should use AI to become the best version of yourself so you can make the best company that you can imagine, and that probably requires that you understand these tools 'cause they will give you an iron suit.
And I just think that's a very elegant phrasing of, uh, I think a issue that a lot of people are dealing with, which is that they hear that AI is the future, and they just can't really compute they can make it the future for them or the company- Mm-hmm ... they sit in right now.
[00:44:07] Jeremy Utley: Mm-hmm.
[00:44:08] Henrik Werdelin: Yeah. And so I, I like that.
[00:44:09] Jeremy Utley: Yeah. Just on that, on that point, one thing I found myself saying when I'm talking with folks, you know, you can even go back to the Jagged Frontier study, one of the seminal studies that demonstrated the impact of AI on meaningful work. It was a study of hundreds of people at BCG, and one of the things that the researchers found is, you know, 12% more tasks, 25% more quickly, 40% higher quality results.
I think they later revised that to, like, 35%, but still, you know, 35% higher quality results. My question, very simply, for, you know, if I'm in a room of leaders is, would your team say your work is 35% better? Do you feel your decisions are 35% better? If no, don't tell anyone else to do anything. This is about you, right?
And I think it's... His point i- that AI is not IT's job, it's my job as the senior leader's, is a very-- It's still lost on most organizations, I think.
[00:45:10] Henrik Werdelin: Yes. Because also I think that nobody's really articulated in a very easy to understand way the why. Like why should you... Like the-- I think when people ask for use cases, they're just going like, "I, I don't, I, I have this button on my phone and I see it, but like how?"
And so, uh, yeah, and I think that brings me to the second thing. There are a lot of things that we do in our daily lives that we do because we always done it that way, and he mentioned brainstorming. I thought it's an elegant and interesting way of kinda like rethinking brainstorming. I mentioned the story with, with Piage and, and the LinkedIn post that I wrote.
Um, I now have, I have different agents that do different things. I use, you know, ChatGPT and Claude and all those different ones, but they all kind of like when I ask, I have a skill in all of them where it says forward slash log, and then it just logs it into basically a repository of things that I've been working on.
And every time I do something that seems like interesting-ish, I just go log and it writes the whole thing. Now every day I get two suggestions of LinkedIn posts I can do where it reads the logs, but it also reads like I don't record all my conversations, but I record those that I think might have interesting new nuggets, and it suggests me new LinkedIn posts.
And so I'm now like posting much more often because I don't have to sit there with the empty paper and go like, "I wonder what to write about today." Mm-hmm. Mm-hmm. And I do have a lot of interesting conversation with much smarter people than myself that often kind of include something that would make a very interesting LinkedIn post.
For example, this story about Piage. And so I now get these two things every day and I mean, like I've never posted more stuff and I think it's really quality work and it adds a lot of value to the people who are reading it. But it's just as simple as kind of retooling the way that I get to a LinkedIn post.
And I think, you know, if we could do more of that, look at like each of the different outputs we have or would like to have more of, or each of the different activ-activities we do for thinking, then there's probably a lot of stuff where we could add AI to the mix and then get to a better output.
[00:47:10] Jeremy Utley: I mean, I'll give you another example for me, which is, which just, just riffing 'cause I think that your point about u-usefulness And, and I also have a specific question about how you do the lo- like how do you log the Bjarke conversation, but I'll, I'll give you my example.
I notice wh- when you say output we want more of, you know, you say LinkedIn posts a good example. I, I think that's great. Another one for me is customized proposals for workshops, things like that, right? Like I'll, I'll do keynotes a lot. Very rarely do keynotes convert to customized proposals, or let alone to then customized workshops.
Part of the reason is I'm doing so much other stuff that I don't have the bandwidth to think about what did we talk about... Like, I mean, if you think about like one area that's super interesting in a keynote is the Q&A. Lots of very idiosyncratic stuff comes up about the organization. Well, that's all in my mind.
But then you know what happens right after the Q&A? Like yesterday I was in Chicago doing two keynotes. I gotta race to the airport. I've got a phone call while I'm in the car. I go through security, and then I get on the plane. I get a nap 'cause I'm so tired, you know. And then it's like, and then I open my email, and all of the kind of call it data which would inform a custom proposal is totally gone-
[00:48:26] Henrik Werdelin: Mm
[00:48:27] Jeremy Utley: by that point, right? And so what I've done is I've actually got a, you know, pen that records meetings that I'm in, and I have set up a skill to basically say, "Read that transcript," and then based on this kind of template I have for a custom proposal, iterate the template inspired by the actual conversation in the room.
And I can tell you, I've never sent more custom proposals, you know?
[00:48:51] Henrik Werdelin: That's awesome.
[00:48:52] Jeremy Utley: But it's because that's... I, I had to go through the thought process of, to your point of saying, what is an output that I would want more? Maybe something I'm neglecting that I would actually really like. And that I think gets b- ties back to Jeff's comments around the, the Pareto principle.
The 20% of our activities that drive 80% of our value almost by definition are things we are neglecting.
[00:49:16] Henrik Werdelin: Mm.
[00:49:17] Jeremy Utley: Almost by definition. The 80% that only drive 20% of value are almost by definition stuff that, that end up crowding out our attention And if you'll, if you will kind of do the hard work of protecting the strategically important 20, or actually saying, "How could I direct AI towards that strategically important minority of tasks?"
There are really cool, um, like, like Henrik's LinkedIn post writer, like Jeremy's, uh, you know, custom proposal generator. Like, they're right there. It's mostly a matter of giving yourself the kind of, you know, strategic thinking space to identify them.
[00:49:57] Henrik Werdelin: It makes a lot of sense. And I think, uh, it's probably like a good little, uh, trigger for people who listen to this to go, like they said, for a few minutes after they've turned this podcast off and going, "I wonder what I would like to do more of," and lets me sit down and make the skill for that.
You know? There's... It's often not that difficult.
[00:50:13] Jeremy Utley: Right. Right. Yeah. I did love the, you know, comment that Jeff made, "Now I'm qualified." You know, if I as a leader am elevating my game, I'm qualified to exhort others, but if I'm not elevating my own game, it's hypocrisy. Mm. I thought that was a really nice kind of very, uh, challenging statement for a leader to hear.
Are you qualified to challenge your team?
[00:50:39] Henrik Werdelin: I like that, too. Got everything, anything else? I know you have a hard stop, so I also want to be the time police today.
[00:50:46] Jeremy Utley: The only other thing I would say is I like his, uh, Jeff's comment that the wrong question is what's the right use case, and the right question is what's the biggest problem we need to solve.
And shifting, uh, focus from kind of use case orientation to problem-solving. The only thing is that
[00:51:03] Henrik Werdelin: I haven't... I think sometimes when you ask big, abstract questions, then you also get, like abstract answers, and I think the, the issue with that is I've tried that, too. And I think sometimes then their mind goes into a way that they framed the problem statements prior- Mm
that is in the world that they live in currently. And I think what's so elegant in, for example, the case about the breeds and BarkBox is that that is only a vision. Even the problem statement is not something you're able to phrase if you don't have somewhat insight in what is now possible, if you cannot free yourself from the scarcity mindset of the pre-AI world.
[00:51:48] Jeremy Utley: Well said. Well
[00:51:49] Henrik Werdelin: said. And so I, I, I, uh, I, I have yet to find, like, a good way of introduce people to the way of asking the thing about what is the problem for our business we're trying to solve, but in the context of what will become.
[00:52:03] Jeremy Utley: Well said. I think that's a really g- It's, it's a combination of the, your expertise and the capabilities of the model.
You can't get the most out of your expertise if you aren't familiar with the capabilities of the model, and you can't get the most out of the capabilities of the model if you aren't bringing your expertise.
[00:52:18] Henrik Werdelin: 100%. 100%.
[00:52:21] Jeremy Utley: As always, if you've enjoyed this conversation, hit the seven-star button that you only get, I think Apple, what, they only give you, like, one seven star per year, right?
So use your seven star on us. Use your seven star.
[00:52:32] Henrik Werdelin: The golden, the golden buzzer.
[00:52:34] Jeremy Utley: The golden star. Hit it. Hit it. Awesome. Share it. Like it. Love it. Comment.
[00:52:39] Henrik Werdelin: And we got a new website on beyondtheprompt.ai and-
[00:52:43] Jeremy Utley: Amazing ... you
[00:52:43] Henrik Werdelin: can go in and you can ask it questions about all the episode. We have a rack model with everything that we've ever done.
You can put in your own company name and it will spit back, back at you what is the right episode for you to listen to. And so go check it out and let us know what you think about it. And with that, there's only one thing left to say, and that is bye-bye.
[00:53:02] Jeremy Utley: Bye-bye