Skip to content

Human Factor Podcast Season 3 Episode 042: More Human, Not Just Faster: Why the Real Return on AI Is the Time It Gives Back to People

Season 3, Episode 042

More Human, Not Just Faster: Why the Real Return on AI Is the Time It Gives Back to People

With Elliott Martin, Founder of Future of Change, Author of the AI-Led Change Management Playbook and Creator of the GEROA Framework


Host: Kevin Novak


Guest: Elliott Martin


Duration: 69 minutes


Available: November 19, 2026

🎙️Season 3, Episode 42

Episodes are available in both video and audio formats across all major podcast platforms, including Spotify, YouTube, Pandora, Apple Podcasts, and via RSS, among others.

Transcript Available Below

Episode Overview

More Human, Not Just Faster: Why the Real Return on AI Is the Time It Gives Back to People

Season 3 | Guest Episode with Elliott Martin

Think about the email that landed in your inbox this year. Starting Monday, the new AI assistant is on for everyone. There was a link to a training video, a line about being an AI-enabled organization, and then nothing. Now think about what happened on the floor. One person built a workflow that saves her team a day a week, and she is not telling anyone, because telling someone means losing the tools or getting handed a day more work. One person uses it to search and calls that advanced. And one person has decided it is not for him and is waiting for it to pass. The system got installed. The people got an announcement. Anyone who has lived through a Salesforce rollout, an SAP cutover or a merger has seen this movie. AI is just the loudest version yet.

Elliott Martin has spent more than fifteen years as the person who gets the technology adopted after leadership decided to roll it out, on large programs in Australia at organizations including AGL, BHP, ANZ, the Australian Taxation Office and Coles Group. His route into AI did not start at work. It started with art, a coastal town that does not exist, blended from the two neighboring Amalfi Coast towns where his grandparents grew up, hanging over his fireplace. Then he asked whether there were applications in the work he does, brought the tools onto his own programs, and noticed something almost everyone else missed. Being more efficient does not help you connect with people. It only helps you do things faster. That observation became the AI-Led Change Management Playbook, a firm called Future of Change, and a framework he calls GEROA: Govern, Embed, Reallocate, Optimise and Accountable oversight. Underneath all of it is a line from his first conversation with Kevin: there is a sweet irony in it taking artificial intelligence to make us more human.

The conversation opens on Elliott’s own programs. Kevin brings Wanda Orlikowski and Debra Gash’s 1994 study of a consulting firm rolling out Lotus Notes, where technologists, managers and users each held a different picture of what the thing was for and never discussed the gap. Elliott says the same room exists today. Stakeholders kept telling him they understood the technology was powerful but could not see how it applied to them, and the reason was that nobody had ever explained the full breadth of what it can do. His answer is the Eight Characters of AI. He walks through three of them with a single board paper: the Googler asks what to include, the character that eats administrative burden produces the paper from last time’s example and this quarter’s objectives, and the Devil’s Advocate reads the finished draft and tests the assumptions behind it. Same technology, same task, three completely different interactions, which is why he says organizations should invest the time to help people see every framing rather than one.

Kevin pushes on the person who says it is not for me, because in his experience that sentence is identity threat. Elliott takes apart the premise first. People do not fear change. Ask the lottery winners of the last twelve months whether they resisted the millions; they walked head first toward the Lamborghini dealership, smiling. What people fear is the unknown and loss: does this make me less capable, less necessary? A leader who sees it that way can tell the two apart and answer each one, with education for the first and with the empathy that humans do best for the second. He adds the cautionary tale underneath every rollout. Hand someone a tool that makes them twice as fast and the first thought is great; the second is whether the next step is that they are not required at all. The opportunity is not to become more efficient at what you do. It is to become better at it.

The second segment is shadow AI: advanced users out on their own with one tool while somebody two desks over pays for a different one. Kevin brings Elinor Ostrom’s finding, from fisheries and irrigation systems, that the rules that hold are the ones made close to the people who live under them and nested inside the larger system. Elliott agrees that governance is the weakest element in every rollout; his own change diagnostic at Future of Change says the same. And the fix is basic. Which tools we use and which we do not, what data can go in, what the review path looks like (the same peer review you would want on anything that matters), and who to go to when you are not sure. Write it down and the shadow problem shrinks, along with the two teams paying for the same licenses twice. The part most organizations miss is the level. Data governance policies live at the enterprise, and an executive team will never know the niche tools a change team or a finance team needs. The charter has to live with the team, which is exactly Ostrom’s nesting.

Kevin raises the generation of leaders trained to run an organization on one to three systems, and asks how Elliott gets them to accept many niche tools instead. Elliott’s answer is that leaders respond to value when they can see it. His example is NotebookLM, which he wanted for its ability to turn complex program content into conversational audio for audiences that were not absorbing it. A program director would not have known the tool, so Elliott wrote a mini business case: here is the complexity, here is the feedback from the business, here is a low-cost, low-effort way to solve it. It was very well received, because he had done the research and solved a problem that benefited everyone on the program.

The third segment is the part of the framework Elliott calls his favorite, and it starts with a phone call Kevin gets asked about more than anything else he has said on this show: the marketing leader whose team won back roughly eight hours a week with no policy and no approved tools, and who stayed silent because telling anyone meant losing the tools or being handed eight more hours of work. Kevin frames it with Lisanne Bainbridge’s 1983 Ironies of Automation, where automating the routine hands the human the hardest part of the job and then forgets to design for it, and with Wrzesniewski and Dutton on job crafting. Elliott lays out three forks in the road. Bank the savings: one FTE gone, a few dollars saved, nothing about the work improved, and everyone left behind concludes that being efficient gets you removed, so they stop being efficient or start looking elsewhere. Pile on more work: no one loses a job, nothing gets better. Or reallocate, the R in GEROA. If Kodak had learned to produce film photography faster it would still have gone bankrupt; what it needed was the time to step back and think. His real example is impact assessments. AI can do most of that work from current-state and future-state processes, so his teams used the capacity to go into the business and ask people how they actually felt about moving from A to B, which is the conversation change teams never have time for. The trust equation runs both ways: answer the eight-hour question the old way and staff stop trusting leadership, and leadership gives staff a reason to lie about the efficiencies they find.

Kevin asks whether leaders see it too complexly or simply find efficiency easier than effectiveness. Elliott’s answer is that humans are poor at the long view because the pressures are always immediate: the next board meeting, the share price at the next market interaction, the bonus tied to this year’s financial report. The leaders of the great organizations of the last decade ran on multi-year visions, and AI, for once, makes the long-term case obvious, because anyone can see what two years of not getting on board looks like.

The fourth segment returns to the lottery line. Kevin brings Jennifer Petriglieri’s 2011 research on identity threat, where the same rollout produces resistance in one person and growth in the next and organizations rarely know which they are producing, and asks what a tailored strategy actually contains. Elliott’s answer is that the change management practice exists for exactly this, and AI deserves more of its due diligence, not less. Find the root cause of the resistance, unknown or loss. Run an education campaign into the why, because many people listening work in organizations pushing ahead with AI and do not understand why. Run a second one into the how, because the share of people who can use the tools already at their fingertips is low. Add incentives, the desire of the ADKAR model, through reward systems and show-and-tells where teams play back how they used AI and the pricing team wins the monthly award. Then measure and reinforce: do people understand the why, do they understand the how, are they using it.

The fifth segment is the one Kevin most wanted the audience to hear. An association executive he knows blacklists any author whose article trips an AI detector; a novelist lost his publishing contracts on suspicion alone; Kevin himself uses the tools as an editor for the newsletter and has said so. He cites Reif, Larrick and Soll’s 2025 experiments in the Proceedings of the National Academy of Sciences, where people who used AI were judged lazier and less competent, expected the judgment, and hid their use, and where the penalty shrank when the AI was clearly useful and when the judge used AI too. Elliott’s story is a communications lead on a recent transformation whose draft was excellent and carried the patterns anyone who works with these tools recognizes, including the sudden fondness for em dashes and the three-negatives-and-a-positive sentence. His feedback was to dial those back. The reply was, what are you talking about, I did not use AI. Reflecting on it, he realized the person was afraid: getting to an 80 percent outcome efficiently felt like cheating, and a lot of our history tells us that cutting the corner is cheating. A neutral organization that never thought about culture produces that denial on its own. So the leadership job is not to be negative, not to be neutral, but to actively promote appropriate and responsible use, with quarterly showcases and awards, until the response changes from I did not use AI to of course I used AI, and look what it found that I never thought of. Kevin adds his own detector story: a long-form newsletter article written by hand that a free tool scored as 75 percent AI, because academic long-form writing is what the models were trained on.

The reflection is the question Kevin asks every guest, framed around the pattern Elliott has watched for fifteen years: no organization has ever dropped a new system on its people, said here you go, and succeeded. If a leader rolling out AI this quarter could change one thing, Elliott’s answer is two-pronged. First, understand that this is the most disruptive technology of our lifetime and that we are at the very beginning of it, and pay the same due diligence and attention to the human journey as to the technology, with the same rigor and the same investment. This is not a back-end upgrade that enables efficiency; it is a disruption to ways of working, to thinking, to people’s sense of themselves. Second, take the time to decide what you want the AI culture of your organization to be, because cultures take on a mindset of their own unless they are nurtured. Kevin told him on air that the answer would be a clip, and it is. The episode closes with where to find Elliott, the free playbook, the diagnostic and the charter builder at futureofchange.com.

About Elliott Martin

Elliott Martin is the founder of Future of Change, author of the AI-Led Change Management Playbook: How to Modernise Change Delivery in the Age of AI, and creator of the GEROA framework for AI-led change (Govern, Embed, Reallocate, Optimise, Accountable oversight). A change management executive based in Melbourne, he has spent more than fifteen years leading change on large transformation programs in Australia, at organizations including AGL, BHP, ANZ, the Australian Taxation Office, Sigma Healthcare, Coles Group, New South Wales Treasury and the Department of Defence, on Salesforce, SAP, operating-model changes and now AI. His path into AI ran through art before it ran through work: an early text-to-image experiment produced the hybrid Amalfi Coast town that hangs over his fireplace, and the question of whether the tools had a place on his programs followed. Future of Change offers the playbook free online, along with an AI-led change diagnostic and a free AI Change Governance Charter builder that walks a team through its own governance questions in about five minutes. He is also the creator of the Eight Characters of AI, a mental model for the different roles AI can play in a person’s work.

Resources

Future of Change (Elliott Martin)

The AI-Led Change Management Playbook: How to Modernise Change Delivery in the Age of AI, by Elliott Martin (free online, with ebook, paperback and hardback editions)

AI Change Governance Charter builder (free)

AI-Led Change Diagnostic (free)

Elliott Martin on LinkedIn

Episode 032: The Measurement Ratchet, the Season 3 Opener

Subscribe to the Ideas and Innovations Newsletter (It’s free)

The Truth About Transformation: Leading in the Age of AI, Uncertainty and Human Complexity (Book)

Research Referenced in This Episode

Orlikowski, W.J., and Gash, D.C. (1994). Technological Frames: Making Sense of Information Technology in Organizations. ACM Transactions on Information Systems, 12(2), 174–207.
Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press. (Nobel Memorial Prize in Economic Sciences, 2009.)
Bainbridge, L. (1983). Ironies of Automation. Automatica, 19(6), 775–779.
Wrzesniewski, A., and Dutton, J.E. (2001). Crafting a Job: Revisioning Employees as Active Crafters of Their Work. Academy of Management Review, 26(2), 179–201.
Petriglieri, J.L. (2011). Under Threat: Responses to and the Consequences of Threats to Individuals’ Identities. Academy of Management Review, 36(4), 641–662.
Ravid, D.M., Tomczak, D.L., White, J.C., and Behrend, T.S. (2020). EPM 20/20: A Review, Framework, and Research Agenda for Electronic Performance Monitoring. Journal of Management, 46(1), 100–126.
Reif, J.A., Larrick, R.P., and Soll, J.B. (2025). Evidence of a Social Evaluation Penalty for Using AI. Proceedings of the National Academy of Sciences, 122(19).
Hiatt, J.M. (2006). ADKAR: A Model for Change in Business, Government and Our Community. Prosci. (Referenced by Elliott on desire as a condition of adoption.)

Key Takeaways

1

Faster Is a Fraction. The Real Opportunity Is to Become Better at What You Do, Not Just More Efficient

2

People Hold Different Pictures of What a Technology Is For, and the Picture Decides What They Do With It

3

One Chatbot Becomes Eight Colleagues: The Googler, the Devil’s Advocate and the Rest of the Ladder

4

Nobody Fears Change. Ask Any Lottery Winner. People Fear the Unknown and They Fear Loss

5

Governance Is the Weakest Part of Every AI Rollout, and the Fix Is a One-Page Charter at the Team Level

6

An Enterprise Policy Will Never Know the Questions a Team Needs Answered. The Team Has To

7

Leaders Respond to Value. A One-Page Business Case Gets the Niche Tool Approved

8

The Eight-Hour Question Has Three Answers: Bank It, Pile It On, or Reallocate. Only One Builds Trust

9

Let AI Do the Impact Assessment So People Can Go Have the Conversation AI Cannot

10

The Smart Answer to Did You Use AI Is Yes, With Governance and Accountable Human Oversight

11

Pay the Same Attention to the Human Journey as to the Technology, and Decide What You Want the AI Culture to Be

Season 3, Episode 42 Transcript

Available November 19, 2026

Episode 042: More Human, Not Just Faster: Why the Real Return on AI Is the Time It Gives Back to People

DURATION: 69 minutes
HOST: Kevin Novak
GUEST: Elliott Martin, Founder of Future of Change, Author of the AI-Led Change Management Playbook and Creator of the GEROA Framework
FORMAT: Guest Conversation
SHOW: The Human Factor Podcast

COLD OPEN

Kevin Novak: Think about the email that landed in your inbox this year. Starting Monday, the new AI assistant is on for everyone. There was a link to a training video. Then there was a line about being an AI-enabled organization. And then nothing. Now think about what happened on the floor. One person built a workflow that saves her team a day a week. And she’s not telling anyone because telling someone means losing the tools or getting handed a day more work. One person uses it to search and calls that advanced. And one person has decided it’s not for him, and he’s gonna wait for it to pass. The system got installed, the people got an announcement. And adoption is failing for the reasons that have nothing to do with the software. Anyone who has lived through a Salesforce rollout or an SAP cutover or a merger has seen this movie time and again. AI is just now the loudest version of it. Today’s guest has spent more than 15 years as the person who gets the technology adopted after leadership decided to roll it out. And he has arrived at the conclusion that almost nobody else is saying out loud. Using AI to do the same things faster is a fraction of what is possible. The real return is the time it gives back, and what you do with that time is where change actually happens.

INTRODUCTION

Kevin Novak: I’m Kevin Novak, CEO of 2040 Digital, professor at the University of Maryland, and author of the book The Truth About Transformation: Leading in the Age of AI, Uncertainty, and Human Complexity, along with the Ideas and Innovations weekly newsletter. Welcome to the Human Factor Podcast, the show that explores the intersection of humanity, technology, and transformation, along with the psychology behind transformation success. This is season three, the season all about practice. Season one was about seeing the human factor clearly. Season two was about what you do once you can see it. This season is about building it into how an organization works so that it lasts after the consultants leave and the sponsor moves on. Two threads anchor the season, the dynamic between artificial intelligence and human beings, which is the clearest case study of the human factor ever run. And it is running in real time. And measurement, because what you measure determines what you manage. Today’s conversation sits on the first thread and reaches into the second. Elliott Martin has spent more than 15 years leading change on large programs in Australia, at organizations including AGL, BHP, ANZ, the Australian Taxation Office, and the Coles Group. He has been the person leading the team that gets the technology adopted. Salesforce, SAP, operating model changes, and now AI. His route into AI didn’t start at work, it started with art. He got into text to image generation really early, and one of the first images he created hangs over his fireplace. It’s a coastal town that doesn’t exist. A blend of the two neighboring towns on the Amalfi coast where his grandparents grew up. After he did that, he then asked whether there were applications in the work that he does. He brought the tools onto his own programs. The programs, they loved it. The people, they loved it. And he noticed something almost everyone else missed. Being more efficient doesn’t help you connect with people. It only helps you do things faster. So that observation became a book, the AI-Led Change Management Playbook, and a firm, Future of Change. The spine of the book is a framework he calls GEROA. Govern, Embed, Reallocate, Optimize, and Accountable Oversight. It answers one question. How do you know if you are using this technology well enough to be transformative? And underneath all of it is a line from our first conversation. There’s a sweet irony in it taking artificial intelligence to make us more human. I love that line. Elliott, welcome to the Human Factor Podcast.
Elliott Martin: Thanks very much, Kevin. Great to be here.

SEGMENT 1: FASTER IS A FRACTION

Kevin Novak: I’m happy to have you. So, Elliott, I want to start on your own programs, before the book. You brought these tools onto a live transformation, and the response was strong enough that you wrote a book about it. But the thing that I keep coming back to is what you noticed while everyone else was celebrating the speed. And there’s some research here on why that happens. In 1994, Debra Gash at MIT studied a consulting firm that rolled out Lotus Notes, which was the collaboration tool of its day. And you’re younger than I am, but I remember that. They found that the technologists, the managers, and the people using it each held a different picture of what the thing was actually for. And Gash called those pictures technological frames. And the frames did two things. They shaped what each group actually did with the tool. And because the groups didn’t know their frames were different, the gaps never got discussed. So nobody was wrong. They were just using different maps of the same territory. And that’s 30 years ago. And it’s the same room that you walk into today. So one person sees a faster search engine, one sees a threat, and a few see something else entirely. So walk me through that arc. What did you see on those first programs? And when did you realize that faster was only a fraction of it?
Elliott Martin: Thanks, Kevin. So I think that whole notion about different people see technology through their own lenses and the framing that they inherit changes what they feel like they are capable of doing to the technology. To the point in the research that you mentioned. I think that’s something which is agnostic of technology and certainly something that I’m seeing within AI. As I’ve started to use AI more and start to speak to more and more people about how they can use AI on transformational programs. So many times over the last six to twelve months, I’ve had people come up to me and say, Elliott, I really understand that this technology is powerful, but I just don’t see how it applies to me. Or they might say a variation which is, Look, I really can see how this technology would be powerful for helping to improve the testing of this application. But in my work, it’s not really applicable. And I think that the reason for this is simply that people, first of all, they don’t understand the full breadth of how this technology can be used. Their framing for how they’re looking at this technology is different. And so, that last example where I had someone come to me and say, I understand how this might be powerful for automating testing as an example. They’re looking at AI through one lens where the person that’s saying, look, I’m just really not sure how it applies to me at all, is clearly looking at it through a different lens. And so everyone has different ideas about what it can do and what it can’t do. And I think that is largely based off the fact that in most cases, on most transformations, it’s never really been explained to the average person going through a transformation or just in business in general. Here’s what this technology actually is. Here’s the way to think about it. And here’s a spectrum of what it can do. And so this question is something that’s come up a lot over and over again. And it’s part of why I ended up creating a framework because I kept thinking to myself, when I’m speaking to people on transformation, how can I get them to understand multiple frames, to the point that you mentioned. And sitting down in front of a whole bunch of literature and trying to outline what this technology can do is certainly, I don’t think, the right thing to do. And so I ended up creating what I call the eight characters of AI. Now, this is nothing more than a couple of different characters that I think humanize to a certain extent the applications and use cases of AI. So if you like, I can maybe bring a couple to life. I don’t wanna go into too much detail necessarily.
Kevin Novak: And I actually have some questions for you related to it. One thing I want to dive into: You kept hearing one sentence from stakeholders, and it’s that I understand technology is powerful. I understand it applies to some people. But it’s not for me. So your answer was the keynote that we talked about called The Eight Characters of AI. And so, in essence, a ladder. It starts with a Googler where most people begin. Somewhere up the ladder is a devil’s advocate, which tests your thinking the way a good friend would. And near the top is McFly, which simulates a future. Feed it a stakeholder’s history and how they think and ask how that message is going to land before you walk into the room. And those are things that people, whether it’s defensiveness or their mind is closed, fail to seize as possibilities. So what changes in a room when people stop seeing one chatbot and start seeing eight different colleagues?
Elliott Martin: So I think the whole mindset as well as the outputs change entirely. So if we take some of the examples, let’s just take a real example. If I’m preparing for a board paper or a board meeting, something that transformations have to do periodically. If I’m responsible for trying to prepare this particular board paper, and I view AI as one of the examples that you mentioned, a chatbot. That’s what the character I call the Googler represents, someone that uses it as a question and answer tool, if you like. Then when I’m preparing for that board paper, I might sit down with my AI and I might say, I’m preparing a board paper. Could you please recommend the key things that I should include in this board paper? Now that’s useful. That’s gonna be productive to a certain extent. And that’s gonna be my interaction with this technology, if that’s my framing. Now, if I move on to the character that I call the debt goblet, which is really representing a use case of using AI to eat away at the administrative burden and produce outputs, not just answers. When I’m producing that same board paper, Kevin, I’m gonna start asking it to say things like, So here’s all the information that I think the board needs to know. Here’s the last example of that board paper. Here are my key objectives. Produce me an output. And so then my output in that scenario is going to be something tangible. If I have the understanding and awareness that the technology is not just able to answer questions and it’s not just able to provide outputs, but it can also act as a thinking partner. You mentioned the devil’s advocate, which is the character that I think represents that use case. Then I might sit down with my AI and I might say, here’s the board paper that I’ve prepared. Let me test my thinking with you. I hypothesize that this is going to be a risk. And my answer to that risk for this transformation program is going to be X. What do you think? Are my assumptions missing anything? And so the point here is that the way in which people will interact with this technology, even in this simple example, will be totally different. How they engage, what they ask, and also the outputs. And so I think that the key thing there is it’s not about using it in one way or using it in another way. I think people and companies should invest the time in helping people understand all the different use cases. All the different framings for how to view this technology such that they can apply it.
Kevin Novak: I have a question later about governance, but obviously someone needs to be mentored through that or helped through that understanding. And that person may not see themselves as that innovative or that open-minded, or there’s a rigidity to what they do. You have the experience where people say, hey, that’s not for me. And a lot of the human factor work says that’s identity threat. Someone is saying, no, it’s not for me because I think it’s gonna take over what I currently do and where my value is in the organization. So how do you lead someone through those use cases and where they’re internalizing those use cases to open themselves up to that?
Elliott Martin: That’s a good point. The thing I would like people to understand about this is it’s important to understand why people might be hesitant. I’m often told, Elliott, we understand that people fear change. I’m sure you’ve heard that before, Kevin, people are just resistant to technology. They just don’t like change. And I would say that actually, that may hold true in some cases, but actually I think it’s a bit of a fallacy. So why do I say that? There is a lot of good change that happens, Kevin. And so if you speak to all the lottery winners over the last 12 months and you say to them, You went through a massive change, you must have been hating that. You must have really been pushing back, when they gave you the millions. People aren’t gonna say that. They would be walking headfirst towards the Lamborghini dealership and they’ll be doing that smiling, even though it’s a massive change. We can take from that it’s not so much every change that people resist. I think people resist primarily two things. They fear two things. One is they fear the unknown. I don’t know what’s gonna happen. What does this mean for me? I don’t know, therefore I’m scared. They fear the unknown. And the second thing they fear is loss. Does this make me less capable in my job? Does this make me less necessary in my job? And so I think when businesses and leaders in general look at it through this lens, then they can start to ask themselves, well, why might my staff be resistant to this? Is it because they don’t understand it? And if that’s the case, you can tackle that problem through an education campaign. Is it because they’re feeling some loss of control? Is it because they feel like they are less relevant in their workplace as a result of this? If that’s the case, then we need to make sure that as leaders and as humans, we can do what we do best, which is empathize and explain that this technology is not designed to take away your importance within an organization. Rest assured this technology is designed to enhance what you can do within this organization.
Kevin Novak: So with the ladder, your eight steps in essence, with that being the basis of what may be going on in somebody’s head. And there’s great utility in what you’ve put together. To help people understand the who, what, when, where, and why and really leverage what is now available to them. But then how do you facilitate that conversation where it is a positive, but they may be feeling anxiety, they may be feeling negative because of what is out there in the press or even within the organization.
Elliott Martin: The key message that I would encourage people to understand is that things are not always binary. Things are not necessarily good or bad. And very rarely is that the case. And I think certainly that is the case for AI. And people in general tend to hear something. And then they attach mentally onto it and then all of a sudden everything’s doomsday scenario. AI has gone from we should think about AI to now the thinking is gone. People for the most part are on board, like I was saying. Let’s go. Maybe they’ve implemented it and they’re starting to use it, but they don’t have an understanding of how it should be used and what good looks like. And so this is where the GEROA framework that I talk about in the book is designed to cover off the key things that I think leaders, not just within the change community, although that’s really what I had in mind when I was building it, but more broadly, need to understand: what does good look like? So to your question, if I’ve got people that might be resistant, or even on the other end of the scale, how do leaders, how do organizations move them? You have to understand what good looks like, and people need to understand that. First of all, that there’s governance, which we can talk about, is a big topic. I think it’s something that organizations aren’t doing very well. But then also helping people realize that look, this technology is not just about enabling you to do what you do today faster. This is actually a big centerpiece of a lot of the work that I do. Because you can understand, Kevin, that if I gave you technology or anyone a technology that said, hey, guess what? This technology is now going to enable you to work at twice the speed. The first thing you might think is great. But then you only have to pause and you think about, well, hang on, what does that mean? Does that mean that the next step is I’m not required at all? That’s not a big leap.
Kevin Novak: No, it’s not.
Elliott Martin: And people that are resistant or that might be struggling are probably thinking this in their minds. And so I think what I try and encourage people to look at is the opportunity here is not just about how we do everything faster. All we end up doing if we do everything faster, and this is a cautionary tale for people going through transformation, is you’ll become more efficient at whatever it is that you do. But the real opportunity here is not just to become more efficient at what you do, it’s to become better at what you do. It’s to do things differently that improve the nature of what it is that you’re doing. And that applies in the discipline that I focus on, being change management, but I think it applies across the board. So that’s what I’m seeing.

SEGMENT 2: SHADOW AI AND THE CHARTER NOBODY HAS

Kevin Novak: So into that, because when we talked, you brought up you’re hearing a lot about shadow AI everywhere you go. So advanced users are out on their own with one tool or another, with somebody two desks over paying for a completely different subscription and nobody knows what the other’s doing.
Elliott Martin: Exactly. All the same subscription.
Kevin Novak: So your prescription is deliberately basic: a responsible owner, the tools that are approved and the ones that aren’t, what they can and can’t be used for, and an escalation path. And I think to your point, that’s what organizations are struggling with. And then you said that almost nobody you ask has that at a team level. And many don’t have it even at the organizational level either, despite the speed of which they’re trying to move forward. And based on some of my own research, there’s a reason the team level matters. And it comes from an unlikely place. So there’s a researcher, Elinor Ostrom, who has spent her career studying how communities manage shared resources. And her focus was fisheries, irrigation systems, grazing land. And interestingly, when I was reading through this, she won the Nobel Prize in Economics in 2009 for all this. And her finding is really interesting in that it cut against what everyone expected. The rules that held were not the ones handed down from the top. They were the ones made close to the people who had to live under them and then nested inside the larger system. Local rules backed by the organization above them. And that’s your charter in essence. An enterprise policy will never say how the finance team should use these tools versus change teams versus the people and the culture team. So the team has to. Why does something this simple go undone? And what happens in an organization that leaves governance to the enterprise policy alone?
Elliott Martin: That’s interesting research, Kevin, because it also highlights and validates the point I was making earlier about people inheriting information better from their leads, that idea of nesting. So that is interesting. In terms of governance, yes, so the shadow AI is a big problem. But to your point, there is a simplicity that can solve it. Certainly it sounds like your experience is similar to mine in that the governance element of AI is the element which is the least well implemented or the least developed along the journey. And I’ve also got some research that I’ve done at Future of Change, which also validates this. There’s a change diagnostic where people from across the globe can go in and answer a couple of questions to try and understand because one of the things I’m also trying to do is understand where people are on this journey. The questions are based on each element of the framework, and the element that is the worst at the moment, based off the data that I have, is governance. I think the data is clear. Certainly from what I’m seeing, what people are telling me as well. And there is a simplicity to fix it, because it’s something which doesn’t require a complex solution in my mind. So there’s two things. What does it require? And then what are the different levels? I’ll go through my views on this.
Kevin Novak: Please, yeah.
Elliott Martin: I think in some ways it’s basic, but it’s important. What are our rules for what we can use and what we can’t use? The what. So it might be we use these tools, but not these tools. Because if we’ve invested in the Microsoft licenses, we don’t then also want to pay open AI or whatever it might be. So very simple understanding of the what. And if you go into a lot of organizations today and you ask people, what AI tools does your organization allow you to use and which ones that you don’t, a lot of the times they’re going to look at you with a blank face. And then, regardless of the tool, what is the data that you are permitted to use? Are we allowing us as a team or as an organization to input personal information? Are we allowing ourselves to put documentation? Are we allowing all information regardless of sensitivity? So I think in that sense, again, it should be simple to answer, but most people don’t take the time to write it down. Then we have our what covered. What we then need to understand is what an escalation path and a review path look like. So if we use AI, what is the review policy? Based off what we put in, do we think someone needs to review it or not? The same way, if we were doing anything, Kevin, and especially with your experience from the academic world, we have a peer review. We don’t write essays in isolation and then say, great, I’ve nailed it. You would invoke a peer review. And the more important that your essay or dissertation is, the more people you would want reviewing it. And so that same thing is simple AI governance. Based on how this is going to be used, what is my review process? If I have to ask a question and I’m not sure, who’s the person to go to? And these basic questions, I think, solve a lot of the problems that organizations have, including Shadow AI. If I had a simple governance charter that says, here’s what we can use, here’s the information that we can put in it, here’s the responsible owners. And if I’m not sure what to do, here’s who I go to. All of a sudden, you don’t have different teams using either different examples of the technology or the same examples of the technologies and paying the same company licenses twice. So that’s the first element. I think the good news is, despite it being clearly the number one element where organizations are lacking, I think the solution is a relatively simple one. Take the time, ask the questions, invest in creating a governance charter. But the second thing I would highlight, which absolutely would link back to that research, is that what organizations do have a lot of the time is a data governance policy. In some ways, what they’re starting to do is they’re starting to expand their data governance policy to include the use of AI. But the data governance policy is almost exclusively at an enterprise-only level. I’ve never heard of or worked in an organization that had a data governance policy at a team level.
Kevin Novak: No, and we’ve done that work, it’s always at the enterprise level.
Elliott Martin: Exactly. And that one is probably well justified. But when it comes to AI, what I advocate for is governance at a team level. And so for example, what I say to the change management community is, as a team, if you’re responsible for the management of whatever the transformation is, have a dedicated change management governance policy. Because I can guarantee you now that the executive leadership team, even if they do come up with an enterprise policy, will not know the questions to ask that we need at our team levels.
Kevin Novak: Yes.
Elliott Martin: There are niche tools for AI that I talk about in the book that makes sense to me from a change management perspective. There will no doubt be other unique AI tools that teams in other disciplines will want. How is the enterprise going to know that? They’re not going to know that. So you need to have that at a local level so that it can be fit for purpose for the work that the teams are doing. Which is exactly your nesting theory.
Kevin Novak: So, Elliott, my question for you, is that most leadership over the past 20 years has been trained that they need one to three systems in an organization. Whether that be Oracle 12i, whatever the number is, or SAP, and even where they’re serving financial, asset management, product management, whatever functions, it’s one to three systems. And so what you’re bringing forward is that there are many niche AI systems to serve different purposes, and whether it is accounting, finance change management, or so forth. So, how do you guide leadership to understand that it is different and it is not just the past 20 years and this is what you do, that there needs to be the consideration of more applicable tools for the different teams?
Elliott Martin: I think the good news, Kevin, is that leadership responds well typically to value, but they need to understand that value. And so what I would recommend, which is certainly what I have done, and I can give you an example of this, is to highlight that value via almost a business case for endorsement. So as an example, you’re totally right that there are just too many niche AI applications for anyone at an enterprise level to say, here’s the exclusive list of tools that we use. They just won’t know enough about the various disciplines across any organization. So in my case, there was a tool that I discovered, over 12 months ago now, called NotebookLM, which has now been renamed Gemini Notebook. And I loved this particular tool. For its ability to generate AI podcasts. And this was, I think, relevant to the work I was doing because I wanted to deliver information to audiences, a lot of what we do in change management. And I really liked its ability to take complex concepts and make them relatable in a conversational style. That’s part of what this tool is able to do, and for my work that was obviously beneficial. But saying that to the program director at the time wouldn’t really have meant much because how would he know? So what I had to do is I put together a mini business case and I said, look, here’s something I want to run past you. This is the nature of the complexity. Here’s the feedback we’re getting from the business to say that they’re not quite understanding the level of complexity. It’s not quite resonating with them. And here’s a simple, cost-effective manner, which is certainly low effort in the sense that it is not laborious from a human effort perspective, that can solve this problem. And the answer, the reception that I got, Kevin, was, if not overwhelming, very well received, let’s say. Why? Because he appreciated the fact that I had done the research to understand what was at the time something relatively new and solved a problem that benefited everyone on the transformation. And so this example is gonna be different for everyone. But I think we can be happy with the fact that leaders will respond well to value. And if we can highlight the value in a given AI application, bring it to life, highlight the problem that it will solve. People on transformations are there to solve problems. And so I think it will typically be well received. Highlighting it through a business case will get the outcome that people are looking for.

SEGMENT 3: THE EIGHT-HOUR QUESTION

Kevin Novak: So I want to jump in to the framework. And you told me that your favorite part starts with a phone call that I get asked about more than anything that I’ve said on this show. So a marketing leader called me earlier this year, and her team had used AI to win back roughly eight hours a week. And there was no organizational policy, no approved tool set, and a real fear that she had created shadow AI in her own department. And she was silent. Because telling anyone meant one of two things, either lose the tools or get handed eight more hours of work. And you hit on this earlier. And for the first time in years, she actually had time to think. And she couldn’t say so. So you told me that is exactly the question the reallocation stage of your framework exists to answer. And you described two honest options: bank the savings, which means fewer people doing the same work, or reallocate the work, which means the people stay and the time goes somewhere better. So there’s a paper from 1983 that seemingly saw this coming. And so Lisanne Bainbridge, a psychologist studying control rooms, wrote a short piece called Ironies of Automation. And her argument was that when you automate the routine parts of a job, you don’t remove the human. You leave the human with the hardest, most judgment-heavy parts of the work. And then you fail to design for that. So the machine takes the easy 90% and hands back the hard 10. And so there’s research on what people do with the time when they’re allowed to keep it. Amy Wrzesniewski and Jane Dutton in 2001 studied what they called job crafting. People who are allowed to reshape the boundaries of their work, the tasks, the relationships, what the work is for, reshape its meaning as well. So the job gets better because the person redrew it. So make the case to a leader whose board is measuring what I call tonnage. What does reallocation look like in practice? And what happens to trust in an organization that answers the eight-hour question the old way? Because I know you’re seeing this over and over again.
Elliott Martin: Absolutely. This is a great question, Kevin. I think there’s actually almost three forks in the road here. I’ll explain them briefly and then I’ll outline what I advocate for. And you covered off a few of these. So one option, as you mentioned, is you just bank the savings, as I call it. This may sound attractive to some organizations, to some leaders. We’ve got to be realistic. A lot of transformations, they have budgetary pressures. They have pressures to deliver, they have budgets. And so option one, using even that example that you gave, they might look at that and say, what this means is we get eight hours back. That’s one FTE, one full-time equivalent resource. So now we can do the same work with one less. They bank the budget and it looks like they’re doing the same thing for cheaper. So there might be some scenarios where people find that attractive. All it’s really changed there is a few dollars and cents at the end of the day. So keep that in mind as one scenario. You’ve effectively saved the cost of labor of a single human. You haven’t changed your processes, you haven’t improved the ways of working. And you certainly have not increased the trust in the people within your organization. If anything, you’ve decreased that trust and people are gonna start to think, geez, Jane over here was working so efficiently through the use of assisted technology that now she’s no longer here. What does that mean? It means for me, maybe I’m gonna be next. Therefore, one of two things, I should probably start looking somewhere else, or alternatively, I better not aim to be so efficient. So you can see that would be one path, which doesn’t sound very appealing in the long run, for the reasons I just said. The next option would be to say, great, we’re not going to bank the savings, but what we’re gonna do is we’re gonna do more. We got eight hours back. So that means we’ve got eight hours of additional work. Again, this may sound attractive, no loss of humans. No one’s losing their job here. We’re just doing more work. That means more values for clients, more values for customers, whatever it might be. And so in that scenario, I’m gonna start allocating Jane more work. And she’s not working longer, she’s just doing more work in the same period of time. Those two paths cover the exact fears that you mentioned in that example. But there’s a third option, Kevin, which is the one that I advocate for. And as you mentioned, it’s the R in the GEROA framework, reallocation. And so what do I mean by this? Reallocation of work. It’s the third path, which is to say: saving a few dollars in option one is attractive until it isn’t, because there’s a whole bunch of other costs in terms of the erosion of trust with my people, and actually the dollars don’t add up anymore. Option two sounds attractive, but what you’ve really done is you haven’t improved. If Kodak had decided that they could produce manual photography faster, it wouldn’t have stopped them from going bankrupt. What would have stopped them from going bankrupt is having the time to step back, look at the future, do the thinking, and innovate. And so, what might that look like from a reallocation perspective? What you do is you don’t get rid of that person, you don’t give them more work, but how can we actually use the technology to reallocate how, as humans, they best do the job that they do? Amazingly, that research was from the 1980s, good on them. Effectively, let them help govern and take some accountability for how they could work better. Because then what you end up with is not faster processes, you end up with better processes. So by real life example, in the change world, which is obviously where I live, one of the big things that we do is we try and understand the impacts of change on an organization. Not rocket science. In order to do that, you need to understand, well, what are people doing today? And then what are they going to be doing tomorrow? And to what extent is that, from a technological, a personal or cultural perspective different. Now, once we really started using AI a lot for this, because the application for AI here is obvious. If you’ve got current state processes, future state processes, you can get 80% of the work done. I was literally leading teams where I was faced with this same problem. And then instead of banking savings and getting rid of resources or just saying we can do the same stuff faster, what we realized is, let’s use the AI to do the impact assessments. But then let’s take that capacity and use people to do what AI can’t do, which is to actually go into the business and have conversations with our stakeholders. Because that is something that AI can’t do. And it’s something that in the change management discipline we don’t have enough time to do. Go and have a conversation with people in the business. Ask them how they’re actually feeling about moving from A to B. Not because we have a process diagram that says this is where the difference will be. Have the conversation, which teams typically won’t have the time for. But if we did have the time, wouldn’t that be a better way of doing things? Wouldn’t you ultimately get better insights and a better outcome? And the answer to that is yes. And it’s already playing out real time on transformations that I’m supporting. And so I think the key takeaway there, Kevin, regardless of what the discipline is, yes, you have three options. Initially, I would say, they might sound attractive. You might save a dollar here, but look at the long-term cost. Ultimately improving, at a fundamental level, the way in which you work will result in more dollars and cents than the one FTE that you saved. And to the point about the trust equation, of course it goes without saying that if you only look at things through the first two scenarios, I’m going to allocate more work to my people or I’m going to get rid of them, of course, the trust is going through the floor. And it’s two ways, because the staff will not trust the leadership. But also, frankly, there will be an incentive for staff to almost be dishonest about the efficiencies that they are gaining. And who benefits from that? Zero. No one. So when they look at it through that lens, the case really should paint itself.
Kevin Novak: So do you think that they see that too complexly? Or, you made this statement earlier, and I align with it, that the world is gray, there is no black and white, there’s this area in between. In what you’re suggesting to them, do you think they may see that no, it’s just easier to focus on the efficiency opposed to the effectiveness.
Elliott Martin: Part of it is that. And part of it, Kevin, is that one of the things we’re not good at doing as humans is seeing the bigger picture and the long term view. Because the pressures are always immediate. The focus is around what the board is going to say at my next meeting. How many people are more worried about where the company’s going to be in three years’ time than they are about the share price at the next market interaction. Not very many. And so that I think feeds part of the problem because they’re looking at it and they’re saying a dollar saved now is a better financial report at the end of the financial year. And that’s the frame for a lot of them, and understandably so, because that’s where these incentives are, that’s where the bonuses are coming in. But if you look at the leaders of the great companies over the last ten years, it was never focused on six months out. The visions were always multi-year. And so I think that there’s learnings to be taken there from people that are a lot smarter than me.
Kevin Novak: But in such a disruptive time, do you see our defaults in those regards actually as an inhibitor to the future?
Elliott Martin: I certainly think that our proclivity to focus on the immediate is an inhibitor for sure. However, because things are moving so quickly, it should be easier or more intuitive for business just to understand that if they’re not taking a long-term view, they’re not going to be competitive. One thing that AI should be forcing people to do, and this is why I think it’s moving so quickly, and why organizations are in many cases throwing it in without the proper due diligence to make it successful, is because, maybe not for the first time, it’s a very obvious example where you can see immediately a scenario of two years down the track, if you’re not getting on board, you’re totally not competitive in existence at all. So our proclivity to focus on the near term is I think for sure an inhibitor. AI is an example where it’s actually making the case for the longer term view and the importance of it clearer than maybe other technologies in the past.

SEGMENT 4: NOBODY FEARS CHANGE. THEY FEAR LOSING CONTROL

Kevin Novak: So I wanna bring back your lottery example because your statement is that the fear is really the loss of control. So you’ve watched it in mergers where people decide they are no longer relevant. You watched it in an SAP implementation where the system now does what a person used to be good at. And AI is only the latest and loudest example. And a researcher at INSEAD published a piece in 2011 on what she called identity threat. So when something threatens a person’s values, the work they are proud of, the role that tells them who they are at work, they respond one of two ways. So they protect the identity, which looks like resistance, or they restructure it, which looks like growth. And her point for leaders was that organizations rarely know which response that they’re producing. So same rollout, same email produces both, and different people in the same room, which means an identity threat and a loss of control are not the same thing. And most rollouts can’t tell them apart. So your argument, Elliott, is that executives are doing the right thing by moving quickly and failing on the one thing that really matters. And it’s a tailored strategy for helping people see that this isn’t a loss. So what does that strategy actually contain that you would recommend that they follow?
Elliott Martin: The good news, Kevin, is who better to help organizations manage that transition than the change management practice.
Kevin Novak: I very much agree, yes.
Elliott Martin: What people miss is that AI is, exactly as you said, the perfect example, maybe the best example, of disruptive technology of our era. And so it needs to be managed with not just the same due diligence and methodology, but extra attention should be paid to the due diligence and the methodology that change management applies. We need to focus on multiple streams of work here. So one thing is understand the root cause of the resistance. Is it fear of the unknown? Is it fear of loss? And then organizations need to systematically put in a tailored approach. So part of that is going to be an education campaign, which is not just about how to use the technology, but why the organization is going down this path. Because I can tell you, Kevin, a lot of people are right now, they’ll be listening to this podcast, they’ll be working in an organization that is pushing ahead with AI, and they won’t really understand why. And if you don’t understand the why behind it, how are you really gonna be able to support and get on board yourself? So there needs to be an education campaign into the why. There needs to be an education campaign into the how, which is, now we understand why this is a good thing for our organization and the industry, maybe more broadly, but certainly within our organization, they need to understand that. Then they need to understand the how. We need to give people the tools as well as the incentives to implement this in their day-to-day work, just because they think it’s a good idea. Strategically from an organizational perspective doesn’t mean that they then have the tools to implement it themselves. So that requires a dedicated education piece again, if we take any of the frontier models, a lot of companies are using Copilot or Claude or ChatGPT. Now, admittedly, things are changing all the time, new functionalities added all the time. But the amount of people in organizations today that understand how to leverage the technology that they currently have available at their fingertips is very low. Partly because it’s being refined all the time, but partly because organizations have not done the due diligence that they should do to get people to understand the how. So they need to understand the why. They need to understand the how. Then there’s ideally the incentives. Just like anything, even if people understand why strategically it’s important and they have the tools and skill sets to do so, is there an incentive for them to do so? This is the traditional desire that change people talk about as part of the ADKAR curve. People need to have a desire to use it as well as the know-how. There’s different ways of doing things. Organizations I’ve seen that are doing a good job of this are actually using reward systems as well as show and tells where different teams across the organization are playing back, here’s how we have used AI. And here’s all the different use cases that teams across the organization have used AI for, and the monthly AI award has gone to our pricing team for their innovation. So you have to have incentives. And then I also think there’s a measurement and reinforcement layer as well which is organizations need to know where people are on each of these. Do they understand the why? Do they understand the how? Are they actually using it? All these things, which will be traditional change management methodology and approaches, may make sense to people in my community, but it’s not intuitive in the least for your broader people working on transformation. Although hopefully when you look at it through that lens, it’s logical.

SEGMENT 5: THE SMART ANSWER IS YES

Kevin Novak: So let me take this in a different direction. So an association executive I know announced that any submitted article flagged by an AI detector immediately puts the author on a blacklist. And recently, and I wrote about this a couple months ago at this point, a novelist lost his publishing contracts on simple suspicion that he used AI. And I even use these tools as an editor for the newsletter. I decided whether to say so, and I have said so: I’m doing this, and it does the same thing that my human editor used to do and doesn’t let me write certain things. So there’s now some hard evidence for the thing everyone seems to sense. And in 2025, Jessica Reif, Richard Larrick, and Jack Soll at Duke published four experiments in the Proceedings of the National Academies of Sciences, and with more than 4,000 participants. So people who used AI for a task were judged lazier and less competent than people who used ordinary tools or did it by hand. The people using AI expected the judgment and then hid their use, of course. And two things made the penalty shrink or simply disappear. So when the AI was clearly useful for the task, and when the person doing the judging has used AI themselves. So your position, Elliott, is that the question leaders should be asking is not whether someone used AI. And that the smart answer is not no. The smart answer is yes, with proper governance and accountable human oversight. So it’s the G and the A of your framework. And you were equally clear that what you call cognitive laziness is real. And I think we can all admit to that on some regards. And it’s that AI will speed the descent for anyone already on that spiral, and that the fear of the irresponsible few should not govern how everyone else is allowed to work. So how does an organization draw that line without a detector or a blacklist? And what does it cost an organization when its best users go quiet because of all that?
Elliott Martin: That’s a good question, Kevin. And I can give you an example of where I’ve seen this, to bring it to life for the listeners.
Kevin Novak: Please, that’s what I’m asking.
Elliott Martin: And then I’ll give you my view on how organizations should tackle this, without a stick, and maybe more of a carrot approach. In a team that I was leading in the last 12 months, it was a large-scale transformation. And there was someone in the communications team who had presented a particular piece of communication for an end user audience. And I was reviewing it. And it was very well written. And this person always does great work. But those people that have spent a lot of time with AI, particularly in getting it to assist with drafts, know that regardless of the model, AI typically has some patterns that it likes to implement, things like the excessive use of em dashes. Never have people used em dashes more. All of a sudden everyone’s a big fan of dashes.
Kevin Novak: All of a sudden now they’re everywhere.
Elliott Martin: Or the sentence construct of three negatives and a positive, where it will construct sentences and it will say, the benefit is not this, it’s not this, it’s this. Anyway, there are little patterns that those who have spent a lot of time pick up. And so my feedback was, look, this is excellent. All the key messages that we needed are there. But I think we should dial back some of the use of these and the use of this sentence construction, because using AI for drafting is fine, but we don’t necessarily want it to become so rigid that it becomes obvious. And the response that I got was what are you talking about, Elliott? I didn’t use AI for this. And I said okay, moved on, and gave my feedback. But when I was reflecting on that, I realized that people, in this case, this person was just afraid of the use of the application. Almost like there was a view that leveraging the technology to get to an 80% outcome efficiently was cheating. And I think that a lot of our history leads us to believe that because it seems like if we are able to cut a corner, in many aspects of life, if we are able to take a non-traditional path to get somewhere in a shorter time, in a lot of cases we would see it as cutting a corner. It’s cheating, in many aspects. And so there is this mentality in the workforce, like the examples that you gave, where editors being blacklisted and this sort of thing are proliferating a narrative that the use of AI is cheating. And so, first of all, I think organizations need to be aware of this, that this may not be something that they’re necessarily driving, but it might be the default human, believe it or not, natural interpretation of the use of this technology. And so one of the key things that organizations need to do is they need to focus not just on here’s the why and the how, but also on the cultural element that goes along with it. It doesn’t necessarily take the example you gave, where organizations are blacklisting people for using it, which has many problems. Even in organizations that are not actively promoting it as negative, the scenario I gave highlights that even in a neutral organization that just didn’t think about the need to promote it culturally, people may have a negative resistance to highlighting it. And so what does that tell us? It tells us, well, there’s a cultural piece that leadership within organizations need to take on board, which is, first of all, not to be negative, secondly, not to be neutral, but to actively promote it and encourage the use of it. And so, for examples like I was giving before where the leading organizations are actively rewarding the appropriate and responsible use of AI, and I say that because it’s always important, via things like, let’s showcase our innovation. Let’s have a quarterly event where different aspects of our organization come together to showcase their use cases for this technology. And then there’s gonna be an award for the most innovative use of appropriate and responsible AI. All of a sudden, in examples like the one I just gave, the response will stop being I didn’t use AI, and become, of course I used AI, I’m being up front, I’m being transparent about it, and look how much time I was able to save. Or here’s how I was able to use it as a thought partner, like the devil’s advocate. Guess what it highlighted that I never thought of. Actively be proud to highlight these things. And that’s all about the culture element. So that would be my advice.
Kevin Novak: And I think, Elliott, we’re jumping the hurdle too quickly. Even with the detectors, there is so much error. On a personal level, I have always written in more of an academic long form. There was an article that I wrote over the summer, and I think it was Grammarly who introduced some new tool and it was free for a period of time. So I go to publish the newsletter on LinkedIn, and it’s telling me that 75% is AI generated. And I’m sitting there thinking, I just spent how many hours writing this thing? No, it’s not. The interesting thing is that that kind of writing is what AI was trained on. So yes, it’s spitting out something that’s very similar to a portion of people out there who write a certain way or even in academia. And I know there’s universities that have turned off the tools because of the error rate and the situation that creates with the students. That could be a whole separate episode.

THE REFLECTION

Kevin Novak: Elliott, this takes us to the question I ask every guest at the end. And with you I’m gonna frame it around the pattern that you’ve watched for more than 15 years. And no organization has ever dropped a new system on its people, said, Hey, here you go, and succeeded. And you said that is why you have a job, and you have been very successful at it. So whether it’s Salesforce, SAP, operating models, and now AI, then the failure is never the software. So if a leader rolling out AI this quarter could change only one thing about how they are doing it, what would you have them change?
Elliott Martin: There’s going to be a two-pronged answer to this, Kevin. What I would say to the many leaders that are going through this, because many are at the moment, is first of all, take the time to understand that this technology, as I mentioned earlier, is without exception the most disruptive technology that we have seen in our lifetime. I’m convinced of that. And I think we are also at the infancy, the very beginning, of this disruption. And with that understanding, when looking to roll out the use of AI in the organization, please pay the due diligence and attention to the human factor. The human journey that needs to accompany the technology. Just understanding that alone and the importance. Do not look at this as a back-end upgrade that enables efficiency. No, this is a massive disruption to ways of working, to thinking, to people’s mindsets, to their fear, their sense of themselves. Therefore, along with the investment that goes with getting the technology right, it needs to be accompanied by the same mental rigor and investment for managing the people journey. That would be my first pitch. And the second point, which I think is equally important and often misunderstood, and this was what we were just touching on, is to understand and take the time to invest in getting the culture of what AI is within your organization right. Because it is a cultural shift, as well as ways of working and technology. And cultures, as leaders should know too well, will take on a mindset of their own unless given the proper nurturing. So to the point we’re just speaking about, do you want your people to be embracing this technology? Do you want them to be a little bit more restrictive in their use of it, depending on the nature of your application. We need to create the right culture for this implementation. So those two things I think are gonna set people up for success. Make an effort to determine what you want the AI culture to be, and definitely realize the human factor that needs to be well managed, along with the funding for the token limits.
Kevin Novak: I love how you frame that. I can tell you that’s gonna be a clip for the episode. Elliott, thank you. This has been a really great conversation. And so before we wrap, is there anything that you want to leave the audience with? And where can people find you, the book, and the charter builder?
Elliott Martin: Thanks first of all, Kevin. It’s been great to have this conversation. I’m very happy to be invited on the podcast, so I appreciate it. In terms of where people can find me, I’m available on LinkedIn. Feel free to reach out. One of the great things about the book has been members of the community reaching out and sharing thoughts. I love conversations like this, so feel free to reach out. You can go to futureofchange.com. The book is a free resource, so it can be accessed there. The diagnostic I mentioned earlier, the frameworks, all that stuff is available. There’s a lot of free tools to help the community. By all means, check it out, and I’m happy for people to reach out with questions or if they just want to have a chat.
Kevin Novak: Elliott, thanks. I have to have you back in season four. This has been a great conversation.
Elliott Martin: It’ll be an absolute pleasure.

CLOSING

Kevin Novak: I want to close by connecting what Elliott has shared to the arc of this season. Season 3 is about building the human factor into how an organization works. Elliott has given us the clearest version of that yet, of what that actually means with AI. Not faster, but more human. The time the technology gives back is the raw material, and reallocation is the decision that turns it into something people can feel. Orlikowski and Gash showed us that people in the same organization hold different pictures of what a technology is for. And they never discussed the gap. Ostrom showed us that the rules hold when they are made close to the people who live under them. Bainbridge showed us that automation hands the human the hardest part of the job and then forgets to design for it. Wrzesniewski and Dutton showed us what people do with time they are allowed to keep. Petriglieri showed us that the same rollout produces resistance in one person and growth in the next. Ravid and colleagues showed us that people react to why they are measured, not whether. And Reif, Larrick, and Soll showed us the penalty for admitting you used AI. And that the penalty disappears when the people judging you use it too. If your organization rolled out an AI assistant this year, the question Elliott might ask is simple: Where did the eight hours go? And does anyone on your team feel safe enough to tell you? If you found today’s episode valuable, subscribe to the Human Factor Podcast wherever you watch or listen. Leave a rating and a comment and share this episode with your leadership team. Subscribe to my Ideas and Innovations newsletter at 2040digital.com or on Substack for weekly frameworks and research on why change succeeds or fails. And connect with me on LinkedIn, where I post regularly about the psychology of transformation. Until next time, remember, it is never the technology, the market conditions, or the strategy that makes change and transformation successful. It’s the people. And the more deeply you understand the human factor, the more likely your transformation is to succeed. This is the Human Factor Podcast. I’m Kevin Novak. Thank you for watching or listening.

Kevin Novak, Founder and CEO of 2040 Digital

Kevin Novak is the Founder & CEO of 2040 Digital, a professor of digital strategy and organizational transformation, and author of The Truth About Transformation. He is the creator of the Human Factor Method™, a framework that integrates psychology, identity, and behavior into how organizations navigate change. Kevin publishes the long-running Ideas & Innovations newsletter, hosts the Human Factor Podcast, and advises executives, associations, and global organizations on strategy, transformation, and the human dynamics that determine success or failure.