Session Notes: AI is Not Your Next Step – Get Your House in Order First

Apr 23, 2026
AI is Not Your Next Step – Get Your House in Order First - Transcript
00:00:00
 
 : Clicker. Do you need that? Yes, I will. Thank you. Yes. I already thought it's here. It will work because we tried that this morning. Sorry. Sorry. Everyone, good morning everyone. Now you're all a bit more awake. set the scene for the day and we start with Tim. Another talk about AI. AI is not your next step and as you mentioned yesterday, you're not in farmer or biotech, but we have a lot in common. So, I'm looking forward to your insights. Yeah, it's on hold. We go. Okay. I have a lot of voice, so maybe we don't need to make this. Okay, we'll see if that works. I'm not like everybody else. I have to rely on my notes. Make sure I hit five minutes. Don't get up at 20 minutes. In fact, I'm going to take 26 minutes. Does everyone else just wait?
 
 
00:01:57
 
 : Well, five minutes a question. Yeah. So, I really enjoy talking on the second day. I love when I get the opportunity to do that because then you have the advantage of the first day and you hear all of the discussions and everything that is either an affirmation of what I'm going to talk about or maybe challenges it a little bit. So I'll leave that up to your guys's interpretation. So thank you for those of you that are here on time. Um, so when historians look back at us right now, as of right now, I believe that they're going to see the um birth of a of the unicorn. And I don't mean the billiondoll valuation companies. I mean the more important unicorn, AI. Oh yeah, we got to do both of these. Yeah. So, what's great about the unicorn? It's quite regal yet cute, right? At least our kids believe so. It's got rainbow hair and that super cool horn, right? But that's not the reason that we're all talking about it, right?
 
 
00:03:12
 
 : Why in the world do so many of our leaders believe that this thing is so essential to their survival? It's because it wields healing powers, right? And for some reason, they believe it's going to heal all of their ills. Now, despite the potential for this majestic beast, I'm afraid what most historians are going to look back and remember is that a lot of us just looked at it and thought, "That's a really cool new workhorse." Because what did we do with it? We attached it to our plow, right? And while most of us are spending our time talking about the unicorn, what we should be talking about is that plow. And that's why I'm the bad guy today because who wants to talk about that thing? It's just mired in bureaucracy and confusion and outdated processes, right? It's just with dysfunction and dirt. Nobody wants to talk about it. And that's exactly because what I am hoping to do with each of you is to help me become your company AI especially in a very practical yet meaningful way.
 
 
00:04:57
 
 : Now, of course, when I say you, I'm speaking more the general you because of course I don't understand how all of your tokens operate. But no matter how advanced advanced you are, you need and I'm saying that with such confidence because the entire issue of this file existed long before we were just looking for an excuse, a distraction from all of the dysfunction that we know already exists. Along comes AI and it's this wonderful shiny new toy that averts everybody's gaze away from our dirt. Now I do promise you I am an AI advocate but I accept the burden today of being your party pooper and that's because our profession project management our industries we are going to rely on this becoming a gamecher and we're not going to be able to do that unless we understand the amount of work that it's going to take to get ready for So, what's the issue? Well, as I see it, there's leadership up at the top. No, leadership is not the main issue, okay? They're just part of it.
 
 
00:06:27
 
 : And kind of like I mentioned yesterday, everything's kind of green and rosy and they're making really cool decisions about how this unicorn is going to heal everything. But then below the leadership is this really weird fog. seems to be what right and this is the point right here in fact I think some of us have actually that term or leadership have we done you know what that actually means It means you actually know that there's dysfunction and you're not doing anything about it. And instead, what we're doing is we're relying on brute force and personal heroics typically of our best people just to make things happen. That is not only not scalable, that is not AI. And you know what dysfunction looks like when you apply AI? It's really freaking loud dysfunction. It becomes really obvious really quickly. So just like those farmers, they're using the plow. You and your people, just like mine and my people are working every day not realizing that there is something so much better. And I'm not talking about AI.
 
 
00:08:08
 
 : Better is signing accountability where there is a gap today. Better is clarifying roles and responsibilities where today there's confusion and gap filling. Better is replacing the outdated with the current. Better is eliminating all of the waste, the obstacles. We hear that all the time, but we don't do anything about it. All of that waste that has become so natural, we hardly even notice it anymore. And you see that brick wall back there, back behind the work. Yeah. Unicorns not getting through that. Unicorns can't fly. They're thinking of Pegasus. Unicorns. What's the one thing, the single thing that is required to light the unicorn through that wall and apply all its, you know, magic poofy power? No, no, you don't you don't get the answer that quickly. I'm the party, remember? No, before we do that, we need to talk about our favorite four-letter word, data. Well, it's not just about data. It's also about process, right? is gonna be able to generate that.
 
 
00:09:26
 
 : Okay. Yeah, fine. It's not necessarily just about process. It's also about ways of working. Right now, let me just start by saying right up front, I'm not going to be talking about the convenience of Gen AI, right? That is incredibly helpful. It's essential for us to learn all about AI. What I'm talking about is the need for analytics and insights. very similar to what our friends at in parallel are doing, right? Being able to understand how we're working and actually provide intelligence on that. And for that, we need the data. Lots and lots of really clean, meaningful, correlated, managed, trusted data. And all of you are thinking, yeah, you know, we've been talking about that. Yeah, but do you actually have it? And I'm sure that most of our companies are actually incredibly good at doing this with a lot of different data sets that you generated that you bought and your analytics are probably incredible. But there's one data source data source that's missing from this model and it's one that I actually had a little bit of a discussion online with a gentleman named Paul Budro.
 
 
00:10:50
 
 : I don't know if you know uh Dr. Budro he's a professor um in the US and in Canada and he is doing totally advanced work on AI in project management and he posted about this amazing new level of artificial intelligence that is essentially the next level of like a mon Monte Carlo simulation but the ability to provide tradeoffs for decision- making and my belief is that no matter how complex something gets It's very much grounded in fundamentals. So I started talking to Paul and said, "Hey, what does a company have to do in order to actually enable that level of intelligence?" I want to read you what his response was. To enable this new level of intelligence, organizations need to structure data across the full life cycle. performance project performance, cost and schedule, risk outcomes, not just risks, resource use, financial variance, and most importantly, post project data such as customer value and acceptance. And they would need consistent definitions of success, risk, and value, and how they all connect. The data that Paul is referring to is not just essential for some potential future state.
 
 
00:12:18
 
 : This is what we need right now. And it's the data set that's missing from this. It's the how we do what we do data set. How we do what we do. Do we have this kind of data set? My company's only dreaming about having this data set. To the point right now, it's more of a question than it is a data set. How do we do what we do? My slide didn't. Okay. So how in the world does this lead to better application of AI? Well, everything that we should be doing, we've heard this numerous times just today, you know this already. Everything should be starting with strategy. A strategy is only executed well or poorly based on the organization that we architect to execute it. That organization then executes the defined portfolio of work. That portfolio is a collection of investments, whether it's a project, a program, an initiative. And what I truly believe we don't give enough credit for is that everything that is done within a portfolio is a craft.
 
 
00:13:47
 
 : Whether or not it's the business craft such as those of our leadership and all of our support organizations, the technical craft of the people actually performing the work on the projects or all of us the project craft. And of course, every craft is just an application of human capability to some set of processes. And every single process is nothing more than a sequence of units of work. This that little dot right there, this is where our business is right. Okay? It's none of those up there. There all of these above this, all there are are containers of expectations. This is where the work is done. And the only way that we're going to know how to apply AI is at that level because the only value that AI offers is based on the value of the people using the AI. It happens at that level. And it's only at that level, whether or not we're using AI or not, that we're actually collecting the data for that how we do what we do data set.
 
 
00:15:17
 
 : Now, your leaders might be able to imagine some future state of your organization, but unless we understand exactly how we run our company today, they can't possibly understand what it's going to take to get from where we are today to that tomorrow. They're just going to randomly apply the uniform What it actually requires is mapping it out. And I mean literally every step of every single process. I believe that this is the foundation of everything. What do I mean by everything? It begins with the accountability for every single process. the dependencies between every single step in every process. Forget about your project schedule. Your project schedule is trying to represent everything that is done in a process because you haven't actually laid out those processes. So you're creating it in this fake way in your project schedule. Trust me, I've done it for 30 years. It creates the foundation of value. We during the work do not create our value. We don't determine what's valuable.
 
 
00:16:45
 
 : The recipient of our output determines whether it's of value. It's the foundation of decision making of change. It's the foundation for being able to apply a new technology or a new technique. It's a foundation to understand the time that it takes, the resources that are required. These are all the things we're attempting to do. you would have them if you mapped everything out. It's not only the foundation for one day being able to set expectations, you might even be able to get to the point of achieving predictability. Okay, so wait, Tim, am I hearing you correctly? Project mapping and process data, that's the answer to all of this? Yes. But no, there's more. Because what we have to keep in mind is AI wasn't designed for us. AI was designed to prove that we could do it. Now they're just getting around to asking people, "What do you need?" And what we really need to remember is our imagination. You need to be able to apply.
 
 
00:18:05
 
 : You and your leaders need to apply imagination. It's about the context around each of us. And while the unicorn represents imagination, the plow is our total lack of imagination. So let's hear it from a guy that knows something about this. Jens Jensen Hong think you might know that gentleman. I want to focus on the second part of this. For companies where the leadership is just out of ideas, they have no reason to imagine greater than they are. So even when they have more capability, even more capacity, they don't do more. And one of my favorite gentlemen in the world is Jeremy Utley. He's a professor at Stanford. The limitation is not the technology. It's not your budget. The limitation is your imagination. And to be able to open up your imagination, we need new inputs. This is not just a matter of doing props. We need these unexpected collisions. You need to be able to do what you're doing here. Get around people that look at the world differently than you do, that ask questions differently than you do, and inspire you to ask questions that you've never thought of before.
 
 
00:19:28
 
 : Because with limited imagination, what we think about having AI do is we say, "Hey, um, are we on schedule? Are are we on budget?" That is great. Maybe we don't realize that today. But that's it's great to a degree for a while. But what we really should be having it doing is it explaining to us based on what's happening, are we going to achieve the desired value that we need for this project? So let me give you a really simple uh simple example. Um okay, so I'm going to have a conversation. Hey, so how'd that project go? Oh, it went really well. It came in under budget. Oh. Oh, how's it selling on the market? Oh, it's not. It's failed miserably and we're having to rush our new product out to get it back onto the market. That kind of project, the data is going to show we did well. What we need is something to be able to demonstrate that you should not even be launching it in the first place.
 
 
00:20:38
 
 : And we've actually been talking about that. It's just that I gave it a really cool new name, the value lens. If you want to get scientific, this is just my own name for ontology, but we can get that in to discuss that later. And what the value lens is doing is with all of that super cool project performance data, which tells us what happened. The value lens comes in and not only does it tell us what it means, it's going to tell us why it matters and to whom it matters. So the value lens enables this new perspective of the data and the storytelling with a different perspective centered on the values that we have set to achieve in every project per our strategy. This is not something that just made up. We're literally talking about setting criteria for the eval for an AI to evaluate all of our data. So what does that mean? It means your stakeholders definitions of value for the project beyond on time and on budget. What are their roles in making that happen?
 
 
00:22:02
 
 : To know that you actually have to understand their roles. your business cases, your business unit strategies, your risk tolerance, all of your priorities and your decision rules for when those values begin to conflict with one another. So, we already need all of this, but we don't have it. And we can get away with it with humans because without having all of this right now, what happens to the humans? We just remain remarkably confused and frustrated. Why are we working on this project? Why don't we keep doing this? What is our goal? Okay. But with AI, we don't get away with it because without that information, the AI will remain remarkably confident as it completely misleads us with its insights. So with this value lens in place, once the data actually starts flowing, it begins shifting us away from sorry this is only my term meaningless dashboards, lighting indicator dashboards to leading indicator decision systems. It becomes this proactive consensus artifact. That's good. We're in Switzerland. We want that consensus and its highest power.
 
 
00:23:29
 
 : It's also its greatest difficulty in creating this because we have to do all of our thinking in advance upfront. And all of our leaders and anybody that has to do that thinking is going to say, "But I don't know anything right now." Right? For those of us that sat in that uh round table last night about the unknowns, right? You have to be able to not only establish what the ambiguity is, you have to document it. And as it changes, strange idea, you update that criteria for what you've learned. You reestablish the value that you're seeking. And in that way, nobody is left to guess. Okay. So now what? We got to work the work. The heck does that mean? Well, every expert that I follow in this area, every single one of them is consistent. They say, "Keep people first. Stuck in process, not tools." This is very important. Keep people first. We're talking technology. Keep people first. Start with process, not tools.
 
 
00:24:52
 
 : Challenge me. Every time you guys research this, you're going to find somebody that says exactly this. And you can achieve both with the action I already described, which is map it out. So what exactly does that mean? Sorry, go back one. You can use any standard gap analysis method. Okay, you guys have done this before. You've probably mapped out processes before. You're now doing it for every single one. But you're not doing it. The people that do the work are, right? Because by us enabling them to become the designers of what good looks like. It demonstrates that you trust them because they are the ones that know what better is going to be. Okay? And in doing that, as they lay out each of the steps of the processes that they execute, one, they're going to determine which of those steps should remain human. And they get to determine which steps should be automated or augmented. And by enabling the people that do the work to do this, it's going to lower their defense mechanism against this technology that you're throwing at them.
 
 
00:26:14
 
 : And it would actually open them up not only to learn and adapt, but to be able to get others to learn and adapt as well. That's just the first step. In parallel, you also need to lay out your data standards and your data governance. You guys are already doing that for remarkable other data sets. Apply the same to your business processes. And third, create your value. Okay? If you're wondering how to do that, if you're wondering I how in the world do I convince my leadership to actually do that kind of thinking and capture that kind of criteria? I'll sum it up with one quote. We don't actually know who said it. It's one of my favorites. Choose your hard. If you don't, your hard will choose you. Okay? What we're all doing today is hard. And the problem is is that our hard is just a l of hard and it just keeps being the exact same thing over and over. If you do the hard that I'm talking about, it's really hard up front and it's constantly getting better and easier as you go through it.
 
 
00:27:37
 
 : Don't take it on all at once. Prototype, right? Just like any other design, prototype it. And if you're very curious about this, I have some handouts on exactly how to create that value lens. And before you jump on me, uh, all I want to be able to say is I believe that when historians do eventually look back at us, they're hopefully not going to write about the companies. maybe some of these your guys' company that push their teams to adopt AI the fastest or first. What they'll do is they'll write about the companies that did it intelligently. They learned about themselves first and then figured out how to apply it precisely where it absolutely could be the most meaningful to the way they work. Thank you very much. Uh, two quick questions from the audience. First of all, thank you. Great story. um you know uh great insights with it as well. Just want me back that picture 

... [transcript truncated]