Loading summary
Transcript56 lines
- [00:00]
A
AI is moving fast, no question about that. But I think the biggest challenge for a lot of organizations is they just weren't built for this kind of speed. My guest today says the winners won't just adopt AI. They'll redesign how decisions, learning and accountability work. Hey, it's John Jantz here. I've got a quick question for you. Are you a consultant, agency owner, or fractional CMO who feels like you're reinventing the wheel with every new client, or worse, giving away strategy for free? Well, you're not alone. And that's why we created the fractional CMO plus certification. It's a three day live experience where you'll license the Duct Tape Marketing proven strategy first approach. You'll learn how to turn strategy and strategy engagements into a product. Our next certification is right around the corner. Head on over to DTM World Certify. That's DTM World Certificate Certify and book a call with a live advisor. Or heck, you can just chat with our AI advisor too to see if this is a fit for you. Hello and welcome to another episode of the Duct Tape Marketing podcast. This is John Jantz. My guest today is Stephen Wunker. He's the managing director of New Markets Advisors where he's advised hundreds of organizations on growth, innovation and new market strategy. He's also the author of a book we're going to talk about today, AI and the Octopus Organization. Building the Super Intelligent Firm. So, Steve, welcome to the show.
- [01:41]
B
Thanks for having me on.
- [01:43]
A
All right, so it seems like I start here. Longtime listeners know I always start with titles, words and titles. So I think the, I think, think when you suggested this show, I think the thing that got my attention probably is for a lot of people it's just the word octopus. It's a metaphor that's used a lot in business. So what's the simplest way for you to explain an octopus organization?
- [02:08]
B
So an octopus has a biology that is weird for us humans. And actually it's very unusual in the animal kingdom in general. It has nine brains, one in its central head and one for each of its arms, which means each arm can sense and think and act independently. And yet all those brains are interconnected with a nerve ring. So they don't need to route everything through the central brain, but they can sort of coordinate semi autonomously with complete contextual awareness. And it's a great metaphor for how an AI infused organization is to operate with distributed intelligence, a lot of local sensing and acting, and yet with complete contextual awareness. So we developed that Model in detail as a guideline for how organizations should adapt to really make the most out of AI.
- [03:03]
A
So I'm seeing a lot of organizations, especially we work with a lot of small to mid sized businesses that really are not compartmentalized necessarily like large organizations might be. And I get the feeling that they, they look at AI as like software. So this is just a better version of Word, you know, or something like that. So what do you, I'm curious if you're encountering that as well or if you see that as a clear sign that they're not thinking about this distributed intelligence.
- [03:32]
B
John, I'm seeing it all over the place. Small organizations, certainly large enterprises. I was with a hundred product managers earlier this week in Denver, talking to a group there and they were reporting it. Even in their software companies. They were getting a lot of requests for these very incremental improvements. And I mean, all that is nice, but the real unlock comes from rethinking the system of work and the workflows. Not through swapping some regular interface with a open text box for a natural language chatbot. No, it comes through taking your 21 step process and making it three steps, albeit maybe three different steps.
- [04:14]
A
Yeah, yeah. Or automating the handoff between your steps. Right, yeah. So the idea, I, at least as I sense it from you, the idea of the eight arms is that we're pushing decisions out in a lot of cases, decentralizing. Where's the risk in that? Where does that break down? Like what role does data play, you know, in making that happen?
- [04:36]
B
Well, look, you need to have decent data in order to make decent decisions.
- [04:41]
A
Yeah.
- [04:41]
B
So that's certainly one. People need to know how far they can go. You also probably don't want to take a human entirely out of the process. Let me give you an example of how to do this in a marketing situation. We worked recently with a marketing department of a major health system and they wanted to infuse AI in what they did. So one issue they would have is that they would work with the different service lines in the, in a hospital trying to create a marketing message for them. But often the service lines didn't even know themselves what their marketing messages should be. So for the lower priority stuff, if like the urology department wants to do some brochure about their services, they were able to put into very simple AI system queries and a sort of interactive process to help urology come to that messaging itself to draft messaging for it. I help them create the output and then the human can get involved and Say okay, well how about we do this? Or we're not quite aligned with the brand values so we need to go in this direction. But it took out so much of the work that marketing was spending on this relatively low priority task and actually was better for the people in urology as well that they didn't have to go back and forth and back and forth every period of weeks. They could do it all with the self service engine.
- [06:04]
A
So one of the things that we've also encountered is a lot of, you know, as organizations, especially if there's not like a leadership mandate for AI. A lot of organizations have various people that have said, well I can figure out ChatGPT, you know, and so they're using it to kind of do their work better. But there's no real central guardrails, there's no brain, there's no retention even of, you know, what they've built. I mean, how do you build a system that is, that starts of course from leadership. And I know part of the answer is going to be buy in from leadership, but how do you build something that, that has the guardrails that you know that we're going to speak like the brand, we're not going to use the wrong language. So that if you have all these people in various departments or various locations even using the tools, you know, how do you make, how do you set the guardrails up?
- [06:50]
B
Yeah, the era of having 900 pilots I hope is drawing to a close. They are distracting, they're dangerous. Yeah, they maybe they got people comfortable with AI, but it is unsustainable. So look, you need to have a common data foundation with some data quality. So definitely invest in having decent quality data, ideally some sort of data lake or other system to provide that free flow of information throughout the organization. There needs to be some governance guidelines and governance process on how AI should and should not be used. But then there also needs to be a mechanism to assess these pilots that are still going to occur. But look, we have a three step system, abc AI ify the present. You want to go AI ify what you're doing now, great, let's make sure it's within guidelines. But go do it. B, become great at experimentation. So what's actually your hypothesis? What are you measuring pre and post? What did you learn? How do you kill pilots that aren't working? And then C, which too few organizations are doing is create the future. So for a handful of things, you can't do it for everything all at once. But in a handful of golden workflows as we call them. What can you do to really rethink what you're doing and make those sort of a lighthouse for the rest of the organization. Do that in a half dozen situations and then move on to the next half dozen and the next half dozen.
- [08:17]
A
So I want to come back to golden workflows because I want you to explain that. But I will say another thing that we encounter a lot is that people are just applying AI to maybe a process that's non existent or broken as opposed to, you know, I always tell people start with workflows first and then use AI, you know, to automate them. Don't tell AI. Create me a workflow for this. How, how do you, how do you work with organizations where they really need to. I think I said at the beginning, they really need to rethink how they even do workflows, how they even serve their client.
- [08:51]
B
So you need to understand what your starting point is and where are all the disconnects. But then don't just think about what might we automate. You want to think about what humans shouldn't be doing because AI could do it better, right? What won't they do? Because it's just too time consuming, but it would be valuable. And what can't they do? Because it's just overwhelming in the amount of detail.
- [09:14]
A
Wait, you forgot whether they hate to do. That's one of my favorites.
- [09:18]
B
Yes, maybe that should be and shouldn't do as well. But look, I. So anyway we have the can't shunk won't. And when you do that then you can really rethink. Okay, with AI, what makes, what does this make possible? So we did this recently, another marketing department in campaign planning and they would spend about half a year planning campaigns for people in the business because there'd be infinite going back and forth and then they have to go to the agency and it all these disconnects and nobody quite knows what they want. And you go through iteration after iteration. We were able to use those principles and come up with the system that was about half the time, could actually be a lot less. But there's still going to be some human indecision that you have to account for. Yeah, but by just reducing the number of cycling because you can real time prototype stuff, get real decisions made right away, make sure everybody has context of what happened in the prior meeting. So you eliminate a lot of those disconnects. You could do it in half the time. Now you can't re engineer every process all at once, but you can do it in some of those golden workflows like that one.
- [10:28]
A
So, so again, you're going back to the golden workflow. You're saying that's basically, you know, a lot of times when people bought into this whole idea of systems, they would try to systemize everything and just get overwhelmed because there were 740 systems and 723 of them didn't matter. So as a golden workflow, in your vernacular, what's a real impact? One that we absolutely have to get?
- [10:53]
B
Right, right. So I looked, I mean campaign planning was a great example because there was real money being spent and a lot of time. There would be others, same marketing department with press releases or even internal announcements about what was going on. People would go back and forth and back and forth. And I mean it's, it was a huge sink of time and you had about 40 people just doing internal communications and we're 28.
- [11:19]
A
28 of them were attorneys though. Right.
- [11:23]
B
Well, I mean, so you do need levels of scrutiny. Right. You cannot just think we're going to replace humans with AI. It doesn't work that way. But you do want those people focused on the highest use of their skills and not helping people make decisions that AI would probably help them make. Even better than a human consultation.
- [11:42]
A
Yeah, yeah.
- [11:43]
B
So I mean this is, we're not seeing a tremendous amount of displacement of humans with AI. There's some in particular in things like call senders or whatever.
- [11:51]
A
Right, right.
- [11:52]
B
But more it's just ensuring that people can focus on the best use of their skills. That's where the real productivity gains are.
- [11:59]
A
I, I suspect I wouldn't study be a paralegal right now though either.
- [12:03]
B
No, that. Well, yeah, that one is already played out. Right. So. And that's example. Right. Auditors, There are some others that are really threatened, but it's a minority of what. Yeah, most white collar jobs are what.
- [12:16]
A
We'Re finding, especially in knowledge work, that what it's doing is not displacing people, but it is asking them to do their job differently. You know, to maybe manage AI as opposed to do the stuff that AI is now capable of doing. If you're a consultant, agency owner or fractional CMO who feels kind of exhausted from starting from scratch with every new client. Well, this is for you. The fractional CMO plus certification is a three day live experience built to help you stop being the bottleneck and start leading with clarity. You're not only going to learn, you're going to license all of our systems, documented processes and practical tools all Grounded in the duct tape marketing strategy first approach, we're going to help you turn strategy engagements into a product you could sell over and and over again. You'll also get one on one support and join a community of marketers who actually get what you're building. Our next certification is coming up soon. You can learn more at DTM World Certified. DTM World slash certified. You. You started talking about the. The eight brains, I think I recall. Don't. Doesn't an octopus have three hearts also?
- [13:31]
B
It does indeed.
- [13:32]
A
So how does that matter? How does that metaphor come into play for you?
- [13:37]
B
So an octopus has different hearts for different purposes. And similarly a company needs to have different operating modes as it enters this very dynamic period of AI led disruption. Also many other disruptions too. So we talk about the analytical heart, which most companies are already pretty good at. The agile heart, which the bigger the company is, the worse it typically is at agility normally. And then the aligned heart of how do you make sure that people understand the common purpose and where people are going, particularly when there's going to be just a lot of turmoil in the workplace. And there is with the entry of AI, being attuned to those emotional cadences in an organization is going to be really important.
- [14:24]
A
I mean, would that. Could you break that down almost department wise? I mean is. Or not even department but function wise like that. What I just heard you describing sounded like culture.
- [14:34]
B
It is, yes. And culture is incredibly important. We also have another chapter on emotion and the cultural side of change. Look, we think about culture brick wall. The culture is the mortar in a brick wall. It is almost invisible, but it gives the whole thing shape and coherence. But you start a brick wall with the bricks which are the hard levers of management control. The way you select people and incent them and measure them, the way you allocate resources around an organization. If you don't have that right, you could put up all the posters you want on the walls. That's not going to change the thing. So get the bricks right and then definitely think about the culture.
- [15:16]
A
But.
- [15:17]
B
But don't just think about it as some isolated thing. It's not.
- [15:20]
A
I like that because unfortunately you do see some examples of performative culture that you know, then doesn't really deliver. You know, when you shine a light on what the company's actually doing. I think also in the subtitle you have the term super intelligent firm. What's a super intelligent firm mean in. In terms of human terms? Maybe.
- [15:41]
B
So there is a fallacy out there that AI is going to approach general intelligence. So being just like human or super intelligent, Just like a human, but even more so.
- [15:51]
A
Right.
- [15:52]
B
And that is based on providing a very human model to a machine. We evolved to who we are because we needed to escape the sabertooth tiger. So we had to be good at a lot of stuff. Yeah, but machines don't. They need to be good at just a handful of things. So the superintelligent firm isn't necessarily super intelligent AI. Where firms get their intelligence, where organizations get their intelligence, is through collaboration of people. It's not just through like brute force processing capability. Right. If we want to say who has the most neurons, well, an anthill wins that contest. Right. So humans can do something which octopuses can't, which is collaborate. And that's how we build civilizations and cities.
- [16:35]
A
Right.
- [16:35]
B
AI has the potential to supercharge that collaboration by making sure the right information goes to the right people at the right time. And it's that use of AI that actually enhances our humanness that is the most high potential use of AI, because that makes us then super intelligence as firms, as organizations.
- [16:55]
A
So one of the things I've, I tell people because I think a lot of, especially in marketing, you know, a lot of people's first reaction was, oh, look how much faster I can do stuff. And while there is a bit of truth to that, I think it's not very interesting to just produce more content. What I think is really interesting is I can iterate 200 versions of something in about the same time it took me to do one. And I think that's where the learning comes, because we all know that we're just guessing sometimes in marketing. And by being able to test faster and experiment faster, that's where we're going to not just produce more, we're going to produce better and we're going to produce more personalized.
- [17:33]
B
I wrote a Forbes profile about a year ago on a company called Movable Inc. That provides email software to companies like L.L. bean and Victoria's Secret.
- [17:42]
A
L.L.
- [17:43]
B
Bean can run a million different variants of an email, literally a million different variants. And of course it's AB testing and it's determining what's best, but it's also using that to be just super personalized in ways that we could never do as human beings. Yeah, but you know, this is all possible now. So yes, we have to embrace that.
- [18:05]
A
So what would you, how does an octopus move look like for like a 20 person firm as opposed to say a 200 person?
- [18:16]
B
You know, the principles are actually Often the same. The maladies may be different. Right. The 200 person, 2,000 person, there's a lot of discoordination and siloing. Hopefully in 20 person it's not. Yeah, but you still need to think about for your prioritized workflows, the things you really want to focus on. How do those principles of moving action closer to the, the suction cups on the tennel, right to the coal face, how does that look for you? What does AI enable in terms of what humans shouldn't be doing or can't be doing or just won't be doing because it's a poor use of their time? How would you really fundamentally rethink things? We talk a little bit about when electricity came. Initially, operators of factories swapped out their steam powered machines for electric machines and it's sort of equivalent to what we're doing now. And that was nice, but it actually took 35 years for the big unlock to come, which was the assembly line. The assembly line could not happen without electrification. But it was that unleashed the productivity because it was a fundamental rethinking of how work was done. So we don't have 35 years this time, but we need that equivalent of thinking about what's the equivalent of assembly line for our companies.
- [19:32]
A
So if a CEO, company founder is listening and they want to like start adopting this, obviously they need to get a copy of the book first. Do you also work with, do you come in and work with organizations to, to install this? Yeah, so that's my day job.
- [19:45]
B
Yes.
- [19:45]
A
Okay, so, so if somebody was listening then and they said, okay, I, what's my 30 day plan? I mean, do you typically. I'm sure it starts with an assessment, but do you also typically, are there things that you often test almost right off the bat? Are there things that you tell them to stop doing right off the bat? And, and again, every instance is different, but are there some common things that you find.
- [20:11]
B
So it has to relate to your strategic priorities?
- [20:13]
A
Yeah.
- [20:14]
B
Right. So AI can be better, faster and cheaper all at once, but to some degree there is a trade off. So what are you actually hoping to achieve and how does that fit in with your objectives? What is impeding that today? And then let's overlay that on where AI can be particularly useful. So we're not just coming in with a hammer looking for nails, but we're trying to understand what are the different priorities in the organization. And then based upon a handful of things that we can really focus on, let's figure out how to redo that Sometimes it starts all the way with the value proposition, but it can also just be very internal process oriented as well.
- [20:55]
A
Awesome. Well, again, I appreciate you taking a few moments to stop by the Duct Tape Marketing podcast. Is there someplace you would invite people to connect with you, find out more about your work and obviously find out more about AI and the Octopus organization?
- [21:08]
B
Sure, you can find the book on Amazon. The website for the book is aiandtheoctopus.com and that is actually a subdomain of our company, New Markets Advisors. Also connect with me on LinkedIn. Feel free to write me. I do answer my emails, so send me a DM and I'll get back to you.
- [21:25]
A
Awesome. Well, again, I appreciate you stopping by and hopefully we'll run into you one of these days out there on the road.
- [21:29]
B
Steve, My pleasure. It.