That Change Show

Revolutionizing Change Management with AI's Invisible Forces

January 14, 2024 Lean Change Management Season 2 Episode 1
That Change Show
Revolutionizing Change Management with AI's Invisible Forces
Show Notes Transcript Chapter Markers

Watch on Youtube: https://www.youtube.com/watch?v=52db0KvcZj8

Discover the transformative influence of AI on the landscape of change management in our latest episode, where we chart a course through the uncharted territories of innovation and adaptation. I share my firsthand experiences since 2022, cutting through the misinformation fog and showcasing the six pivotal areas where AI is reshaping the change management arena. From crafting dynamic training to analyzing intricate data, and streamlining communications to enhancing project oversight, this episode is a treasure trove of insights into AI's role as both a strategic tool and a collaborative companion for change professionals. We tackle the often-contested debate about AI's impact on jobs with an optimistic lens, positioning AI as an ally that augments our abilities rather than eclipsing them.

Embark on a narrative journey into the heart of AI's contributions to content creation, a theme that's central to my latest literary venture and professional toolkit. I recount the adventures of teaching AI to emulate my writing style, a time-saving marvel that, with just a pinch of fine-tuning, revolutionizes content production. We'll explore how AI breaks down technical jargon to foster clear communication across diverse teams, retains the context of previous prompts to assist ongoing projects, and even ventures into the realm of visual content with varied results. By the end of this episode, you'll be invited to join a conversation on AI's burgeoning role in change management, leaving with a renewed perspective on AI's potential to complement and enhance our roles as change agents.

Jason Little is an internationally acclaimed speaker and author of Lean Change Management, Change Agility and Agile Transformation: 4 Steps to Organizational Transformation. That Change Show is a live show where the topics are inspired by Lean Change workshops and lean coffee sessions from around the world. Video versions on Youtube

Speaker 1:

Alright, here we go, season 2, episode 1 of that change show. So I guess really the only reason for calling this season 2 is just because I haven't done one in a while. So I was putting these shows out every week. So normally every Sunday I'd go out for a walk, think of a topic and try to do a short, 20-30 minute show that would give people some actionable insights for what they might want to try, given. Whatever the topic was and for those of you that don't know, I moved last year and it seems like everybody I know moved last year so life got in the way and it was basically moving, selling stuff, packing stuff, cleaning houses, doing a bunch of stuff for a lot of the year. So things have kind of calmed down a little bit. It's a new year, I've got a new book coming out next week, january 22nd, whenever and thought I would get back to doing these on Sundays. So let's cut the preamble and let's get into this week's topic. So the next book, like I mentioned, is about how to use AI and change management. So I've been using AI since yeah, depending on what you want to call AI, I've been using it since, you know, 2022, I guess, and just kind of lurking and watching how it's affecting organizations, particularly in a change context, and there's so much misinformation out there. I started working on some content last week and decided, hmm, you know what this actually should be a book, because I can't cram all this stuff into a short blog post. And really what it's kind of intended to do is dispel some of the common myths with AI and change management, because it depends on who you're listening to. On LinkedIn, you know the people who are on the cutting edge of technology say it's going to revolutionize things, it's going to change things, it's going to take jobs and of course, the people who are in those jobs obviously get angry about that. No, ai can't do what I can do as a human and all this stuff. It's just a fancy search engine. It can't replicate human intelligence, it can't do all of these things, and the interesting thing is that both sides of the argument are true. What I wanted to get into in this video was just to get an introduction to what I think are the six most important things that are going to revolutionize change management with AI and by video. Obviously, if you're watching this on LeanChange TV or YouTube, you're seeing the video version. If you're listening on the podcast, go there. Go to leanchangetv. You can catch the video versions. Sometimes I will share some visuals, but I'll describe them as I go as well. So, and vice versa, if you would rather have just the audio version because I don't really put a lot of production effort into the video part of these just hit record and go. You can grab those anywhere. You can grab them in Spotify, apple Music and whatever your favorite podcast. Thingamajigger is All right. So let's get into six things that I think are really going to revolutionize change management. Now, the hard part with this list of six is there's a lot of overlap between them. Plus context matters. So I've worked as a change agent where it's just been me. So I've either come in for an agile transformation or as an external consultant for a systems implementation or something, and you could argue I'm the only change manager that might be. You know, a 20 person organization to a 3,000 person department in an organization that doesn't have any dedicated change people. So these six things in the context is really what matters, especially when you talk about, you know, can AI take the job of a change manager? Well, in some cases yes, in some cases. No, but don't think about it as taking a job. It's not like companies are going to implement AI and then fire all the change managers and AI is going to take away their jobs. But AI can negate the need to have to hire a change person in certain cases and context. So leave your angry comments down below, if you're watching this on YouTube or just posted on LinkedIn, about how much of an idiot I am and it can't do that. So the six things I was just thinking about these six things as things that I've been tasked to do. So I don't know if this holds true for you as a change agent, but I've had different engagements where you know the tasks and activities seem to change on a weekly basis, so it really depends on the context. I'm taking the six things that are the most important from my perspective, so let's just list them and then I'll just do a little quick overview of each of them and then what you could potentially use AI for these things. So one designing training sessions you can use AI to help you with everything from creating content for that, suggesting exercises, doing pre surveys, post survey analysis, stuff like that. Next is the obvious one, which is data analysis, and that encompasses pretty much everything. It's kind of a catch all, and data analysis will touch all of these six things, but I'll talk a little bit specifically about what are some data analysis analysis things you can do. Communications is another obvious one. This one really makes communications people mad when you say AI can do some of their work for them or all of their work for them. But in some cases it can, in some cases it can't. Let's not tell them. So the fourth one is think of AI as your change project or project management co-pilot. So there's a bunch of stuff that we need to do when we're managing the day in and day out of change work, or managing the budget and the schedule, administrative things. Those types of things. Ai can substantially help with that. The fifth one is better meetings. When I say the term AI, I mean the entire ecosystem of tools and things you can use to make your lives easier. Now, some of those things will directly be AI. Some of those things will be other tools, like automation tools and other things that maybe don't fit into an AI bucket, but there's still things that are kind of useful. So just when I'm tossing the term out, that's what I mean. And then the last one, obviously, is content creation, which, again, much like data analysis, kind of fits into everything. So let's get into designing training sessions. So here's a really cool one. So what I did I think it was last week I was in an organization and they wanted to run a couple of retrospectives and a couple of general training sessions and things and much like every company, we were mixed, so we were remote and we were in person. Maybe the first session it was a 50-50 split. The second one, maybe 70% were in the room and 30% were online. Now, designing training sessions includes everything from doing pre-workshop surveys, post-workshop surveys, suggesting activities, real-time analysis of all the collaboration and exercises that you're doing on the fly, all that type of stuff. So what I did with this one is I already have the sessions designed. And when I say I have the sessions designed, if I'm going into a day-long workshop, I'm prepared to run a five-day workshop. That means I don't typically do out-of-the-box, standard copy-paste training for anything when it's more of a consulting or an internal training thing. So I'm prepared to do a bunch of things. Now, taking that stance is important because it always helps me figure out what's the most important thing for the people, where AI comes into play, as you can analyze their sentiment, their feedback, their questions, what they want to get out of it, given the topic in real time, and this is what I did in Miro, for example. So I was using GPT-4 for this. Gpt-4 is the newest version of OpenAI's model. If you're watching this, two days later after I recorded it, they're probably up to version 5,000 because the industry moves so quick. But a cool thing is I asked people you know tomorrow we're going to do this session. This is the topic what are your most important questions you hope to get an answer to Took a screenshot of that, dumped it into GPT-4 and it analyzed it for me and looked for patterns and told me what was most important for people, told me what the outliers were, and then made some suggestions about you might want to focus on these three topics or learning objectives. Now, granted, those are all things my brain could have done, but it took 30 seconds to do. It. Set a five minute timer. They tossed their questions in, grabbed a screenshot, pasted it, gave it a couple of prompts, boom, it gave me the data. So it freed up my brain to do other things, which was focused on the facilitation, now getting a little bit deeper into designing training sessions again. I don't know if this rings true for you as a change agent, but for me. Sometimes you're thrown into a session or you're assigned to a change project where you might not know a lot about the subject matter in it. I will say this rings true for change managers who have come to my course, who have been thrown into agile projects, because that's typically not the background they typically don't know. I've actually had people come into my course on a Monday and Tuesday and say you know, next week I'm leading an agile transformation. What the hell is agile? I'm like geez, all right. Well, for starters, you shouldn't have taken the job. But second, let's try to get you up to speed as quick as possible. So you can ask AI, give it a topic, give it context, give it what your role is, give it some objectives and say can you generate five ice breaker ideas with all these parameters? It's going to spit some out. You can say I like the fourth one. Can you give me a little bit more information, maybe some instructions? Can you write me a welcome email, these types of things, so it really can act as your co-pilot. I was working in an organization where we were coaching people to be our replacements. So I came in as an external and we were coaching internal either managers or team members who were going to be agile coaches and they were amazed at how much time and effort went into designing these types of sessions. So if we were doing a half day session on whatever topic, odds are we would spend a good chunk of a half to a full day getting it ready and that's everything from, like I said, the pre-surveys and sending out the calendar, invites and planning the activities and all these types of things. So that can be really useful to offload a lot of that brain processing to AI. So the second one is data analysis, and this is really what AI is designed to do. I'm not going to get into the technical bits of how it does this. I might save that for some future videos depending, but I'm not entirely sure most change folks are all that technical in the first place. But data analysis really, you think about talking to AI like it's a person, so don't talk to it like it's a robot. You can ask it open-ended questions. You can ask it. I've been tasked with creating a change record that describes what's changed from last quarter's plan to this quarter's plan. Simple example. You can put that into AI and you can say what data do you need and what format do you need it in. So you can give me a five bullet point summary, blah, blah, blah, and it will spit out an answer for you. But the data analysis can be super helpful with surveys. Now the other thing that the data analysis capabilities will do is I see a lot of surveys where they're binary questions, yes-no answers or rating scales, because those are easy to analyze. You can dump that into Excel and you can get a trend. It doesn't give you any insights. You don't have to do that anymore. You don't have to use any yes-no rating scale questions at all, even if you want to know how much people agree with something. You can have an open text box that you're really focused on the questions and getting people to pour their thoughts into the question as opposed to just going, ah, it doesn't apply, it doesn't apply, yes-no, maybe, whatever. So it doesn't just make analyzing the data easier. It actually changes your approach, where you can get more useful insights and more information. So Gilbert Kridner and I did a state of change management survey. I'll put the links in the description below, so you know, back in 2018, and our goal was to just create something that was by change agents for change agents, because I think we can all agree, a lot of the stuff that you see on LinkedIn is garbage and it's really biased from the perspective of the vendor or people who want you in their training sessions and yes, I've been guilty of that too, so you can leave angry comments below as well, but we wanted to do something to find out what's really useful for change agents, what do they want and what's working for them. That's it. So, um, if you go to leanchangeorg slash AI, you can find that survey and it's wide open. You can add to it and then you can use our custom little AI implementation to ask it questions. So I would ask it questions like analyze the responses for the first two questions and give me a sentiment analysis on it, so you don't have to ask. You know, do you fully agree? Do you slightly agree? Those types of things? You can ask AI to analyze the sentiment for you and look for patterns and stuff, and that's across the board. That's one tiny example. There are so many other things you can use it for If you're in an organization that wants to predict the outcome of a change. I'll just quickly go down this dirt road, because that always gives me a chuckle. You can get some insights from AI if you feed it your historical data. So if you are tracking your success rates, if you're tracking, you know what's working, what's not working, employee turnaround, all these types of things you can dump all that stuff into AI and it can give you a prediction when you ask it. We're going to start on this change. Here's the reason why we're doing this. Here's a few bullet points about what it is, etc. Etc. Now can you tell us what the likelihood of this succeeding are, what problems we might run into, etc. And that's really the magic part. Now, again, people will say, well, you can do that in a room with people. Well, sure, but you can do that in a room with AI, with people. So when I talked about how it's not, in some contexts it may negate the need for a coach or a change agent. It could, because a manager could just open up, you know, any AI tool and do this, but they're not going to get rid of all the change managers, obviously. So you can think of it as something that's going to help you get more insights, to help you make decisions faster. The next one let's move on to communications. This one's fairly obvious and also the most misunderstood. People will say things like Well, everything it generates is really wooden or robotic and it doesn't really sound like a human, and that depends on how you prompt it. Now, if you ask it, we're doing a change. Here's the topic, here's the parameters. Can you generate a communication? Well, yeah, it's going to generate a communication Based on its interpretation of what you're asking. Now, if you start to make it a little more personal and say things like Can you write the communication from the perspective of each of the big five personality traits, or, depending on what your flavor of temperament and personality surveys are, from each of the 16 Myers Briggs types, or each of the four temperaments or each of whatever, and then it will actually customize it, so you can do some things like that. The other thing that you can do is people also misinterpret communications as somebody who's just locked in a room typing emails all day, which obviously isn't the case. So you can do things like have it generate talking head videos from whatever content it's producing, if you're using any kind of internal social media site like maybe you're using SharePoint or you're using, you know, whatever these enterprise tools are that look like Facebook but they're internal. You can generate specific messages for that medium and stuff like that, so you can format for medium. You can analyze feedback so you can ask it for trends and open rates and click rates and stuff like that. So it can be super useful, not just for writing content and stuff, but also analyzing that data. All right, so the next one is a change in project management bot or assistant, and this one's gigantic as well. It can help with so much. And again, I'm only really relating this to my experience as a change person. So back in the mid ish 2000s we were using a three reports from Toyota. So basically every quarter we would get together with whatever managers needed to be involved, depending on what the objective was. We'd create an a three and then we would go execute that and every week we would do a check in. Every month we would do a bigger check in and retrospective, etc. And then at the end of the quarter we'd review what happened, what our objectives were, what our measurements were. Did we hit it? Yes, no, and then we would do the next one. So what I did, just for fun, is I dug out those old a threes and I asked AI do you know what an a three is? And obviously it does. And I'm going to feed you in AI in a big messy Google document. The headings of the document are the titles of the sections. I'll just give it to you. Can you tell me if you can interpret it? Boom put it in there and it said yes, I understand this, that and the other. And I said cool, can you prepare that? I think it was like an a three report is a one page report. I think we all know one page isn't enough to hold all the details for like a 1200 person company transformation initiative. So we had a little summary of one pager, but we also had seven pages of really getting into the details. So when you start looking at your countermeasures and your metrics and stuff, you've got a bunch of data in there. So I said can you summarize this in a few paragraphs for an executive that would care about these things? And it did. I said now the overall objective of this transformation is to do this. And here are some other parameters, given what we did last quarter from the A3 I just gave you, can you create an A3 and suggest what you think we should do for this upcoming quarter? And, honestly, all the details are at leanchangeorg slash AI. If you grab the book because it generated a ton of content I'm not going to have time to go through all of it here and it was actually pretty close to what we did. Now you do still need, obviously, the people in the room to make the decisions, but like I said for the other points is the heavy lifting is done and it's super easy to do. That's just one example. Now think of all the other things you do as a change or project manager, or if you're the only change agent and you're sort of taking on the role of the comms person, the project manager, and you're doing all the change stuff. This is going to save you so much time. I really wish that this existed when I was working on, you know, big agile transformations in the mid and late 2000s and things like that. But you can do tons of different things. You could feed it all of your status reports and have it look for trends. You could feed it if you're in a regulated environment. So I was in a company where I think they had to provide monthly Freedom of Information Act reports. I think you know when you're talking a 6,000 person department, there could be 30, 40 projects going on at a time, project managers producing weekly status reports. So PMO is never going to be able to keep up with this stuff. If they're getting you know 30 to 40 weekly status reports, they have to go through that, roll that up to a program view and communicate that outwards, upwards, whatever it is. You can dump all those things in and say you know, generate me a executive summary for all the projects related to this program and it will try to figure it out and do it for you. So it can save a bunch of time, depending on what those kind of tasks are. And the obvious thing with project and change management things is it'll help, point you to things that can potentially help so you can give it a change challenge. You can say, all right, well, we've been doing this change project for X number of months. Here's the problems we're running into. Do you have an idea about what I could try? It would give you some generic responses and then you could say, oh, can you do that again? But here's a little more information about the objective of the change, so you can talk to it like a human, but that can be super useful. The next one is better meetings, and this is everything from doing real time translations to creating searchable transcripts, to creating things like if you're collecting feedback on your meetings, you can have it analyzed, which ones are good, which ones are not so good and what people think about it and sentiment and stuff like that. Now, this isn't only just AI related, but there's tools like D D. Dash ID is the name of one of the tools that can do some really interesting analysis and transcription creation. Noteyai is another one that plugs into Google Meet and to zoom, and it'll do this in real time and it will also allow you to create analysis, not just searchable transcripts based on text, but interpretation as well. So you can use these things with very little barrier and at the end of the meeting, you can ask people okay, just on a sticky note, if you're in person, what did you like about this? What didn't you like about this? It can take five minutes. You can take a picture of those stickies, dump it into AI, dump in your transcript, and it can give you some insights into, maybe helping you figure out. You know this could have been an email and not a meeting, but there are lots of other logistical things such as, you know, prepping your agenda, creating summaries, creating meeting meeting meet mini there, meeting minute notes easy for me to say. So that's very much similar to what you can do with data analysis, analysis and stuff, but again, you think of it as how much time and energy it's going to save doing these things. That can get you some insights very, very quickly. The last one is content creation, and you'll notice that content creation is kind of sprinkled in all of these six things, because this is the obvious one that people think AI can do. Oh, it can make me an image. Oh, it can write me a blog post. And then, of course, the other side of that argument is well, doesn't sound like me, sounds like a robot, etc. You can train it to talk like you. You can feed it some things that you've written and say can you analyze this for structure, grammar, tone? Now I'm going to give you an objective Can you use that same tone, structure, whatever, to write this piece of content? And it does. It does okay. I've tried this. I've fed it pages of my books and stuff and said you know, can you tell me what my writing style is like? And it spits out some things. And then I say, okay, can you write something using my style? And then I give it a topic and some bullet points and it's okay. You still do have to modify it a little bit yourself, but it does save a lot of heavy lifting. So the things you can do with content creation is I haven't been asked to do this much, maybe a couple of times. I think this point's maybe more relevant in an enterprise setting. So just assume you're working on a giant three year implementation where there's five vendors and 25 teams and all this type of stuff and you're getting flooded with integration documents and all these types of things. You can argue the chain or the story. The project manager would be responsible for those, or the technical product owner or the product owner slash, product manager, slash whatever. But if you're the change agent, you can also like part of your work is changing how those projects are done and executed. So you can bring in excuse me some of those elements. You can take those documents and say I'm going to give you the tech spec for blah. Can you write me a non technical overview that will describe the benefits of it to somebody who can't even turn on a computer. So again, you can talk to it like a human, but it can analyze. Again, you're creating content, it's doing some data analysis. That's why these things are all kind of mixed up together, but it'll save you a whole pile of time. You can do what's changing snippets, and that could be things like you know the software is going live. This afternoon Somebody went oh crap, we need to get a change manager on this and they assign you to the project and you've got a bunch of documents from the product team that described what's changed for this release and release release notes and stuff like that. You can take previous release notes and the new ones and say write me an overview about what's changed, stuff like that. Really, the only limitation of AI is your number one, your open mind to your level of creativity. And three, learning how to prompt it and talk to it like a human. Yeah, like I said, if you don't give it context, think of a conversation you've had with anybody right, or you've. You've posted Something on LinkedIn about I have this problem in this organization and somebody gives you an answer and you go. You're an idiot. That would never work here. What's missing is the context. Ai is the same. It's going to give you a stupid answer if you ask it a stupid question or you don't give it any context. So the same goes with the content production. You can use it. There's some plugins for open AI, for doing mind mapping, for doing Infograph creation all that type of stuff, which is really cool. Images I found as a hit and miss. There's a bazillion plugins out there and a bunch of other Software that helps you create images with AI. So you might have seen some of these on tiktok or Instagram, if you're. If you're there, where you can upload any personal photo of yourself and it will turn it into a professional looking photo. You know the headshot with the crossed arms and all that nonsense, but images just by themselves are a bit tricky. I dabbled with it for the cover of the book. The main problem is AI cannot reproduce the exact same image twice. So it generated a book cover that was perfect and exactly what I wanted, but the text wasn't correct. It was a 3d image that looked like a product shot and I'm like I just need a flat image so I can put this on the book, so can you recreate this exact same thing but take all the text out. And it gave me a totally different image and it still generated a book Because remember, ai, at least if you're using GPT stays in context so it remembers what you were talking about in the thread and it thinks you're talking about the same thing. So the content production part is it's always going to remember what you were talking about. I've done a couple of prompts where I say forget everything I asked you in this conversation. Now generate this. Then it will follow that instruction, but it still does save the context so you can still go back and say remember when I asked you that question a while back about this. Can you regenerate this type of thing? So that's probably a lot of information to digest. Those are the six things I think are going to be the most important for change. I'm going to keep doing a few of these things and probably get into some explainer videos Again. Go to lean change TV if you want to find previous episodes, or subscribe To the YouTube channel, because I always post them on YouTube and I post them in your favorite podcast reader and in YouTube. Drop your questions below or go to lean changeorg slash AI. We've got an open AI survey that's just running perpetually and we're actually going to use AI to run it and analyze the sentiment. It's all going to be open for everybody to take a look at, and what I really want to figure out is what is it that is important for change agents, how can they get the most out of AI and, most importantly, how can they not worry about it because it's you know? Just a very short story. To close off and this is the opening story from the book is I was working as a web designer when the dot-com bubble burst. There were five of us on a web team that maintained a public website and an internet site. I would do some simple applications to support the marketing department. When you look at what technology has done today, there is absolutely no way you would need a team of five to do that now. The technology didn't take our jobs, but it didn't negate the need to have a web team that big for that small of a context. Now you can do the same work the five of us were doing with WordPress and a high school co-op. I'm not angry about that, because it was actually good, because now it forced me to not just be a web designer, and I think it's the same with change managers. You know, if the only thing you can do is write comms, then, sorry, you better get your ass in gear and start learning how to do some other stuff, because you are gonna get left behind. And I'm sorry if that sounded mean, but that's the nature of technology business people want to do more with less. And the leaders in today's companies Are gonna say why do we have an agile coaching team with eight people? Why don't we just use a bot? Get rid of them. You know Google scrum, master layoffs, and you'll find thousands and thousands of people who've been laid off because the people who are managing the company via spreadsheets go oh well, that agile coaching team doesn't build, isn't billable and they don't provide any value, so just turf them. We got to get rid of 20% of our staff, and I don't think this should be a fear thing either. So maybe I said it a little bit harshly, but the point is, you know, much like Developers today are called full stack developers. They're not just one language, one trick ponies, like they used to be. Same thing is gonna happen and change. You know, if the only thing you can do is one or two things, it's time to upskill, and AI can help that. Because I can. Actually, you can ask it questions, you know. Here's the context. Here's the things we do. What types of skills and things should I be looking to add to my repertoire and stuff? So remember to hit like and subscribe to get notified when new episodes come out. And don't forget to subscribe in your favorite podcast reader. And don't forget, go to lean changeorg slash AI. There's a bunch of goodies there, especially the state of change management survey, which is AI driven, plus all of the stuff about what I think is going to help change agents the most. So I'm your host, jason. Little Thanks for checking out season 2, episode 1 of that chain show. Hopefully I'll get back to a regular schedule. Enjoy your day. I.

AI's Impact on Change Management
AI's Impact on Change Management
AI and Content Creation