AI-Powered Knowledge Management in Architecture Firms

Welcome to the Troxell Podcast. I'm Evan Troxell and before we dive in, I want to acknowledge Chris Parsons and thank him for his partnership throughout this series. This marks the seventh and final installment of the KM 3.0 journey that we've been on, a project we created together as a lasting resource for the AEC community, something that captures where knowledge management has been, why it matters more than ever, and how leaders can respond to it's been incredibly rewarding to build this series and watch it resonate with so many of you working to move your organizations forward. In this episode, Chris and I talk with Todd Henderson. Todd is a thinker, advisor and practice leader who has spent his career helping organizations navigate complexity, build resilience, and rethink what it means to operate in a world that rarely behaves the way we expect. His work crosses strategy, leadership, technology, and human behavior, and he brings a clarity and honesty that's especially needed right now. Our conversation spans a wide range of issues how AI is reshaping knowledge work and expert driven fields like architecture and engineering, why VUCA volatility, uncertainty, complexity and ambiguity is now the constant backdrop for professional practice, and what leaders must understand about the fragility of institutional knowledge. We get into the rising cognitive load inside firms, the limits of traditional knowledge management tools, and the shifting identity of the expert. When machines can retrieve information instantly, Todd brings a deeply human lens to all of it the role of language and narrative and leadership, the tension between identity and adaption, the psychology of AI adoption, and what it takes to build organizations that learn rather than simply react. It's the perfect conversation to close out the KM 3.0 series because it elevates the discussion from tools and tactics to the broader mindset and operating system firms will need going forward. As usual, there's an extensive amount of additional information in the show notes. You can find them directly in your podcast app. If you're a paid supporter of Troxell and if you're a free member, you can find them over at the website, which is trxl co. Please help the podcast by sharing episodes with your colleagues, and I hope you'll stay through the end of the episode to hear Chris's and my closing thoughts on the series. So now, without further ado, I bring you my conversation with Todd Henderson and Christopher Parsons. Todd Henderson, welcome to the podcast. Great to meet you.

Likewise, Evan. Thanks for having me.

All right, I'm going to pass this straight off to Mr. Christopher Parsons, who has a relationship with Todd. They've been working together for quite a long time. And I'm actually really excited to hear about Infinite Todd at some point during this episode. So take it away, Chris.

Oh, yeah, good. Good foreshadowing. I've known Todd for close to a decade now at this point, and I have come to think of him as one of the most generous and inventive knowledge leaders in our community. So he's spoken at K Connect twice, once several years ago, about lessons learned with one of his coworkers. And this past August, he spoke at our conference and he gave a talk that has already become one of the most discussed sessions of the year. It was called Rethinking KM teams when your newest teammate is AI. So we'll get into that today. We're going to learn about. Talk about what it means to build a learning and knowledge program when AI isn't just a tool, but it's a collaborator on your team. So Todd leads a bunch of things. You'll hear about his plates that he's spinning, but learning and development, knowledge management and AI at Boulder, he champions their synthesis platform. There he's a director of practice improvement and his intranet they call Biscuitnet. We may or may not get into why it's called Biscuitnet. That's their intranet platform. So Todd is a huge continuous improvement person that will come up for sure. And he's one of those people that's always experimenting. So he's curious, he's optimistic, he's quick to turn ideas into action. He once described himself as the person that puts the fork in the light socket just to see what happens, which I think that's a huge part that actually describes a lot of people that I hang out with. So it's a fun group to be with. So over the years, he's just been a great contributor to our community. And I think this talk, you'll see that he just brings playfulness and fun to knowledge management and business. And I've always appreciated that. But Todd. So with that, Todd, let's dig in and welcome Nick KM 3.0.

Thank you so much. I read something recently, Chris, about how, I don't know, in the context of what we do, it's like we're in a jar, right? And you can't read the label on your jar. And so it's on your own jar. On your own jar, you can sort of read other people's jars. But. So that introduction was very heartwarming for me because it's like, I know what I think is in the jar. It's nice to Hear what other people see.

Todd's personal brand is curious but dangerous. That's what it sounds like to me.

Yeah, it's like that Christopher Walken sketch, the Saturday Night Live one with the. Where he's with all the plants, you know, and he puts the googly, googly eyes on all the plants because you don't turn your back on a fern, you know, that way he can make.

Eye contact with a fern and make Smurf staring contest.

Exactly. Cool.

Hard transition to AI Search. No, I'm kidding. That's a natural segue.

Yeah. Yeah.

No. So one of the things Todd, I usually bring in early into beta when we're building new stuff, because Todd can see past what's broken and can see what's possible, which is very helpful when you want to collaborate and build something new. And so, Todd, you were one of the first three firms into AI Search. When we introduced it to synthesis, you were right on the edge of what it could do. So we appreciate that. And I'm kind of curious for you what that was like, what drew you to wanting to be part of that beta and that process and kind of like what you learned kind of early on and how you started. You were one of the first people in our community to evaluate. Is this working? Is it doing what I would hope it would do? What even do I hope it should do? Like, take us through that. What you remember of that? That was, like, last October is when we started.

Oh, gosh. Well, okay. A thing about me is, you know, my favorite dish at any restaurant is always whatever I haven't tried. Right. I live in novelty and curiosity and. And what's new and what's out there. And so I just. That's just how I meet the world Is. Is by. Oh, there's a button I haven't pushed. What happens if I push it? Right? What's that alarm, you know, sound now? So that. That kind of thing is right up my alley. Just the. An ambiguous situation or a new situation, unclear situation. I kind of love. That's where I love to live. The dark side of that is I get bored really fast. Like, really fast. Like, I went skydiving once, and I was bored on the way down. Yeah. I was like, okay, well, this is cool. I can turn left. I can turn right.

Got probably less bored right at the end, though, Right? Like, that's. That would be interesting to know how that turns out. Yeah.

So I love. I think I might be part guinea pig, you know, genetically or something like that. I. I love that kind of Testing stuff out and then see if I can break it, see if it's broken, whatever. So that is context. I mean, what, what was fascinating for.

Me.

Is, you know, I always say every, every tool works at both ends. And being able to search biscuit. Net, our intranet in this other way, this much more, you don't have to know where stuff is. You just have to write a question. Was this big, just a, like, oh, wow. And then kind of. Right. Follow behind it like a, oh, no, we really should have been investing in our data because now it's going to pay off. Or it would have if we had, you know, been doing this all along. And so that's Chris, that's what comes back to me first, was that like, oh, I get it now I have.

A question about this. So AI combined with search combined with Internet. Right. But I want to know where were you in your kind of tool curiosity when Chris came to you in regards to AI in a more practice based or just even consumer level?

It's been going so fast.

I know it's been a year and it's been a lifetime.

Last October was a decade ago. I was using large language models. ChatGPT. I think at that point in time I might have been a Perplexity fan. I forget. And using it a fair amount. Sort of, sort of dabbling in it, kind of, you know, poking it like, what, what does this do? You know, I'd gotten past the stage of, you know, write this in a funny style. Okay, that's okay, cool. You know, we, we, we wrote some new Shakespeare and we're done with that. And then was sort of, sort of like, well, how does this become useful? And of course at the same time, those tools are sort of rapidly evolving and gaining their abilities. But I think that's kind of where I was, was, this is important and happening and I need to know what it is. And it's cool as heck. And also trying to find like, okay, where does it, where does it go? What's it for?

Tracks. Because pretty early on we asked you to, I mean, again, thank you. Like, we asked you to do a webinar like six weeks after you started using the product to show people like what you were learning. And you showed a really cool example of kind of critical knowledge transfer from a couple subject matter experts. One of them was Darcy Hernandez and other was Emily Nightingale, where you stitched together multiple different kind of AI workflows to do something super cool. So I wonder if you can kind of tell us about those early projects.

Yeah, all of that was coming up simultaneously along with AI search and so forth. I think what sparked this for me was all of a sudden we have tools that transcribe recorded speech and they do a pretty good job. Knowledge architecture built that into synthesis. So any of the videos that we post there have a transcript and you can like, huh, that's cool. And, and teams does this and that. That technology became available and I don't know, somewhere the idea came to me like I could take that transcript and then use that to write intranet content. And actually I could have my friend chatgpt write, write the Internet content and I can kind of, you know, be the director of that. And so out of that came this pipeline that I've sort of iterated over time of get expert talking, record it, transcribe it. Those basically are effortless at this point. And then go and run with it and use AI to mine it. Right. Whether it's analyzing it for patterns, whether it's crafting a summary or content, end use content from it, whatever. This is, this pattern that I just do all the time now. And that was right around when it was born. And so the two examples, Chris, that you mentioned are two different experts in our firm. One's my partner Darcy, who leads our, one of our offices, Southern California. And she actually, actually I think this is where this really kind of, kind of landed for me as like the, aha. She was lamenting like, ah, I have something to share. And it was about the politics of healthcare construction in California. Very niche thing, but, but kind of a deep topic. And she's like, I know I should share this on biscuitnet. I just, I, I'm never going to find the time. I just don't have the time. Yeah, sure, yeah. And I think in that moment I was like, I have an idea. How about you and I have a conversation and I'm going to send you a draft. Right? All I need to do, Darcy, is get you for a half an hour because it's the writing is the part that sucks. If you're an expert in X, as long as X is not equal to writing, if you're an expert in healthcare architecture, probably you're not great at writing. Experts at writing are great at writing. You know who else is great at writing large language models?

It might even be hard for an expert in writing to write about writing, though.

That's probably true. Yeah. Yeah. And that hypothesis that, hey, if I can get the writing out of the equation for the expert, and really all I need from the expert is Talk to me. Which is a thing they tend to enjoy doing.

Yes. I was gonna pipe in with that.

We love to talk about what we know. Yeah, we're good at it because we enjoy it.

Yeah.

So it's just a source of pleasure. Right. It's the writing is the painful part. So that's. So that's. That's a thing I've. Now that. That little. That sort of, you know, pipeline is something I do. Just kind of standard operating procedure, you know, the. The. The knowledge that. That all of us, you know, experts hold. When it's at rest, it's unreachable. If you can get it out in the air and like, literally in the air in the form of wave vibrations that a thing like a microphone can pick up. Yeah. Then you can do amazing things with it.

Well, they hate writing. But I think the other thing, Todd, that I've seen you do is they also don't know what they know. And so you've been pretty good at asking them kind of big, simple questions or following up or like, wait a minute, hold on there. What do you mean by that? And clarifying and digging for examples. And I think what I've seen you produce is a thing that they probably maybe not have been able to write on their own because they wouldn't have access, you know, just a top of memory to some of those things. Like, they would be much more buried. Right.

I think that's it. And, you know, talking about, you know, the run up to this meeting with Darcy where I wound up interviewing her, I even, you know, we sort of had a, you know, like, well, what. What are the points you want to cover? And. And I said, okay, well, cool. I'm going to just facilitate this. And you just talk. She's like, oh, you mean I don't have to worry about kind of getting us through this to the end? I'm like, nope, I got this. And it is. Let her just drop into her expertise and not really worry about it. And. And so. And then. Yeah. And then I can. I know enough that I can poke and prod and dig and.

And then you can restructure so she can talk about it out of order and kind of meander off to a piece. And then you have the ability to kind of.

Yeah.

Synthesize that that's not too on the nose. And then. And then put it back in the right order. Right. Like. And then I think the part that you said. And then share a draft. So then it becomes back and forth. And she's like, actually, I Missed these three things. Shoot.

You know, Abs. Yeah, she doesn't need to tell or any, you know, they don't need to tell a coherent story from start to finish. And we're just going to write it down for them. It's like just pop the cap and let it fizz out. As long as we catch it, we got it right. Those transcripts become this precious gold mine. And it doesn't have to be pretty, just get it, just capture it.

Can I ask about the output from that? And obviously you had this back and forth then with the draft. And so then it goes into the Internet and it becomes a source of this information that people can look up. I'm just curious how it connects to the end of that whole workflow. The people who actually need the information, do they read it? What did they think about it? I mean, I'm just curious from like an actual like rubber meets the road. Practice improvement. Right? That's what you're doing there. So. Because you can. How many times have we all done this? Like Chris has been in it, I've been in digital practice. The stuff that you're doing, it's like you put it out there and then you don't hear a thing about it ever again, right? Never.

That never happens. I'll pivot to a different example that maybe illustrates a kind of a more a richer sort of end use case. And that is the other one that Chris mentioned, Emily. Emily's one of our just rock stars, just generally project manager. And she is really good with Bluebeam. Bluebeam, if folks don't know, is this PDF, it's kind of, it kind of for architects replaced Adobe Acrobat as a thing that reads and edits PDFs and a lot of us spend a lot of time in Bluebeam. That's where we mark up drawings these days since we don't tend to print them out that much anymore. Anyway, it's a complex program. It's got a billion settings and same thing. Emily was like, she's known to be this resource and I actually, in this case I approached her and I said, hey, Emily, what do you think about doing a quick masterclass on Bluebeam for project managers? Right? And in this case we invited some other folks. So she had an audience that she spoke to and they could ask questions and naturally recorded it, transcribed it and so fast forward, right. On Biscuitnet, there's now a page, Bluebeam for project managers. Actually, I think I kind of rephrase it because you don't have to be a project Manager to, you know, it's like Bluebeam for like busy people. And you can watch the video. You can watch the video. Synthesis automatically transcribes the video. So Synthesis Search doesn't need anything else from me in order to. If someone says, ah, what is the difference between a Bluebeam studio project and a Bluebeam studio session, which sound like the same thing but actually are different. Synthesis gives a lovely answer to that. But on the page where the. There's the video, I still have a written summary that, you know, I use chat GPT to develop because a lot of people would rather skim. Right. They want to kind of go right to the thing that they want. So, so with, I mean. And that took 10 minutes to create. I mean, it just, it's so fast. Right, Right.

We.

One of the plates that keeps spinning is our onboarding program. It redesigned our onboarding program. Well, that page, that content is a part of onboarding. And so new hires, unless they're entry level, they get a kind of a very basic intro to Bluebeam from that blue, that Bluebeam themselves produce. But if you're, you know, have some experience we have, you go do this and you can either watch the video, you can kind of read it, and off you go. And so that's a way of trying to operationalize that, that knowledge and meet.

People in different, you know, I don't think, I tend to say learning styles, but like, you know, different preferences or if you're in a hurry, you don't want to watch a video. You want to get right to the shortcut key.

You want the answer.

You want the answer.

Yeah, exactly. Yeah, exactly.

So you took this. So timeline wise, we're kind of in like Q4 of last year, October, November, December kind of thing. And then we started talking the beginning of 25. And I'll just, I'll skip some of the kind of the back. The history there. But like, you ended up with this idea of creating a medical planning accelerator.

Yeah.

Which I just want to know, and I want our audience to hear more about. What is a medical planning accelerator? Why did Boulder Associates need one? Should we all have one? You know, what is this?

This was one of those, this was one of those moments where just a need came up. I'm like, holy cow. This is what I've been training for. Right. And I didn't even know it.

You didn't know it ahead of time.

Didn't even know it. So another partner of mine, let's call her Kate because that's her name, was lamenting that and so Kate, Kate, like Darcy, is a rock star expert. Her expertise is medical planning, especially the early phase of medical programming and user group meetings and really complex medical planning. Incredible expert. And she has built a cadre of medical planners around her, a growing team around her. Word came to me that she was lamenting like, ah, it takes, it takes too long to get people kind of up to speed. Right. So we have a few people whose job is actually medical planner, but we have a whole lot, I should mention our company, 80, 90% of our work is healthcare work. And so we have a lot of folks that you know, do typical projects and, and may not even use our actual medical planning and like 200 or.

So employees, something in that order of magnitude.

250. Yeah, exactly.

And then how many medical planners of those people?

Six?

Okay.

Something like that. So they're a, they're a kind of a precious commodity. And this is, this is a relatively new thing. I mean this is just, this all was sort of born like last year. So this is, you know, new and growing and evolving. Anyway, the problem, we need to get the sort of the most project teams working on healthcare projects up to, up to speed, faster. We're hiring a lot of folks who are maybe experienced architects but are new to healthcare and there's a lot to learn there. Right. We also, there was the observation that when teams do engage with our medical planners, they don't do so great. Kate talked about getting a phone call. Hey, I have a user group meeting tomorrow. I'm hoping you can run it. Well, that's not great. That is not a best practice. So how do we just raise our game on some of the basic stuff, but also how to use these experts that we're growing and developing and also how do we take those medical planners and actually keep elevating their work as well? So that's sort of the rundown. It struck me as the most perfect knowledge management. Like what a, what a juicy KM challenge. This is a poster child for a wonderful knowledge management challenge. Right. So I did some research and you'll see the threads of what we just talked about coming through where I just went and I interviewed all of medical planners and to do that, you know, I did a structured interview. I kind of interviewed them the same things. And I'm not a skilled interviewer. So first of all, I had chat GPT. Help me. Like, hey, help me, help me plan an interview. Here are my goals. I want to surface the lived experience, if you will, of our medical planners. I want to understand what Their world is like how they got to where they are, where are they feeling friction, blah, blah, blah, right? And it helped me figure out, you know, like a half an hour kind of a, you know, just an interview plan, a series of questions. So I started interviewing these folks. Then I feed the, I feed the first transcript back in and I say, hey, I did one of these interviews. How did I, how did I do as an interviewer? First of all, give me some coaching because, you know, in all of this I want to be, I'm not trying to hand off my work. I actually want to get better at all of this stuff. And are the questions working? Do we need to tune anything? And so we're able to, you know, just rapidly iterate. I mean, I had, sometimes I had a half an hour between interviews and I could revise the interview plan, you know, easily within that time, get the interviews done, feed them all back in. Here are six or eight interviews, whatever it was. And now it's a task of mining that gold for the, the nuggets, right? What are the patterns? What are we seeing? What's, what's missing, what's, you know, and it was just this kind of just fantastic process. It was so easy because I'm not trying to, if I had to do this manually and sort of read through these transcripts or re. Listen to them or read my notes. A human brain doesn't have the, the bandwidth to do that easily. AI, not a problem.

I think what's interesting about that process, I mean, because you've started out, I think Chris mentioned it in the introduction, which was when AI is your newest teammate, right? This was part of your talk at KA Connect. So you really, I mean, that's becoming very clear that that's how you're treating it. And I think what's interesting kind of through hearing you and through my experiences, and I would love to hear from both of you, your experiences in this. But it's like somebody is there and they're always available and it's the person that knows about the thing, right? So it's like when you're interviewing Darcy or Kate or, you know, it's like you just have access to them then, but they only have so many hours in the day and they're working on all these other things and you actually don't have access to them. And so I think that's what's super interesting about this. And then multiply that across multiple offices and people having access to this information in a very like, I get to ask the question kind of a way. Not just I get to read the report that was generated from through writing or through an interview or a video or whatever. Right. But I think that that's. That's like a super interesting shift in the way that we're interacting with technology as it represents what people in your firm are authoring. And so I'm curious from both of your, you know, speaking about AI as a teammate or a collaborator, does that kind of. Is this the experiences that you're having? Because I know there's a lot of people who just still use it, like, oh, answer my question. And I think you're really speaking to this kind of back and forth between you and IT and bouncing ideas, getting feedback, all of those kinds of things.

I think Todd's point earlier, it's like, oh, yeah. And then, oh, no. I think another way of kind of saying that is, you know, we use these tools internally. Like, we use our own technology. And I have written more content on our internal intranet. I've probably written five to six times as much content this year as I wrote in the previous year. You know, like, probably higher than that. Because what you realize is if you want that to be possible, if you want people to be able to have conversations with, you got to get it.

Out of your head.

You got to. Right. But you have to treat. This is the thing. If, like, if the AI is a teammate, you have to think about its learning and development as well. Right. So how is it possibly going to answer that question? And it's got all these knowledge gaps, and part of what you're now trying to do is, like, you're trying to externalize what you know, and not just what I know, but, like, what our team knows or, like, what nobody knows. Like, you could. There's a gap here. It's like, nobody here can answer this question. We need to go find the answer to plug the gap. Because you're trying to build, if you imagine it's like a foundation, like, you're trying to build a solid knowledge foundation for your company. And you don't want there to be cracks and holes and gaps. You want to build on rock, not sand. All the metaphors, right? And so when you get that metaphor in your head of, like, we're trying to build and then also maintain this solid knowledge foundation, and you're thinking about these future teammates that may be AI, it just changes your whole approach. Like, as I'm designing stuff, like new features, a lot of what I'm doing is then I do a lot of that through audio Mode, which I know, Todd, you do as well. And at the end of that I'll say, like, let's kind of tell me our design decisions that we made while we were designing. Like, what did we decide to do, what not why did we make these decisions, what are some workarounds? And then I'll log that because I'm not going to remember that three months from now or six months from now or 12 months from now, so I can personally go back and get answers to those things. But that opens it up to my team too. So they're like, why aren't we supporting this feature? Well, it's because of this, this and this and we have to do in the future. So I think it's a, it's a mindset shift, Evan, and an awareness that like, if you take the time to write all this down, and I think this is what's different from the past of knowledge management is everybody, especially subject matter experts, get it. They're like, oh, I see, I can retrieve it, my team can retrieve it. You know, it's setting us up for AI success. So it just feels like that ROI on doing all that knowledge work has just gone through the roof.

That retrieval part is so important, right? Because how many times like, okay, it could be a Revit model, it could be a Rhino model, it could be a Bluebeam document, but you have to go in and look for that information if you want to find it, or it's just forgotten about, right? Let's be honest, how many files are on the server and it's like nobody knows what's in them and no one's going to go look and somebody's moved on and we forgot about it. But what you're actually talking about is just kind of this actively live document. At least it has the potential to be that.

I think it's kind of like a digital twin for your people, right? If you think about it, you want to keep the model of your knowledge up to date, you know, digitally, while you're building it in your carbon based analog brain or whatever it is, you know. And Todd, I think like partly what's going on at Boulder, right, is like you're kind of a key thing about knowledge management. You can't manage all your knowledge. If you try and do it all at once, you're going to just fall apart, like there's too much. So. But what you've done is you prioritized a key area around medical planning that's like, let's establish at least a foundational level knowledge in the company around medical planning.

Right.

And then we'll move on to the next thing and the next thing and the next thing, right.

It's the classic advice of pick. Pick a problem that matters to do something that somebody cares about that improves your business. Don't just do something neat for fun. You have to, you know, it has to be targeted. And so here we had a, you know, with the medical planning thing, we have a clear and present need for something. Right.

Which all lived in people's heads, right. There was nothing written before or was there any. Anything, right?

Very little. Yeah, very little. You know, and, but, but that I think a key thing is that that first step of, okay, I hear what you've asked me to do, I'm going to go study the situation and get a clear idea of kind of the landscape first. Was critical because that.

Todd, let me. I'm sorry to interject, but like, no, there's. There's something in, in the nuance of the way you did it that I think is important. What you didn't do first is sit down with Kate and try and exhaustively interview Kate. You went to the people that worked that were more junior or kind of like mid career, right. You did the. You call it going to Gemba, which I think would be like, you tell us what that means and kind of like why you did it that way. I think the sequence was important.

It was important and very deliberate. In fact, I didn't interview Kate on purpose. I have access. Kate's sort of the, the customer for this, in a way. But the work is going to impact mostly these other medical planners, so I need to understand their world. So I've been shaped immensely by lean, right? Lean is this sort of philosophy that comes out of the Toyota manufacturing system. It's come into healthcare architecture, especially starting in the Bay Area. So it was really, really important in my kind of middle of my career. And it just shaped my thinking about all kinds of things. When Chris introduced me saying he's Todd's all about continuous improvement, boy howdy. I mean, the way I make coffee in the morning has been iterated 10,000 times. I'm always like, get a little bit. But anyway, I got excited about coffee and I lost the train of thought. So Gemba, there's this notion. So there's a lot of Japanese terms that, that, that show up in my mind and in my world. And, and one is this notion of genchi, genbutsu. It means going to Gemba. Gemba is like the place where the work happens at Toyota. It's expressed by an executive coming out onto the shop floor. They've heard there's a problem with one of the machines and the executive is there and talking to the person who uses the machine and understanding the reality of this and they might. There's this example of, of one of them, like reaching into this pan of oil and pulling up, you know, you know, filings of metal, like something's really going wrong with this machine. And it was, it's this expression of you really do need to go see what's going on. And in so doing, partly it's, it's just that, understand what's going on. But you are showing respect and that is, that is just so important, right? You're showing respect to this person, who you are perhaps to, you know, change their world in some fashion, designing a process or a standard or who knows, right. I don't think you can ever go wrong by showing too much respect. Just, it's kind of like communication. Can there be too much communication? I don't know. So that notion of, hi, I'm Todd, tell me what your world is like, tell me what's going on for you, right? Helps me understand what's going on, but also helps build relationship. It helps them be interested in something. The alternative, and I have tried this is, hi, I'm Todd. You're doing your job wrong. Guess what, it's your lucky day because I fixed it.

I have a template for you to try.

Here you go.

I have a standard.

I have a standard for you. You're welcome.

Yes, here you go. I'll stand over here if you want to run out and get me some flowers or candy, you know, because no, that does not work.

So you're both trying to understand their lived experience, but also their hopes, their dreams, their needs. And so when you end up shaping knowledge products for them, you know, it's not you and Kate dreaming up what would help them because Kate's so far down her journey, she doesn't know. And you don't know anything about medical planning, so you don't know. So your design is customer led knowledge product design.

It's customer led. And, and Chris, it's, it's influenced by getting to be a beta tester for the things you guys roll out. The, the, the sort of discipline of I don't go very far without making sure the people who are actually going to use this have, have given some input, right? Have helped shape it.

What's wild about that, Todd, is like we start really early with that process. You're talking about. And even so like, like, let's say I, there's 7 of people and I kind of spend a bunch of time with them and we figure exactly what they need. We go design it, we build it, we give it to them. You have to stay with them because the what everybody hoped and dreamed isn't once you get into 3D and you start using something, it's not like you thought it might be and you have to stay with that group like two, all the way through, you know what I mean?

And it helps when you've built bridges and they've already see you as an ally, not an enemy or someone who's going to change their world, you know, make change at them, but. But is trying to make change with them for the things that they've told you they want. That's a much better, you know, talk about building a foundation.

So you, you, you went to Gemba, you spent time then what did you do? Like, or what did you learn?

Yeah, yeah. Well, you know, take these transcripts back and kind of just work back and forth with, you know, in this case of chat, GPT of like what, what's here? What are the patterns? What are we seeing? And you know, a couple things popped up. One, ah, some of our medical planners aren't clear what the medical planning role means. Well, that's a problem.

Welcome to architecture.

Yeah, well, and it's one of those moments. It's one of those moments where it.

Means a lot of things.

A problem well stated is half solved. It's like, okay, I can knowing that. Guess what? You know, half an hour later we have a pretty good medical plan at roll description figured out now, right? I mean that was a side, a side gig.

Just to pause on that though. Is that synthesized from these people who are not Kate, or is that here's what people think it is then you like Kate, this is what people think they're supposed to be doing. How close are they?

So just to sort of do a little side quest on the medical planner role description, those interviews with the planners helped surface that we need one. We have the lack of clarity. Also our project teams, remember, our project teams aren't using medical planners well, they don't know where they start and stop. We all need this clarity. So for that, I actually did interview Kate and one other senior medical planner. I also turns out we have a job description. So we have roles, but we also have jobs. So we do have people who are hired as a medical planner. So I was like, well, give me that job description. That's probably a good feedstock for this. And I took all of that, all of the interviews and the job descriptions, and then I was like, help me write a medical plan or job description. Role description.

So the difference is a role is like, I might be playing this role on a project, but I'm. My job description is project. Job captain or project architecture, whatever it is.

Exactly. Okay. Exactly.

I love this whole idea of this thing that you didn't just stop at. Well, we fed the system the foundational information, and now we can just live off of that. Because I think a lot of. I mean, that's. That's what happens, right? It's like, okay, check the box. Yep. Uploaded the HR docs, uploaded the employee manual, uploaded the job descriptions, the roles, the responsibilities, the definitions, the glossary, the acronyms, all the things. Okay. But I think what's super interesting is that you actually asked a question. Then it's like, what? Well, I mean, it's your job, right, Todd? How do we improve on this?

Right.

How do we make it more clear? How do we get clear? Because you could interview two different medical planners, and they could have two different descriptions about what the role of a medical planner is. I mean, there would be a lot of overlap, of course, but there's still gonna be outliers. There's still gonna be different ways to talk about things. There's different ways to abstract information and communicate that to people. And then you get the. To use these kinds of tools to really help bring clarity to everybody in the office. And then it just becomes this thing that I. I don't know what the cadence is, but you. You can't. You kind of need to do this, and I think over and over, and I didn't finish the sentence, but. But in architectural practice, how many things are there in the practice that it's like, well, that's done, and now we never have to think about it again. We're just going to keep doing it the way we've always done it. That's like the dirty word on this podcast, right? Is the. We're going to do it the way we've always done it. Because that doesn't. Like, you're. You are literally in a role of finding better ways to do the things we already do.

Yeah, yeah, yeah. And you know what the. The other thing is? I believe in the 80, 20 rule almost as a. As a. As a lifestyle. So that role description, it only has to be 80%. Good.

Fine.

Because we can iterate. Yeah. Plan, do, check, adjust. Plan, do, check, adjust. I remember.

Right. There was a big thing, though. There was like, a question. It's like, yes, are medical planners responsible for selecting medical equipment? Yes or no? And nobody knew. Like, that's a big sometimes piece of the role. Sometimes.

It depends.

Depends.

It depends. That's an answer.

Well, the medical planners were clear. We don't do that. We're not medical equipment planners, which is kind of its own thing, very technical. Our project teams really want to hand this off to somebody. So they were strategically unclear themselves. And. And yeah. So, you know, that surfaced in these various interviews as a. As a concern, as a lack of clarity, as a thing to get like, we should figure this out. And. And then. And. And kind of, you know, it came up a few minutes ago. It's nice, Chris, you talked about this. You sort of throw everything in the hopper and then you can kind of get it back out again knowing that there's this issue around. Hey, we need to draw this boundary. Our medical planners are not your equipment planner. But how do we do that? Tactfully, I was able, when using AI to write this role description. That piece showed up clearly. And then using it to write a revised page on biscuitnet, same thing. And it was able to phrase it more tactfully than I was. Than I was able to, because it's good with language. You know, there's another thing.

Good.

There's another thing I want to about this, that the whole kind of. From these interviews, that was one of those, like, blew my mind moments where every now and then, I mean, most of what. You know, most of these conversations with ChatGPT, it's just like, yeah, banal, banal, banal. And then you're like, oh, hold on. And at one moment, it used this phrase, the Kate. Effectively. He said, oh, just blithely. It's like, oh, well, you know, you need to worry about the cat effect. And I was like, what? No one. No one I interviewed said that phrase. We don't. That's not a term that we use. What do you mean by the cat effect? And it came back with this really poignant thing of everyone who is. You know, all these medical planners, they talk about the incredible impact that Kate makes when she joins a meeting, when she comes and intervenes and does something. She gets a lot done fast. They look up to her. They want to be her. And there clearly isn't sort of a second in charge. There's Kate and then there's everyone else. And that is a source of systemic fragility. That is a vulnerability. Oh, My God. Yeah, that is. That's gold right there. Yeah, I'm. It's doubtful I was going to come to that on my own.

That is insightful for sure.

Yeah. The kind of risk profile of that.

Oh, man. And then it's like, now I can work on that.

Closing that gap.

I can work on closing that gap. Yeah, yeah. And the things we're, we're still working on is strategically like, okay, I need to transfer some critical knowledge. I think the book is sitting right there from Kate to these other folks and broaden it out. And the other thing about this also is I can show this. I wrote a report from those original six interviews of like, here's the picture that I'm seeing of medical planning and gave it to Kate and she's like, she can be like, aha, I see it, I get it.

But it's citing the Pete, those people's quote. You quoted those people. Right.

Literally. So like here there was absolutely.

This is in their words how they feel, right?

Yeah.

I don't know what their job is or.

This isn't Todd trying to, you know, whatever. Do you know, it isn't Todd's opinion. This is Todd as a sort of amateur anthropologist studying a situation and reporting maybe journalist, maybe is better, better metaphor and reporting it back. And that has, that has also been very powerful because what, what I'm really trying to do, and I think it's, it's bigger than knowledge management. I'm trying to understand the structure of problems because it's a problem presents as some kind of behavior that is happening and you wish it wasn't, or behavior that isn't happening and you wish it was. And the classic, the classic management move is go tell those people to stop doing the thing or start doing the thing. Yeah.

Nobody wants to hear either one of those.

It never works. It has never worked in the history of your managers.

Symptom, right?

Yeah, it has never worked because the symptom is the result is the. Is the bottom of this long chain that that includes the beliefs that your organization holds and your structures. What jobs are there, what information flows are there? And if you don't try to understand the structure of the problem, you're not going to change the output and just yelling at the person that is or isn't doing the thing.

Yeah. I mean, you're literally describing cause and effect, but everybody's just looking at effect most of the time. Right. And you have to go to the source. Like, this is the gamba.

There's a 1911 physicist named Rutherford, British guy, was studying the structure of the atom. The belief at the time, and this is in England, so that the atoms, they knew there were particles in them. I think maybe they knew they were electrons. I don't know if that term was in use or not, but they knew there was something. It wasn't just a marble, right. That, that, that model had been replaced with one like a plum pudding. Right. So some kind of a, some kind of a, a, a matrix with these little things dotted in, you know, like a, like a cookie. He was testing this. And so I promise this lands in a relevant way soon. He shot a beam of alpha particles at a very thin layer of foil, of, of gold. Gold foil, gold leaf. And if the prevailing model of the atom was true, what you'd see is the scatter. You know, it would hit, you know, it would go through the pudding parts and it would hit the particle parts and it would scatter off. Like you'd see a particular kind of, probably sort of a kind of a homogeneous scattering thing. So he had like a detector, a particle detector around this setup. So they could see like when they shot the beam, like where it, where it went.

Yeah, like a game of pool. Like, like when you.

100%. Yeah. Okay, 100. What he saw instead was most of the radiation, mostly off particles went straight through. That was the. Aha. Atoms are mostly empty space. They're not a solid thing. Occasionally some of them would veer off a little bit. Very occasionally. Some would like ping back at a sharp angle. So they hit something in the middle. Right. And so this is this, this was the, you know, you can't see an atom directly, but you can fire radiation at it. And by the way it ricochets off, you can deduce what the structure is inside. What I'm doing is the same thing. You can't see the mental models that people hold. You can't see the beliefs. You can ask them and they'll tell you what they want you to know, that they think, they believe.

That's a well qualified statement, right.

So, so these interviews that I do, those, those are my alpha particles, right. And I'm, and I'm pinging them off. And the way they ricochet, the answers they give lets me understand the shape of the problem that I'm studying, the structure of the problem. And sometimes that structure is like the structure of the firm itself causes the problem. And just the ability to handle all of the language tasks around that, using AI is what's making this possible. It was making this a really Fertile time for what I'm up to.

So now you've defined it. Well, talk about solving it. The last time we talked about this in August, so five years ago in. Right, in AI terms.

In AI terms.

In AI terms, you know, you were, we talked about descriptions, you know, you built some templates. You were messing around with an AI roleplay surgeon. You were messing around with some different stuff. Like what, what are you finding is working to help close some of the gap? Like what experience have you experiments have you tried? And then what do you think's working to close that knowledge gap between Kate and everyone else?

Bunch of things. Not all of them. They have stuck which is somewhat predictable. Easy win is that role. Role description. Right. That's just, you know, one and done. Right. That's the sort of lowest level intervention. It still needs to kind of be rolled out and implemented. But, but the, one of the things, you know, Chris, you mentioned the a role play and so one of the things I learned from our medical planners is like it's kind of hard, I mean running these user group meetings. A user group meeting is a group of. And I think it's particular to healthcare architecture. I'm not sure other markets necessarily have this. Maybe they have an analog, but it's pretty intense in healthcare where you have a meeting of physicians and nurses and department heads, really intense people sometimes, sometimes big ego people and invariably unbelievably busy people you might be meeting at seven in the morning because that's all you, that's all they got. Right. So these meetings are really critical and they're, they're not easy and they can be a bit of a, you know, they can be a bit stressful, hard to do your first one right. What if there was some way to help a person practice in a safe place where you can fail? You can get thrown to the ground and it doesn't hurt because you landed on a pad. And so my early ideation was, man, is there a way I could build a homebrewed AI simulator for this? And probably that's possible a little beyond my. I think I'd spend enough time on it myself that someone would be like, where has Todd been for the last three weeks? But right around then LinkedIn came out with something. It was right before Ka Connect. It was like the week before Ka Connect. So they've had these role plays. You can go on there and you can role play giving feedback to a high performing employee or a low performing employee and you use a voice mode and just talk back and forth. What Came out just right there in late July, early August was the ability to make your own, where you could define your own. The Persona, the task, what's it graded on? And so I built a very quickly, just a kind of a virtual surgeon. And you're going to. We do a lot of surgery centers, ambulatory surgery centers is one of our very common project types. And so I sort of built a kind of a. By the nature of it, it has to be very, very kind of small and targeted. It can't really be a real meeting because just it's not designed for that. Right. But this thing of like, hey, I need to introduce myself, describe the point of this meeting and ask you about your workflow. Right. Kind of, kind of just a couple things and it's been interesting. It's stressful. I sort of told it like you're a surgeon, you're real busy, you're not sure why you're here. You know, you're not sure why these healthcare experts need, need you to design this thing. Right. So you might express some skepticism. And boy does she. And it's.

You're trying to create a, you're trying to put these people, give them reps under duress, you know, figure out.

Right.

Yeah. To have seen it before so they don't see it for the first time live.

Yeah, yeah. It's cool. It's cool. It's interesting. It hasn't frankly gone anywhere because it's just, I think it's just not quite there yet. It's so, it's too constrained because you just, you can only describe. You have like 500 words to describe it. I mean, it's just, you know, it's just not full fledged for what I'm trying to do. It did spawn another idea. I was showing it to Kate and she's like, you know, it'd be cool actually if you could do something that was like when we go to project interviews, that was like a trial run of doing a project interview. Well, that I was able to build in just chatgpt. I loaded up all of the kinds of questions. Every marketing department keeps a word file or a spreadsheet of quest of interview questions. And so I made this project interview role play thing that's actually quite, it's quite cool. It has different, it has like three different levels. It can be like easy mode where it'll give you feedback, you know, or there's like tough mode where it's like skeptical and it wants. It's not going to just take your first answer. It's me like, I don't know. Give me some evidence that that's true. I mean, it's so much more interesting. I'm still sort of shopping that around. I. I think this is a little bit. This may feel a little bit like, hi, marketing department. I'm Todd. I've solved something that you kind of have your fingers in here and they're a little bit like, thanks, I'm curious about that.

So Kate thought it would be really cool.

Yeah.

But is what I'm inferring here, but I don't know, like, Jim and Denise and Tony didn't think it would be really cool. Like, they weren't necessarily thinking this is the way that would help them learn.

Yeah, I think so. I don't know. Chris, you, you said it before, like, sometimes people ask for something and then you give it to them. They're like, yeah, it's not. It's not quite right. So I haven't put a ton of effort into it. You know, just a priorities. It hasn't been.

But the goal is still clear.

Right?

You still, like, we need to get people reps at being able to do user meetings or project interviews. Is that. Do you still feel like you've diagnosed things correctly based on the preliminary work you did, and now you're just figuring out the best way to solve the problem?

Figuring out the best way to solve the problem. Maybe even just waiting for the technology to advance to where it's even easier because it's just on the cusp of this being trivial, it feels like. But in the meantime, you know, other things, like, I had this idea, you know, there are TED talks. What if there were med talks? We have all these. We have all these tools that the planners have created. In true to form, we haven't really told anybody about them. And guess who knows how they work the best? Kate. So we've assigned each of these tools to one of the planners. Their job is to learn how to use it at a Kate level or a Kate like level. They can interview Kate, they can talk to Kate. Kate's going to coach this process. Whatever. The only rule is Kate may not do any of these presentations. It has to come out of one of the other planners.

Ha ha.

And then it's going to do like TED talks on one of these tools. So teach BA how to use this tool, what it's for, why to use it, where it falls off and when. You should contact one of the medical planners to go deeper. Right. It does a couple things. It elevates that planner. It Transfers some, some of Kate's knowledge. It ties that planner to that tool. So Kate is, is Kate's able to, to elevate and be at Kate's level. And these folks are starting to be that second level of, of expertise, but on a narrower piece. So it's, it's bite size and Kate's.

No longer a bottlenecked on supporting that tool. So she's been able to let that, let that responsibility. It's so interesting because when people say we want to backfill Kate a lot of times, like we need to find a, we need to make a clone. And that's usually very, very hard to do with someone that's been working for decades to build that expertise. But we're going at is decomposing important parts of her knowledge in this case that are tied up and embodied in tools and distributing that. Right.

Is it, is it the, the best or the only way to kind of solve the Kate effect? I don't know, but it's the one that like meets. It just checks a lot of boxes right now. And it's kind of fun, frankly. I mean, med talks. Who doesn't love that, love the little drippy sound. I mean, you know, the whole thing.

I like how you're introducing people into this puzzle of scale.

Right?

Because I think a lot of times when you talk about scaling knowledge or whatever it is skills, you're usually, I think, approaching it from like a technology standpoint. Well, the tool handles the thing, whatever it is. And I think it's really interesting that you're saying, okay, now this person is going to be the. Become an expert because that you, you're investing in that person, first of all, which I would think they would really appreciate. They're giving, you're giving them the time to do this, right? Because everybody's got deadlines and all these things. So you actually have to be really proactive about that investment and say, no, like we need you to learn this. There's value in that. And then we're going to take that and you're going to become the spokesperson to deliver that information to an even broader group. Maybe it's a small group at first, and then you get some feedback, but you build in to the culture of the firm then how you approach scaling knowledge, which is more than just a technology solution.

Right. But you also record that. So you're also externalizing it and digitizing it at the same time, which helps with retrieval. And like, so you're doing both things at the same time. Right.

It just. Exactly. And it's eating away at the structural problem. The structure of this problem is that we have this wide base and then Kate. Right. And what we need is, you know. Yeah, my hand gestures aren't going to do well on a podcast, an audio podcast. But we need a pyramid scheme, not a tall, skinny tower that's unstable with just Kate in it. And it's a way of solving the structure. And again, a tool works at both ends. So this will shape the culture of the firm a bit. The place. I want to put this carefully. It's not about taking anything away from Kate or how people see Kate, but it's broadening the base.

Well, I mean, if you think about the amazing person that Kate is, Kate is a rock star. Kate is an asset. There is a liability side to that equation when it's the single person.

Right.

So absolutely. I mean, it's. You have to diversify that knowledge throughout the firm.

Yeah. And Kate's delighted, You know, Kate's delighted by this. She doesn't want to, you know, that circumstance. Right. She's probably terrified of buses, you know, as a result.

Or vacations.

Lotteries, maybe. Not lotteries.

You know, I can't go on vacation without, like, and actually turn her phone off and worry that she's not going to get. You know what I mean?

Oh, that's a great point. I love that.

One of the things that's so interesting about this is like, you know, when I first got started knowledge management, you know, people say knowledge is power and people, like, want to hoard knowledge. And like, I don't feel like in 2025 I hear much about knowledge hoarding anymore. I feel like people are busy and their knowledge is changing so fast that they're like, I just want to help. I just want help, you know, Like, I don't want to be the, the bottleneck or the choke point and hold this organization hostage. That's not my intention at all, you know, and.

And on top of that, my calendar is back to back meetings I have. I can barely just do what I'm doing right now. I actually can't solve the problem that I am in the middle of. And that's, that's, that's really what's becoming my role is I'm someone who works, you know, on, on the structure of the business, on the firm. Yep. Instead of for it. And so I can take these things on and I can just be like, sit with me for a half an hour and, you know, and then I can run with this. Once I capture your Expertise.

You're working on the important but non urgent things for Boulder. You know, while a lot of people are working on urgent fires, you're just trying to prevent fires from happening in the first place.

Yeah, yeah.

Before we get to Infinite Todd, which I know everybody's waiting for, I'm going to keep pushing it out a little bit further. Deal you had. These are my words. You correct where I got this wrong. You had enough success with Kate and the medical planning accelerator that the firm came to you and said, like, hey, dude, like, project management, like, medical planning is impacting this percentage of the business. Project management's a big challenge for us.

It's.

It's impacting the whole business. So talk to me about taking things, the things that you've been doing in medical planning and taking that over to project management and what that looks like.

Yeah, the, the as a segue. I'll mention that a frustration with the medical planning thing is that AI has made me really fast and everyone else is still going at their pace and are overburdened and over, you know, over scheduled.

And so it's like limitless. Like, you're like, you know, Bradley Cooper. I don't know if you guys have seen that movie, but he gets the drug that, like, lets him operate at like speed, but everyone else is still moving. Okay, yeah.

All right. I wouldn't, I wouldn't. That would be a bit boastful. But, but, but, yeah. And so I love getting stuff done, but I'm constrained by just the, the pace that, that everyone else goes. But that means that I actually have bandwidth to kind of keep pushing multiple things. And so just like Chris, you said this other opportunity of, wow, you know, we've gotten, we've gotten to a certain size and we're not going to get to the next size if we don't just sort of figure. Figure out the things that frankly, other firms have figured out, like project management. Like seriously standardizing what it means to be a project manager. What do they do? What authority do they have, et cetera. I'm actually getting ahead of myself. So, you know, the first thing I did was, guess what?

Gamba.

Bunch of interviews. 100%. These I kept confidential. I picked folks because I kind of wanted to, you know, wanted to sort of dig because this is unlike the medical planning thing, which is, you know, important, but it's sort of a narrower problem. This is a deep cultural thing. This is, this is at the heart of, of architecture, of interior design, of. It's at the heart of revenue. It's at the heart of doing any of the things that we like to happen as a result of doing a project. Whether it's sustainability, whether it's happy clients, whether it's making money and make. Make money and stay in business. Runs through project management. And, you know, we're 40 years old. We have some entrenched thoughts on this. And so I knew this was like, careful. Go. Be careful. Right. Dig deep, but be careful. So did these interviews anonymously picked folks randomly who'd been through our psmj. We do a PSMJ bootcamp like a lot of firms do, only. Only what we do is something special. It's like Fight Club. Never talk about it again. The interviews are great. Just surfaced again, trying to find out what is the structure of the problem. I had my own. As someone who raised up in this firm, I had my own sense, but tried to put that aside and just hear it. And it was all those things. Our PMs don't know what a PM does, even though they'd all gone through the same training, but we'd never talked about it again. Right. Unclear authority. They would even say like, yeah, the lead designer, they just do what they want. And I'm like, but the work plan. They're like, whatever. Right. So not feeling like they actually can manage the project. Oh, and that kind of the biggie is so. So welcome to older associates. Oh, you're a project manager. Cool. We're going to send you to this, this boot camp, you know, fall of every year, and then guess what you get every other month, A meeting with the CFO to talk about how your projects are doing financially. Good luck. Bye. And that's it. That is. That's it. So those meetings, you won't be surprised to hear, were not well loved. On the whole, I'm understating the situation. So do these interviews kind of surface all this? And it's just like, Yep, you have PMs that are scrappy and inventive. Good on them. They survive through grit and determination in spite of the fact that there is no supporting structure around them at all. Except they once in a while get to answer for financial outcomes that they really aren't prepared to even impact that much. Shocking that this is a concern that's coming up. Right. And so I think the report that I sort of wrote, you know, with ChatGPT back and forth three or four pages, and I really looked, you know, from a system systems thinking lens at this, like, at the level of belief, what is. What is holding us, you know, where we are at the level of sort of structure and authority and rules and everything else. And when I gave it to, you know, the folks I report to who are project managers, right. They are living this as well. They're a light turned on in a unique way where I think they were able. I was able to paint a picture of the problem as lived by people who are in it. And. And again, quoting them because it's all fed by transcripts. So every time I just check my instructions to chat, GPT is like every point you make, illustrate it with something somebody said, tell me who it was and what timestamp they said it, because I'm going to go check. If you don't do that, sometimes it makes things up. And this was this amazing moment of like, being able to share, like, here is what this problem is. We're all focused on the outcomes, right. The behaviors, but really we have to be working at this other level to get this straightened out. And that was really motivating. So I've. This is what I'm. This is kind of half of what I'm doing right now is, you know, judging project management. So a really good role description. Right. There's just a lot of parallels with the medical planning thing and that's gone through a bunch of iterations and it's looking really good. And then we had another boot camp. And so one of the things that was so clear, that was so clear in these interviews. No, I'm sorry, I'm going to retract that. I'm going to say it this way. One of the things that AI surfaced from the interviews was the void was something no one said. One of the questions was, hey, when you think of strong pm, who do you think of and what makes them a strong pm? And AI was able to go, you know what? Nobody actually ever answered that question. Your PMs generally don't know who PMs are. They are operating in a complete vacuum.

No community of practice.

There's no community.

The knowledge management term. Right.

There's no community at all, even just awareness that each other exists.

Yeah.

Like one of our PMs. And she's like, I literally cannot. She actually said this out loud. I can't name another pm.

Wow.

What? Yeah. Holy cow.

That's so interesting. So who are you learning from? How are you getting better? Like, you're just on your own, just, you know, all running experiments, just.

You're just guessing, you know, And I'm better than that. These folks are better than that. You know, they're. They're smart. They have, they have colleagues that they can talk to. But, but the craft of being a project manager is a distinct thing from, from architectural design.

But when one person learns something, Boulder doesn't learn it. Right?

Yeah.

There's not organizational learning happening there. That's individual only.

That's right. That's right. And also we have, we have a, a feedback loop of. To be a successful PM is because you are scrappy and you figure things out and you kind of come up with your own way of doing stuff because we don't have a way for you to do stuff. And that feeds it. And that, that becomes kind of the, the culture of project management is these rugged individualists. Right. So. But that cult, that, that loneliness. There's a profound loneliness amongst our PMs, at least in terms of connection to other PMs. And so. Right. The two big things that I'm working on are clear role definition, which is more than just writing down the role, but it's also like we need to operationalize this role. I need the project executives to own that they're going to support this and that the PM has some authority to manage the project also just at a basic level. And this funny. It's funny the, it's funny that the debates you end up having. But like we need to be able to point at a person and say, you are the PM on this project. That's not universally agreed to. Some, some folks have the. It's like, well, PMs, kind of a vibe. Like these are some things that need to get done and we're going to kind of like when you're, when you're an eight person firm, that's fine.

Vibe PMing.

I love vibe PMing. Right. When you're, when you're small or your projects are small, that's fine. But we're not, right. And we have aspirations. So the community and the role, those are, those are the two, the two main prongs of this. But they do a lot. I mean the role itself is cultural, cultural evolution work because it's. We don't have any roles that are as kind of like there's accountability and then there's authority that comes with this. If we're calling UPM you I should 3D print some badges. It just occurred to me, right, we're giving you some power here and we don't talk like that, but I think that we need to.

What's so interesting, Todd, is it so much of the noise around and the talk around a. I shouldn't say Noise. Because it's not all noise. But so much of it has been around, like, efficiency. Like, the end state is you don't have to do the stuff you don't want to do anymore. Like, there's been a lot of, like, things that end in kind of, like, products that take away work that nobody wants to do. And the profound way I think you're describing AI is as an organizational designer, because that's what I look at what you're doing as is you're helping redesign Boulder, and it's a research assistant. It's helping you document, it's helping you explore solutions, it's helping you redesign the firm to do really human work. And a lot of the solutions you're coming up with are human. Like, we just need to be more connected, more clear and more aligned. Like, and it's helping you 100% get there faster than. And I just feel like that's so interesting to me.

Yeah.

Like, what kind of teammate are we trying to hire? And maybe we're going to get to Infinite Todd. But, like, if you think about what you want AI to do and who you want it to help, you become, like, this is helping it. You become a more connected and human and clear and distributed and all these things.

Right? Yeah. I think there's a. So I deal a lot in mental models, right? And I think there's. I see on LinkedIn and just everywhere, this sort of model that, like, if you give something to AI, it's just crap, it's just slop. It's necessarily a lessening of what you can do. You're deskilling yourself. And I'm like, well, if you use it that way, I think that could be true. If I said, here's my problem. Solve it for me. And then I just run off with that, then I guess so. But, you know, at the end of the day, I'm an architect and I'm not working on buildings anymore, but I am solving the architecture of these problems of the firm. Right. And the AI is. Well. And maybe this is finally the bridge to infinitad. The AI is just the thing that gives me superhuman bandwidth, superhuman attention, superhuman recall, and can write stuff pretty good in seconds, and then I can edit it in minutes and, you know, make things happen at a. At a. At a pace that it just couldn't even imagine before.

Okay. So Infinite Todd Brain Trust, Council of Experts. Go.

Yeah, so, yeah, I mean, it goes back to the medical planning thing. I. I was, you know, out running one morning and sort of thinking about this challenge and I was getting really quite cranky because I'm like, I have got all these things, all these plates I keep spinning onboarding and the Biscuit Net and half a dozen others. And this is such a cool, wonderful KM opportunity. But I don't, I have no team. This just talking about structural problems. It's me, myself and I right now doing all these things. And then, you know, then it, then it hit me like, well, guy who is trying to tell people to use AI a lot, why don't you eat your own dog food? Yeah, and I kind of had. The one thing I like to do sometimes it's amusing, is take something very literally and see where it goes. And so it's like, yeah, what if AI was my team? No, really, what if AI was literally my team? What would happen? Right? And, and one of the things I, I thought of was this, this, this council of experts. Like, I, what I, part of what I need in this work is advice is just input and, and kind of that, that, that back and forth to, to chew on ideas and bounce things around. And so yeah, I had this, this notion out on a run, came back and I just, I, I, you know, in a ch. I said, you are a council of various experts, you know, containing these Persona. Persona. I was like, I want a knowledge management expert and I want a learning and development expert and a psychologist and an organizational development and blah, blah, blah. I think I, you know, I think I put in like eight. And then, you know, the first query I gave this group was hey, you're here to help me. Who else should we invite to join this council? And AI was like, well, you need, you know, blah, blah, blah, blah, blah, blah. And it's like 20 some odd Personas. It picked names for them and gave them each a little kind of role description. And it's, and so the way I use this. And it's kind of this funny thing, right? Because it is, it's just chatgpt. It isn't actually 20 something different Personas. I had wanted them to kind of debate or disagree, but it can't, it can't disagree with itself. But what it does do is it puts on different hats and so it lets me inspect an idea from a bunch of different perspectives. And again, most of it's like, ho hum. And then every now and then you're like, oh, there's the goodie, right? That's the good stuff. And I use this all. So I have a, I have a personal ChatGPT account that I pay for. And then I have my work one and I have one of a sort of a similar council of experts in each one slightly different. Right. And I bring it. This isn't the place to bring like, what's the population of, you know, Canada. It's like I'm trying to think about how to help project managers feel more community. You know, just bring it kind of a big, wide open thing. And it, it's wonderful. And I just. Just go back and forth and sometimes it just digs up a nugget of something that's like, yeah, I could have thought of that. I don't know if I would have just because human limitations.

So it'll come out from a lens, like a learning and development lens. Here's what I would do. But if I were a organizational psychologist, I would say like, this is the problem or a community.

Yeah.

Building expert or something like that.

The way it tends to work is kind of like, you know, there's that. That rule in improv of like, yes. And like whatever the person before you does, you never just go, no, you kind of roll with. Feels a little bit like that. Each one sort of like, I agree with what she said. And I'll add, you know, kind of the organizational development lens or, you know, psychology. And there's one of the Personas is the overworked architect. And you know, he generally jumps in with like, don't forget, you've got five minutes to prove to me that this isn't a waste of my time or I'm out, you know.

Right.

The answer can't always be more. Right.

Yeah. The answer can't always be more.

You need the skeptic that's helping to push back on the enthusiasm of the other counsel.

100%. Yeah.

Yeah.

And it's, it's just been a rich. A rich resource to just kick stuff around that's very blue sky thinking, you know, generative. I mean, truly generative thinking. I use it. Use it all the time.

It's kind of in the neighborhood of the simulator you were kind of trying to build. Like you can get reps with these different. I know it's a little bit different, but like, there's this kind of. How can I run this simulation first before I go put it in front of real people to sharpen my thinking.

Totally. And it, you know, and it's sort of a way of maybe what it does is it enact a certain thoroughness of thinking. Like I want to. Well, two things. A certain thoroughness of thinking. So think about it as a KM Person. Think about it as a busy architect. Think about it as a psychologist, cfo.

Right.

I'm actually. Right.

Yeah.

I'm not a psychologist. I don't even know entry level psychology. But I bet you ChatGPT does. Right? So. So it's kind of the thoroughness, but also just access to basic, you know, not amazing psychology, but like, you know, 101 level. Fine. That's better than what I've got on board right now.

Is.

Isn't the traditional guidance to write it at like a fifth grade level, for example? Right.

Like it.

And so I assume it's probably kind of in that realm with a lot of things because that's what it's basically trained on that.

I mean, what else can it give but what it's trained on? Yeah, but it, but you know what, it has read everything on the Internet. I've read one, one millionth of what's on the Internet. So it has a lot that it can show me.

Yeah, I like what you pointed out about that idea of the nugget because I think there's so many things out there where it's like that big giant thing, that podcast, that movie, that series, that talk that I went to, that seminar, whatever. It's like, oh, generally most of it wasn't for me, but oh man, there was this one thing and that one. And so I'm just curious from your point of view, like, you've brought that up a couple times now. And that to me is a big takeaway from the type of interactions that I'm hearing you talk about. That really is resonating. I think it's valuable.

It's. It's funny to me how, how a lot of people, you know, sort of, sort of take AI or interact with it and they're like, well, I don't want to use AI because I make mistakes. I'm like, have you ever worked with a human being?

Have you ever made.

Have you ever worked with someone straight out of college and tried to get them to do something? I don't know. Have you ever tried to get an architect to write something? Yeah, it makes mistakes sometimes. Those hallucinations are fantastic. That's what's called design in my world. Right. So you understand the nature of the tool that you're using and you approach it with the appropriate respect or, or, or whatever the word is.

Whatever.

Use it for what it. Empathy, you know, and, and you go, yeah, 80% of what it gives me is going to be meh, fine. Yeah, I'm here for the 20.

But you're being. You're patient. You know, that's part of this, too, right? Is you're just like, you're after the. Yeah, it's.

Yeah. Yeah. Well, it's a conversation. I mean, it's a. In one sense, it's just math. It's just software. In another sense, it is a very different creature. And it. Yeah, you go back and forth and you tunnel into the parts that seem interesting and you kind of, you know, let go of the stuff that's like, yeah, I got that. And I don't know, gives me access to a thoroughness of thought, a quality of thought, depth of thought that is, for me, has been a game changer. And I'm still the editor. I'm still the director of it. I kick the stuff out that isn't good. But when I see something that's good, I'm like, holy cow, that's amazing. Off I go. And that's just pure win.

It's. You know, we named our newsletter Smarter by Design, kind of celebrating this idea that, you know, architects should design their businesses with the same intentionality they design buildings and bridges, infrastructure architects and engineers, AC firms. And I think that your story that you've been sharing is very much a design story. And then AI is a design partner for you in helping kind of level up Boulder Associates as a learning organization and figure out, like. And I kind of want to maybe just spend a couple minutes as we close on, you know, you and I have talked about this, Todd, that we think that, you know, the future of learning and development and how firms in AEC learn is going to radically change in the next five years. And you've kind of laid a lot of groundwork in this discussion around kind of the AI piece of it. You're also in our beta for the learning management system that we're building. And I know that long before you got into that, you know, we've had conversations for years you've been exploring really doing learning and development in a really meaningful adult learning, principle driven, like, smart way. And like, and I'm curious just, like, how these worlds are connecting in your brain between what you're doing on the AI side and all that stuff we just talked about. And now you're thinking about how do we upskill and educate Boulder as a learning organization going forward? And it's like, where are you on that, man? In 300 words or less. I'm kidding.

Yeah. Just a small, small ask right there.

Yeah. Yeah. What comes to mind is the everything bagel from the. The movie Everything Everywhere, all at once. Right. Because. Because not only are these capabilities. I want to jump back to something from a while ago. You talked about roi, Chris, half an hour ago. The return on our investment is going up in knowledge management activities and the investment is going back down. Yeah, Right. And both of these are happening fast. And so, yeah, that ro. The. The division of the two is changing, you know, radically. I don't really know where this goes. I didn't know what I needed an LMS for before I got the chance to poke around in one and kind of unlike, have it shape. Have it started inhabiting my thinking. Right. And so there's that, there's the evolution of the technology. There is also. I think I'm probably drifting away from your question, but we are in the most vuca time I can imagine, right? Volatile, uncertain, complex, ambiguous. We are fundamentally mostly healthcare architects. So I have this mental model that we are on a planet that is not doing great. It's getting ready to kick us off. We are in a field architecture that is really upside down. Our business model makes very little rational sense nowadays. It's not looking good. We do a lot of healthcare. Health care in America in particular does not make a lot of sense. Not looking good. It's shaky. And there's AI. So my overall view taking, like everything is just wobbly and radically wobbly. And I think. I think architecture has been overdue for a disruption for a while. And I think the disruptor is here. I think we know who it is, it's AI. What that actually means. That's the thing about disruption, right. Whatever the question is here, agility is the answer. Not clinging to what we have done before, not clinging to who we think we are, not clinging to business models, not clinging to beliefs and ideas. Get ready to dance. Because. Yeah, yeah, right. Because it's time. And so really the way I meet this moment is I don't know what AI is going to do. I have concerns. I have, you know, I see possibilities. But what I know for certain is I want to understand it deeply. I want to be only. I want to be in the front figuring out what it can do rather than find out what it did.

I think. I love what you just said. I think the way I would translate that to someone who's spending a lot of time with people in that beta is I've been using the phrase continuous onboarding a lot. And there's this notion that onboarding, if you just hear it said with no qualifiers, it's like cool. The first week that somebody's at the company, or maybe in a really good example, the first 90 days. But the truth of it is, it's like you're always onboarding into something new. You're onboarding in a new software. Now you're a job captain. Now you're working on ca. You've never done that before. Now you're working in retail because you worked in health care. Now you've never. You're doing an ambulatory thing because you've never done one. You've only done Nikki's like. And then the technology changes. So, like, that get ready to dance thing that you said to me is another way of saying this. Continuous onboarding, it's like learning and unlearning is going to become as important as learning in some ways. Because the thing that we just thought was the best thing to do two years ago is no longer even close to true. And we have to be willing to let go of it. And it's like a beginner's mindset out of Zen, you know, it's like you. Because things are in a vuca environment, it's like the best we know right now, and we don't know how right now, how long right now. It's gonna. It's kind of stressful, but it's the only way through it is to. I think that is the truth of our times.

Now you can deny it and look away.

Right?

Good luck to you. You know, it puts me in mind this movie reference is probably too dated, but like the Matrix where there's that big old USB connector in the back of everyone's head, and you can be like, hey, I need to know how to fly a helicopter, because I just got in a helicopter and it's like, okay, cool. Like, in the limit. I think that's. You know that. Like, I need to know X now.

I think it's so funny you just brought that up. I was just having this conversation with my wife last week about that scene in that movie. And the question we had is, like, why don't you just upload everything in the beginning just so that you have it all? Like, what is it because the brain has limit? No, I mean, maybe we got to. It's like. It's the anatomy, right? The brain has limited capacity to hold information. So therefore. But then if you want to upload the helicopter thing and your brain is full, do you have to then delete the Spanish module?

Like, yeah. Anyway, in your head, you have to go through this menu of. Yeah, well, I don't need to remember what A cookie is anymore. You know, whatever.

Auto delete. Like they get the helicopter thing, but I set it to delete two weeks from now because I know I'm not gonna. Anyway.

Oh, no. I deleted my family.

Yeah.

Whoops.

You know.

Yeah. Yeah.

That's great, Todd. I mean, I think. I think you've really gotten at something Very true there. That and I think the role. So the kind of. What your role in practice improvement is, is if people are getting ready to dance. You're a little bit of the. What, the orchestrator? The guy cranking the record player. You know, you've actually. No, I know what you are. I like that you started with a thing about the jar. And you are telling Boulder what's. What the jar is. Or at a big level. But you're telling project management what's in the jar. You're telling medical planning what's in the jar, you're telling Bluebeam what's in the jar. That you know, that people. And it's. It's. You've got this kind of way of being both in the business, but then also kind of looking at it from outside at the same time.

Yeah, yeah, yeah. What they don't know is all of the jars, they're just hot sauce. All hot sauce.

I. What I. What I'm hearing is that Todd is super proactive about identifying issues that nobody is really asking for help on. Architecture school, like to go back to this idea where Todd was like, okay, what's the source of the problem? A lot of times the source of the problem is in the training. Right. It's just like, this is who you become as an architect. You are a lone ranger. You have to solve all your problems. You get to work on a project by yourself. You get to present the project by yourself. You get to build the model, you get to do the drawings, you get to do all the things.

Yeah.

And. And rarely do you actually get to ask for help. Like you're forced upon you the death's crit. Right. Like, you don't. Some. Some architects were like, yeah, please, please help me through this. Help me see what I'm not seeing. What are my blind spots? Of course. But I think a lot of times, like, well, there's deadlines and we got stuff to do, and I'm going to do it the way I was trained or the way that I was mentored or whatever, and I'm just gonna figure this out. And then there's Todd. And Todd is coming in and saying, well, I see a bottleneck, I see an issue. I see a future problem here. Maybe it's not a problem right now. I see if. Okay, I need. I see a scale problem, I see this kind of problem. I see that kind of problem. And what I love about this is, like, the proactive nature of working on the practice versus working in the practice, because in the practice, you've got blinders on. Like, you've got a project, you've got a deadline, you've got a hierarchy, you've got the politics, you've got the culture, you've got all those things. And if you really can step outside of that and start. We need more people like this in more firms. Right. Because that's what's going to elevate not just that practice, but the profession. And I think that, like, hopefully, through conversations like this, that that's a big takeaway and that people will step into those roles, because that's my goal.

Designers kind of thing. Yeah.

Right.

Right.

Todd, do you. I'm curious that. Because I think I'm trying my best to. Yes. And what Evan just said, my feeling is that some of these things you're going out and proactively discovering, but you're also getting inbound from people as they're finding out more and more what you're capable of. And now you've got to prioritize which of those things to take. But you're also. Not necessarily. What I'm also taking from this conversation is you're not necessarily accepting the premise of how they've defined the problem and what their proposed solution is. You say, like, if I'm going to do this, we're going to go. We're going to tear it back to the studs, and we're going to go to Gemba and we're going to find out what people think and build back to, like, validating this thing is the way that you've described it is that.

Yeah, I think that's. I think that's. I think that's basically right. Part of what I do, in fact, part of what I've always done, is I've been at this firm for 23 years, and I didn't have this term until not that long ago, is I moved the Overton window on things. Right, Great. Yeah. The, The. The range of allowable thought. And I have. I have sensed that, at least for my firm, the wavelength of change is about three years. So it. I have to start talking about something, and in about three years, you know, we're ready to sort of, like, move on it.

Like a Doppler which is. You talk about it and it takes three years for it to kind of.

Yeah, it just, it's just like at first, it's like, here goes Todd. Right. Talking about equity in 2017. The heck's that? Right. And that's not, that's not a boast. That's just. It's the nature of the.

The diffusion of innovation curve. Right, Kind of.

Diffusion. Yeah, Diffusion of innovation and just new ideas just need. People just need to kind of encounter them a few times and then.

And then you got to incubate.

Yeah, yeah. So when I was proposing the role that I now have, you know, I gave this list, like, they're like, well, what would you do? Like, well, you know, and kind of just from. Not from deep research or anything, but just from having been in the firm and of the firm. Like, well, these are some things that I think need work. And so some of those are now actually coming back as somebody else's idea.

Nice.

So there's a, you know, there's a. There's a. Yeah, the Doppler effect, the kind of ripples.

But what you're not saying, though, I think is what you're inferring is that they've come back as someone else's idea, but now it's better for you to. This is the better time to try and fix project management because it's not your idea to fix project.

Well. And you're a partner to them.

Yeah, right.

Yeah, make it happen.

Yeah, exactly. And also. And also it's like it's got leadership's attention. I mean, you know, this one, project management is such a great example. I cannot think of a thing more centrally like, we're going to move the needle on things we care about. This is it. It's not revit skills. We got those. It's not, you know, what, whatever. I don't know. This is super for us at this moment right now. This is a huge opportunity. So it's the right thing to work on.

Yeah.

Right. And then after that, something else will be the right thing to work on.

And the next most important.

Good.

Yeah, exactly.

Yeah, that's great. Yeah.

No, but I mean, I think that's a good lesson for people that aren't in knowledge management or new to it. It's a like just what you described is like one of the best pieces of advice. It's like find a problem that really matters to your company and then help them make progress on it. And yeah, not just doing random acts of knowledge management because they're fun science Experience like, you should do some exper. Maybe you need like 5% or 10% experiment time to just see what happens. But like, your core work should be the kind of things that you've been describing with us today.

You know, my. Like where I kind of started with. With some of this, like the interview with Darcy. That was in a sense a beta test. And that was a pretty small thing, very constrained notion, but it sort of proved the potential of a way of acting. And like, hey, I can very quickly generate value from something. And even though that wasn't enormous, I think those types of things, even if it's not the most important thing, but just like the opportunity that presents itself, kill it, get it, do it, you know, and it starts to build credibility and build your own abilities.

Solve a real problem.

Yeah, solve a real problem.

Doesn't have to be the. Yeah, that's fair.

If.

If project management came to me as like, okay, Todd, you have this role now fixed project management, that would not have gone well. It's actually good that I've had some. A ramp.

Yeah, no, I think that fair. I've revised my previous statement and I substitute it with your new one.

Good idea.

Gavin, do you have any parting thoughts or comments?

No, I think I gave them. I feel like this is a model for firms to really pay attention to because it's rare. I would say on some level that this stepping out of practice, working in the practice to working on the practice, something that I've talked about a lot over the last five years and I. It's super, super important. So appreciate you coming on and sharing what you're. What you're doing at Boulder.

Well, thanks for having me. This has been just a joy and a pleasure.

I've enjoyed it. Thank you, Todd, for being so generous and transparent and I've been. If you can sense that I'm not wanting the conversation to end. That's two for two reasons. Partly I can't think of a kind of topic I'd rather discuss so I could keep going. Also, Evan and I have been working on this series for the last seven months. This will be our last episode of in this kind of run of this series. I'm sure Evan and I will partner and do fun stuff together in the future. But like, this is the end of the 2025 welcome to KM 3.0 series. So I just wanted to say big thanks to Evan for making this possible. Like, I've really enjoyed collaborating with you, all the people in the series. I've learned a ton. I know I'VE gotten a ton of great feedback from our community. Random people at Autodesk University saying, like, I've been listening to the thing. I love it so much. So thank you. Evan. You've built a really great thing here at Troxell, and we really admire and respect the work you're doing.

Well, thank you for helping make this kind of thing happen. I believe more of these kind of partnerships do need to happen, and I can't wait to see what KM 4.0 looks like.

You're working on that, right? Slow down. Hold on. Hold on, buddy. Let's go with 3.1. Sounds good right about now.

You're the software guy. I'll defer to you on that.

All right, thank you, everybody. Thank you both.

All right, thanks, guys.