KP Unpacked

The Data You Share Is the Advantage You Lose

• KP Reddy

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 53:29

What happens when AEC firms ban Claude because they don't know where their project data goes?

In this episode of KP Unpacked, KP Reddy and Nick unpack the regression happening across construction firms: people disconnecting Claude, companies banning enterprise AI tools, and employees carrying two laptops (work and personal) to keep building with tools their firms won't approve. A 3,000-person AEC firm just banned Claude entirely. The result? Everyone's using personal instances on company time, and the firm loses all institutional knowledge being built in those sessions.

But the deeper conversation is about IP anxiety in project-based industries. In AEC, there is no enterprise, the project is the enterprise. If you're a civil engineer on the Tesla factory and Tesla says "don't share our data with LLMs," how do you even comply when Claude's connected to your email? The answer: firms are hitting pause out of fear, not strategy. Meanwhile, KP delivered his first Zero RFI keynote at Building Transformations, and the feedback was split. Some GCs realized Zero's tools could drive risk to zero, which raises an existential question: if owners don't need insurance against risk anymore, why hire a general contractor?

Key questions answered:

  • Why did a 3,000-person AEC firm just ban Claude entirely?
  • What happens when employees carry two laptops to keep using AI tools their firms won't approve?
  • How do you protect client IP when Claude's connected to your enterprise email?
  • Why are AEC firms regressing on AI adoption instead of accelerating?
  • What feedback did KP get from his first Zero RFI industry keynote?
  • If Zero can drive project risk to zero, why do owners need general contractors?
  • What are owner-controlled insurance policies (OCIPs), and why don't more people use them?
  • Should firms invest $200/month per employee for enterprise Claude, or keep blocking it?
  • Why do some firms still run on-prem Exchange servers instead of migrating to cloud?
  • How do law firms handle attorney-client privilege when connecting email to LLMs?
  • What's the difference between major muscle tissue (Procore, Autodesk) and connective tissue (Zero's tech stack)?
  • Why is Microsoft Copilot "good enough" for 700K Accenture licenses but not for startups?

If you're an AEC firm struggling with data privacy policies while employees build workarounds, wondering whether blocking AI tools protects you or puts you further behind, or trying to understand what happens when risk mitigation becomes automated, this episode will force you to ask whether hitting pause feels safe, or just delays the inevitable.

Listen now.

Hair Grayness And Tech Myths

SPEAKER_03

Hey Nick, how's it going? Back in business, baby. KP intack. Might we might be wearing the same clothes as last week, I don't know. Depends on when this episode airs. The only way to find out is if you watch us on YouTube, right? Gotta watch both episodes to know. By the way, I had I heard someone say um earlier today that they watched the episodes. Really? There's people out there.

SPEAKER_00

There's people out there. Well, I mean, I have to wear a hat, so I can't that's a because I can't compete with your hair, so I just put a hat on.

SPEAKER_03

Your hair joke and pretty funny. I didn't even think about that today. I should have totally like gone to the salon or something. You need um you need to get an Instagram account for your hair. I had so my wife's sister was here with her boyfriend, gosh, it was probably two weeks ago staying at our house. Her boyfriend has a 13-year-old daughter, and she anyway, she found out she she guessed that Sarah was 18 years old, which I was like, you're like your age, like the the age, the age guesses are I couldn't quite say this in front of Sarah, but you're the the the age range that you're that you're you know reading on humans is not totally super accurate right now. So, but anyway, she made the comment, she's like, Oh, okay, well, how old are you? I was like, Well, I'm 36. And she's like, Really? You're 36. How is your how is half your head turning gray if you're only 36? Wow. And I about fell to the ground. I was like mortified that that she just called me like an old man and my hair hair's gray. I'm already self-conscious about it.

SPEAKER_00

You know that you can you can do stuff, you can do stuff. By the way, you know, in the medical field, my understanding is they can actually cure grayness. Can they? Yeah, they can actually cure grayness, but there's not enough incentive to do so, right? Baldness, there's an incentive, yeah, right? We're gonna keep you hooked on it. But the color industry is so massive that there's no there's no reason to solve the problem. There's no reason to solve it, right? That's like when people talk about the light bulb that lasts forever or the tires that can last longer. It's like it's that it's that same conspiracy theory, right?

SPEAKER_03

Well, I'd buy the product.

SPEAKER_00

What's that?

SPEAKER_03

I'd buy the product. I'm ready to reverse my gray hair. Maybe there's a peptide. I'm sure there will be. There will be. There are there actually are some that are like in and around that idea. Yeah.

Why We’re Reviewing Posts

SPEAKER_00

All right. So, what do you want to chat about today?

SPEAKER_03

We want to go through some we're gonna go through some posts today because last week we just did a rift. Yeah.

SPEAKER_00

Do we have any reader listener mail these days?

SPEAKER_03

I haven't got any reader mail recently, but you probably I mean, it's coming from your inbox, so you tell me. How many episodes have we recorded? I mean, gosh, I think we've been doing it about a year now.

SPEAKER_01

Yeah.

SPEAKER_03

So I don't think we've missed too many weeks. I mean, we I'd say we've probably got what 30 episodes?

unknown

Yeah.

SPEAKER_00

Oh, at least. Yeah, I think so.

SPEAKER_03

It's not bad. No.

The AI Skills Gap Emerges

unknown

All right.

SPEAKER_03

So a couple of KP posts. Let's just run through the gamut. So let's start with some Substacks.

SPEAKER_01

Yep.

SPEAKER_03

The post I'm bringing up now is the skills gap that AI just created in AEC. Subtitle, if you are still using chat as a search search engine, you are two years behind. So I read this post and I thought it was really compelling. The you know, my so you reference your book, right? So let's bring up your book as like the as the the re the research entity behind this post. And you reference the changes in models since you wrote wrote your book. So when you think about when you wrote your book, the model capabilities then, and your prognosis on what was you know what was going to happen, and then and then relate that back to the industry and kind of what's what's manifested adoption-wise, like yeah, walk us through yeah, origin of that post.

SPEAKER_00

So I think there's a couple of them. And I kind of highlight all the things that have happened, right? One, we didn't know much or anything about the idea of an MCP server. We were not talking about agents back then, right? We were very focused on oh, I can create a funny image and/or I can create some text, right? It was more in that world, and then you look at that, and yesterday, Autodesk released their Revit MCP server. A couple days before that, Trimble released their SketchUp MCP server. So I think I think there's been a couple dynamics, right? So now, and then you throw agents on top of that. So, in a world of like what we were when I was running around giving my talks. I mean, think about it. Like I was giving my talks about, and literally the talk I was giving around my book was this changes everything, was the talk. That was the title of the talk. This changes everything. And the entire narrative there was my dad bought a PC. He told all the engineers in his firm that his son was gonna replace them all with the PC. My my 13-year-old son and his computer are gonna put all you engineers out of business. Yeah, well, well, that didn't happen, right? The internet came along. Oh, internet's gonna put everybody in business. That didn't happen. Before that, Excel spreadsheets were gonna put accountants out of business before that. You know, Lotus 123 before Excel, actually. And the funny thing is, one of those accounting firms, Anderson Anderson and Arthur Anderson, embraced the technology and started Anderson Consulting, which became Accenture. That was their hedged bet against oh, accounting is gonna go away because of spreadsheets in the IBM PC.

SPEAKER_01

That's a fun fact. I didn't know that.

MCP Servers Explained In Plain English

SPEAKER_00

Yeah, yeah. Largest professional services business in the world spawned through the fear that it spreadsheets and the PC were gonna put accountants out of business. So I think when you know, and so the point of my talk was always this changes everything, was my inkling that AI is going to change everything, right? So this is not a you know, this is not a fire drill, this is not a head fake. And a lot of things I talk about, it's been interesting about like the evaporation of middle management. It was a hypothesis. I didn't have agents in my vocabulary when I wrote the book, right? I mean, I was yeah, Claude, what's what's anthropic? I wasn't using anthropic when I wrote the book, it's in Chat GPT, right? So I think when you look at everything, and then you look at like MCP servers, which I'm gonna explain to people because a lot of people there's a lot of people that throw it around. And then there's people that don't understand the power of it, which is it's pretty simple, right? If you take an API, which is how you access data out of software, yeah, and you throw an AI layer on it, yeah, and then you take another API and throw an AI layer on it, now they can talk. Yeah, which means I don't have to match and field match exactly. I don't have to say first name equals F name. Yeah, right, it figures it out, right? It figures out everything in between with limited programming experience, right? So now you know you connect Revit MCP to your Claude account, and you start dropping in some drawings from a previous project. You maybe even take some pictures of a building you like, and you say, design an office building for me. And then it's designing an office building, right? So I think so. The whole point of my book was like, Great, this technology stuff, it's gonna take over, make your life easy in theory, right? Yeah, so now so now what skills do you need to have? And and and that's where I was like, hey, like I'm doubling down on what I wrote in my book, which is you have to be good at building relationships, you have to be a good storyteller, you have to be more human than ever before. You don't get to sit behind your computer typing away to a screen and emailing your emailing your clients anymore. You know, and I think I had a blog post a couple months ago. It's like, you know what I love about AI? Everybody gets to, you know, I'm long, I'm I'm long AI and I'm long golf. Invest in anything golf because everybody's have much more time to play golf, leisurely activities, yeah. And they should be playing golf with their clients, right? Oh, I'm too busy, I've got so much to do. No, let your agents do the work, go play golf with your client. And I think, I mean, I think that's kind of why I wrote that. I had a bunch of people text me, like, dude, like I actually just had you know, during our little break, I just had somebody of one of the largest engineering firms in the world say, Hey, I read your post, let's chat. Because I think a lot of the stuff I wrote in my book, people, it was like, Oh, yeah, great, KP, cool ideas, future.

SPEAKER_03

It was it was early, yeah, yeah, for sure.

Custom Design Gets Cheap Again

SPEAKER_00

But it's all like it's all happening too. I mean, less than two years later, it's all happening. Yeah, and I don't think I thought about the timeline, right? But I think specifically in AEC, where so much of our work is you know, kind of on the computer these days. Now all of a sudden, you know, with some of these automated CAD things, there's another tool called Blender that our team's been playing with. I mean, we're not far from just designing a hospital, right? And some people would say, Oh man, could take away my job. And I'm like, was your job to draw? Or is your job to spend time with customers, understand what their needs were, and think about something unique and different? Right. Impact on society, impact on, you know, do we do all our buildings need to look the same? Probably not. I mean, I mean, that's why you know we we invested in monumental labs, you know, and they do, you know, they basically automate sculpture for buildings, right? People are like, why would you do that? I'm like, because I think the design constraints we've had in society, which is if I build a big white box, it's cheaper. If I design a big glass building, it's cheaper, and the contractor knows how to build it.

SPEAKER_03

I think ornamentation in the previous era, basically before the last couple of years, was not economically didn't make any economic sense. It just was economically infeasible for the last 50 years or so.

SPEAKER_00

And even if it was even if it was feasible, there's a little bit of like, well, how do I draw that in Revit?

SPEAKER_03

Sure, yeah.

SPEAKER_00

How do I if I can't draw it? Right. If I can't draw it in Revit, like do the monumental guys have a Revit plug-in?

SPEAKER_03

You know, it's like if you can't draw it in are you are you gonna draw gargoyles for the front of your building and Revit? Like, no, no one's no one's maybe now with AI, maybe with the MCP server, you will. I mean, monumental is gonna build buildings with gargoyles on the front, and it's gonna be spectacular. Yeah, yeah, because gargoyles are the best, they're the best, yeah. Absolutely. Yeah, okay. Well, sorry, finish your thought.

SPEAKER_00

Just on the so I so I think that's the point, right? So we're if if you if you take my book and say, hey, here's all the things that have happened since then, which wasn't that long ago, and then you look at our industry, I mean it's it's a it's a problem, right? I don't think you know we have a a class of engineers and architects that believe like sitting behind their computer and drawing in Revit is designing, right? They think that running a CRM and a newsletter is customer management, client management. I do think the contractors probably do a better job just because they have to be, you know, in the field, right? They have to be at the project. But at the same time, I think like people want more, you know, they want stories, they want to be engaged. I think one of the things, you know, and people get mad at me about this. Everyone I ever talked to that's going to get their MBA, I try to talk them out for the last several years. I was like, if you want to be middle management in a Fortune 1000, go get your MBA. Because it's a generalist business degree, go get a master's in finance, go go go to Kellogg and get a master's in marketing, right? Go deep in your master's instead of being a generalist business person. And I think those that have heeded my advice, right? Now with AI, like if you're really good at finance and you're an expert at finance, you can push Claude to do things for you.

SPEAKER_01

Yeah.

SPEAKER_00

But you can also do all the other things at a mediocre level. You can do some marketing things at a mediocre level, but at least you're an expert at one thing. So the idea, a friend of mine that worked at Bain, he claims Bain invented the term, which was the expert generalist.

SPEAKER_02

Yeah.

SPEAKER_00

I don't know if they did, but he's he was very, you know, the Bainies are very they're they're away.

SPEAKER_03

I like that term. I never heard it, but I like that.

SPEAKER_00

Yeah. I mean, isn't Claude an expert generalist? Yes, absolutely. Right. So now what? You know, now what? So I think in my book, I talked about a couple things: the death of management consulting, the death of middle management. I have a chapter called Deming is Spinning in His Grave, which is the idea we've created so much routine process and checklists around everything. What that actually does is drive mediocrity. Whereas with AI, we you and I don't have to have a template, right? We don't have to have, you know, think about, you know, I think I had another article, like we're talking about pitch decks. Don't send me 10 slides. Send me a hundred slides, send me the details, send me the entire data room.

SPEAKER_01

Yeah.

SPEAKER_00

Because I can I can now infer from all your data, I can infer what's important to me versus you presenting what's important to you. And I think that's where our industry, where there's been a lot of pride in, I would say, the mechanics of the work versus the relationships of the work, not just the relationship with the client, you know, the relationship with the users. If you're building an elementary school, maybe go spend some time in classrooms, go spend some time with kids. You know, understand what they're what they're going through, what they need, right? And design based on that. Instead, it's like, oh, we don't have time for that. Yeah, we don't have time for that. I gotta go back and do some Revit work. So I think I think that's where that went to. And it's funny, one of the emails I just got was someone saying, Hey, I we, you know, I just talked to our executive team. We need to figure out to spend more time with you. Like, yeah, like your your sub stack is getting sent all over the firm.

SPEAKER_03

So the skills gap to address that point in particular, what do you view as the current skills gap in AUC? Is it relation, is it relationship? Is it spending time on the human side? Like, is it the design, is it the design and the creation, the feasibility analysis? Like what, like where do you see people not spending time and where where do the skills lack?

SPEAKER_00

I think using, I mean, in the world of using AI to build highly custom things, right? We we we suffer from a this is the way we've always done it, and we've always done it, and it's been fine that way, right? And and there's certain things, right? Laws of physics. But you know, for example, you know, we have one of our friends who runs a big engineering firm, and you know, they design steel, reinforced concrete, and timber every time, just to see what the options are, like just to just to see what it looks like, see how it prices out, right? And what some would say is like you can't afford to do experimentation, clients don't pay you to experiment, yeah. Right, clients don't explain explain, you know, pay you to experiment with a new type of material. And my point is like, no, like that's the job. I mean, engineering is not about cut and paste work, architecture is not about cut and paste work, and you know, when you think about our ability to start off with a new a new page, right? Like, well, let's design another bridge. How did we design the last 10 bridges? Well, let's design it that way, yeah, right, which is a lot of how people are thinking about this stuff. So I think there's that, right? The creativity and work to really think about the uniqueness of that, whether it's the customer, whether it's the user community of that asset, whatever it is, right? And to do highly custom things that don't cost anymore. Doesn't cost you creative in a way, creativity doesn't cost you anymore. Right. Then I think it's the relationship things, which is being, you know, I I find it interesting how lawyers and tax accountants have become the trusted resource for a real estate developer. Right? Not their architect, not their engineer.

SPEAKER_03

When you say trusted resource, what do you mean?

SPEAKER_00

Like, hey, I just got, you know, I just met with an investor, they want to fund us building 10 office buildings. You're popping out of that meeting. Who is that real estate developer calling? Their banker, their lawyer, their tax accountant. They're not calling their architect going, hey man, I just had this amazing meeting. We need to get together tonight for dinner and start sketching some shit out. Let's blow their minds, right? Yeah, it's more like guess we have to hire an architect.

SPEAKER_01

Yeah, yeah.

SPEAKER_00

And so, but I think that that mechanics of like who do you call when you're excited about this new project? It's it's the lack of a relationship. Yeah, and and I think those relationships above, you know, we're we're doing a hackathon next week with clients. They're like, If you why do you do this? I'm like, because you know how you build relationships, you can do trust balls, right? You can go build houses as a group, doing habitat for humanity, and those are all fine things to do. My view is like, let's go build product together, let's go bang on some bang on some keys together and do it for a day. And at the end of that day, we're gonna be much better acquainted than we did at the beginning of the day.

SPEAKER_01

Yeah.

SPEAKER_00

Right. And so if you're not in the mix with your clients, you'll never you'll never create enough shared experiences versus the client says, Hey, I just got this, sends you an email, right? Hey, I need you to send me a statement of work. And your response is what? Sending them a statement of work versus saying, Oh my god, this is so exciting. Do you have time for lunch today? Would love to take you out to lunch and celebrate and spend the afternoon talking about their vision and their you know, their thinking and their concerns. And yeah, no, yeah, they sent me an RFP, I'm gonna respond to it.

SPEAKER_03

It's become so uh transactional, right?

SPEAKER_00

But the industry loves to check they'd love to blame the clients for this. I don't think it's true. I don't think you can blame, you know, they're like, well, you know, that's that's how our clients are. They want us to respond to RFPs with the lowest price. Well, you know, I've gone and talked to these clients of theirs, and they're like, we've never talked to them about price. We've never said that.

SPEAKER_03

We've never said that prior they're just like this hidden assumption they're operating on.

unknown

Yeah.

SPEAKER_03

Yeah.

SPEAKER_00

And it and it's then it turns into like this weird urban legend. Everybody's like, oh yeah, like you gotta be the cheapest. There are plenty of engineering firms out there doing work that are not the cheapest on any of their projects, but they have great relationships. You know, I mean, like you know, one of our dear friends, Tom Scarangello with the Ron Thomasetti, Tom is a relationship guy. He spends time with people, he spends time with clients. He is that engineer that if one of the clients has a challenge, they're calling him. You know, they're calling him. I I don't, you know, I don't mean this to be pejorative, but I don't know that Tom's gotten any awards for his structural engineering prowess. I'm sure he would admit there's a lot of people at the firm that are much more technical than him in the weeds, and that's fine, right? But the clients love him.

SPEAKER_01

Yeah.

SPEAKER_00

Clients love him, people love him. Yeah, yeah, right. And he spends time and he asks those questions, and he, you know, he's he he he's great that way. But I, you know, he's very empathetic, he's you know, very outgoing, and I and I think he does care about his clients. And I think a lot of times, you know, if I'm sure he would probably be the you know, his disappointment of not getting a project many times isn't the project. Hey, I didn't get the project, I don't get to work with these really cool people that I like.

SPEAKER_03

It's probably how he probably views things, and so I think which is the right internal scorecard for that line of work, right? It's not just about sending the RFP or getting the invoice and doing this, you know, doing the statement of work. It's no, I just want to work, I want to work with on interesting projects with interesting people. And I think that sort of energy can that is conveyed and it stands out as a differentiator.

SPEAKER_00

And how but how are we how are we training people on this stuff?

SPEAKER_03

I mean, you tell me what engineering firms are going through like intrinsic, you know, intrinsic value relationship building training or sales training. Like I like to me, that that skill set that you're describing that Tom has that's not a sales training thing. Like he's not persuading anyone, he's just genuinely curious and interested in what they're doing.

Judgment Ownership And Titles That Motivate

SPEAKER_00

Yeah, 100%. He's very authentic about it. It's not like a game, right? He's like he's very interested and curious. Another thing, you know, I think I talk about is like human judgment. You know, I think human judgment is a powerful thing. I think humans make great decisions when they have all the information at the right time in front of them. And I think I don't know, I always think about how my brain works around decisioning. Like, how do we think about these things? And I think so many times we in our industry, it's it's it's not a decision, it's like, hey, this is what's comfortable, this is what I'm told. You know, so I think that's another thing that people really um struggle with. I also think like because of the way this industry works, you're not always in proximity of the owner, right? If you're the drywall contractor or the trade, you're working for a sub, the sub's working for a GC, the GC's working for a developer, you know, like you're so far removed that I don't think you really think about like the framing of hey, this is for the owner. And I was talking to a uh a CEO of an uh of an engineer, of an architecture firm actually the other day, and I was like, the number one thing, you guys are worried about getting young people into this industry, the number one thing you can do that doesn't cost you a dime, not a dime, is letting young people have attribution to a project. I don't care if you're the guy that installed the door handles, I don't care if you're the person that printed off the plans, you have the right to say that's my building. Yeah, I worked on that building, I helped design it, that's my building. Yeah, I've I've been infirmed. We're like, don't go around telling people that. That's my building. I'm the I'm the architect of record.

SPEAKER_03

Sure, you didn't earn that yet.

SPEAKER_00

Yeah, you didn't earn that yet. You don't get to go around telling people it's your building. How dare you? And I was like, what does it cost you?

SPEAKER_03

It's like I'm sorry to transition this to VC, but you know, famously, Andrees and Horowitz, who's arguably the most famous venture firm out there, they give everyone who starts new at the firm a partner title.

SPEAKER_00

Oh, really?

SPEAKER_03

Yeah, yeah, it's at least every it's like everyone or like the majority 50% a partner title. And the reason they do that when most firms run the exact opposite way, like you have there's a hierarchy, you have to earn your way up, you know, there's like a real, yeah, there's a real hierarchy there. Is because it doesn't cost anything to give someone a partner title, and yet that is what they want most. Yes. What does every junior VC want? They want to be, they want to be called partner. This is true of you know of of our of our industry too, and you know, young engineers, what do they want? Right? They want to be part of the partnership. And it's just an interesting, like, I'm not saying that everyone should go out and you know, and and bring the new hires on as partners, but like the reframe of what what is cheap that actually gives the the employees something they deeply crave and want, and oftentimes that's a title at the beginning, and yet it also gives them like more ownership in what they're doing, like it solves for the skin in the game issue. If you're a partner at the firm, you better produce, right? Or you're gonna get cut. Like, of course, you got to make your ends meet. Yeah, and so I actually think it's really smart, and I think there's something to that, to your point.

SPEAKER_00

Have you seen this trend in startup world, like founding SDR, founding account exec?

SPEAKER_03

Yeah, yeah, yeah. Founding engineer, yeah.

SPEAKER_00

I found it really were really weird at first, but then I kind of applied the same thinking. I'm like, it doesn't cost you anything, totally, right?

SPEAKER_03

And maybe that's what does that customer success person want most? They want to be a founding member of the team, yeah.

SPEAKER_00

I'm the founding customer success manager, right? It doesn't mean you're the founder, that's not what it means. Yeah, but I just all you know, within a period, a window of time, I saw it popping up all over LinkedIn. Founding STR, founding account executive. You know, I was like, wow, this is like kind of interesting. I thought it was weird. I'm not gonna lie. I was like, this is weird.

SPEAKER_03

It's a little, it's a little gimmicky. I mean, I yeah, I think it I think it's gimmicky, but the here, but here's the thing you're not solving for the public perception, you're solving for the internal scorecard of the person you're hiring. That's what they care about.

SPEAKER_00

Yeah, I can't pay you as much because we haven't gotten our series A, but you get to be the founding account executive. Totally. I'm in, I'm in, done. Right?

SPEAKER_03

Yeah, it communicates a different level of stature if you were founding, you know, if you were the first, if you were the first hire on that particular team, right? It's like a it's like a it's a way to communicate that you were in from the very beginning. And like if you're just if you just have that account exec title, the SDR title doesn't communicate that.

The Bar Is Still Great Work

SPEAKER_00

Yeah. No, I I think look, I think it's like super interesting. And and I think if you look at the hierarchy, I remember when I was in engineering many moons ago, I started off as a staff engineer one, and then I was a staff engineer two, and then I was a staff engineer three, and then I was a project engineer one, right? We had pay bans, and it was all in an employee handbook. And in my second year as an engineer, I drove more business than any single engineer in the firm. And my boss looked at me and I'm like, hey, do I get a promotion? He's like, I can't. You haven't been here long enough. I'm like, what does that have to do with anything? What is you know, what did my years of service have to do with that I'm crushing everyone around me and putting money in the bank for the company for the company? And they're like, no, that's that's how that works. That's how that works. And I and I think that's a lot of this stuff that you know, because you know, I made a a point on our internal Slack the other day was that if you're gonna use Claude, you know, the bar is still greatness, right? You have to deliver great. Yeah, and you should not use Claude if you don't know what great looks like.

unknown

Right?

SPEAKER_00

You should not use Claude to do your work if you do not know what the great deliverable looks like in order to iterate it to get it to great, right? So it should never be a tool that I've never delivered great. I'm gonna use the tool that way. I should only understand what great looks like. Now, if you don't know what great looks like, you should use it as a training tool. Like, what does the how should this be? What should the so it's your training partner, yeah, to understand how to get to great, right?

SPEAKER_02

Totally.

SPEAKER_00

And and I think that's one of those things where everybody wants to use it as you know, it's a hammer looking for now. I need it, I need a deck. I mean, I've been seeing some really beautiful decks that say nothing, yeah. That say nothing. It's like, oh, you think because it's pretty, it's meaningful, it's pretty and dumb, it's horrible, right? Because you miss the point, right? And I mean, I'm seeing so many founder decks these days. You can they all look the same, right? You can tell because they all we all we know what it looks like. And the math doesn't math, right? None of it works, none of it's and then you ask them 10 questions, you know, because you know it is when startups pitchy, we play stump the chump. What about this? What about that? What about this? Help me understand that. They have no answers. And so I think it's just as important to use these AI tools as a as a thought partner to learn things, not just to crank out deliverables, right? Yeah, so I kind of said like if you don't know what great is, you shouldn't be creating deliverables.

SPEAKER_03

And I think there's a way to understand if someone knows if if what they're building towards is great, they know what greatness looks like, have them send you an example. Yeah.

unknown

Yep.

Data You Share Becomes Lost Advantage

SPEAKER_03

Whatever you're building towards, send me an example that you're aiming for. I find that's like a pretty good, yeah, it's a pretty good test. Like if they can't, they send you a couple website examples on you know on the design side, or they send you, you know, a couple models that they've seen in the past that are that are good and they're mediocre. I think it you know tell tells you all you need to know. Okay, cool. You up for moving on to your next post? Sure. The data you give away is the advantage you lose. I think this is an interesting concept because obviously the incentive right now is to give away a ton of data to lots of different parties. So, yeah, walk us through this post.

SPEAKER_00

Yeah, so a lot of it, I think, you know, so we've seen a regression, you know, in my little sample size, right? Where people are now not using Claude. Companies have disconnected it, companies have banned it. I think some of it came from the glass wings stuff, like people not understanding about you know, mythos and all that stuff, right? I think some of it was alarmist due to that. And then a lot of people, you know, it's it's pretty interesting. You know, I talk about like in our world, there is no such thing as the enterprise, the project is the enterprise. And what that means is everything we do is project-based. And if I'm the user, let's say you're this civil engineer working on the new Tesla factory, and Tesla says, Hey, I don't want you sharing any of our data with the LLMs or with your internal LLM or your SLM or whatnot, you don't know, right? Because maybe you connected Claude to your enterprise email, maybe it's connected, like you have no control over it. And I think in a world where we are all in AC, right? We're just so fixated on IP for good reason, right? For good reason, then we're not really, we don't have control, right? Because whatever corporate system we have, whatever cloud's doing, people don't know what cloud's doing, people don't know where their data is, and so I think what you're starting to see is people are saying, like, well, I don't want to share that with open AI because I don't know where it goes, I don't know how they're using it. Um, I don't want to, I mean, I always kind of, you know, I kind of regress back to like early days of spreadsheets. There were the spreadsheets on the P drive and like the public drive on your or the project drive at your company, and those were all the standard pricing spreadsheets and stuff. And then there were the spreadsheets you kept on your personal drive on your hard drive that you didn't share, right? Because it's like that's my IP, I'm not sharing it with everybody, they'll just screw it up, they'll screw it up and then play me, right? And so I think we're very much that kind of industry where if you're the project engineer on a big construction job, it's not the company's project, it's not the company's client, it's your client, it's your project. And I think people are realizing that they don't know where all this data is going and what the value is, and therefore, like in many ways, they might feel like they're putting themselves at risk. So I think there's been a series of events, but generally I think people are kind of hitting pause now. I know it's a 3,000-person company that just banned in our industry that banned Claudius, and so I asked the the VP there, I'm like, so what does that mean? And he's like, It means we're all carrying around two laptops or work whatever personal. Wow, I was like, really? He's like, Yeah.

SPEAKER_03

If you're the if you're that concerned about it, why don't you just set up on-prem? Because they don't know how to.

SPEAKER_00

I mean, yeah, there's like but you know Nick, I run an AE firms these days, they're still running an exchange server on-prem because they couldn't figure like they're still trying to figure out how to get from on-prem to the clouds.

SPEAKER_03

It's perfect time to just jump straight back to on-prem.

SPEAKER_00

It's like belt bottoms.

SPEAKER_03

Add some GPUs to the to the racks.

SPEAKER_00

Yeah, it's like fashion, belt bottoms will be back. Just give it time.

SPEAKER_01

That's hilarious.

SPEAKER_00

I mean, think about it, they're like, we never trusted the cloud, so we stayed, we kept everything on-prem. I'm like, are you kidding me? Like, how often does your email go down? I mean, I mean, literally, they were telling me that sometimes their email goes, the email server goes down all weekend. And I'm like, what? Like, what do you like? What do you do? That's insane.

SPEAKER_03

It's crazy.

SPEAKER_00

Yeah, it's very crazy. Yeah.

SPEAKER_03

By the way, yeah, it's an interesting post, and I think that the the data, like that data question will, you know, everyone will have have to address. And I think there's a lot of you know, looking in the mirror you have to do about where you're headed as a firm. And you know, we've talked a lot about how to protect your IP. There's a lot of leverage you can get from AI with your IP, as long as it, you know, the privacy and everything is is secure. And then yeah, you can always go on-prem if you're a security nut.

SPEAKER_00

Yeah. And I it was interesting, I was reading something about like law firm law firms, right? If they're if they connect their email into an LLM, they're technically in violation of attorney client privilege.

SPEAKER_01

Yeah.

SPEAKER_00

But I'm also like, but for my stuff, if you're gonna spend less hours, please use it. Like, I don't care.

SPEAKER_03

Yeah, I have a friend who's a lawyer and they block it on his his when he opens his email, like you can't access it. Fascinating.

SPEAKER_00

Not everybody has a not everybody has a burner laptop. 100%, which is even worse.

SPEAKER_03

Good good thing. It's just easy to copy and paste text from an LLM that's a text-based model.

SPEAKER_00

Oh yeah. So I mean I look, I I think this is something the industry will struggle with. I think there's the the other problem is there's plenty of people that are kind of anti all of this, you know. They don't, you know, they're okay with their employees going out and getting their own chat GPT account because they don't have to pay for it. I mean, I have one group, 2,000-person firm. They're like, it's like you guys need to move to enterprise cloud, then you can have controls, then your corpus of data stays within your firm, it's not staying in people's personal accounts, blah blah blah. They're like, KP, we looked at it, it's$200 a month per employee. We have 2,000 employees. Wow, that's insane. We can't spend that kind of money. I'm like, okay, don't I'm not here to force you into something.

SPEAKER_03

But force for the trees, force for the trees.

SPEAKER_00

Yeah.

Zero RFI Keynote And Risk Going To Zero

SPEAKER_03

So I think a good way to uh close would be to ask you what the most interesting responses you got from the building transformations talk that you did. Because I imagine they were all across the spectrum, and probably just give people a preview for what that event is.

SPEAKER_00

Yeah, so building transformations, it used to be called Can BIM, which is the Canadian BIM thing. It's a Canadian organization. It kind of, I think during the pandemic, like a lot of these orgs struggled, but so it's been really cool. Like I give a lot of credit to a lot of the GCs. So Jitkey at Suffolk, Ty over at Henself Phelps, HANU over at DPR, the guys at Alistan, all of them kind of rebooted it. And and what I found really interesting is so it's a an organized nonprofit for GCs, by GCs. Yeah. So instead of like other for-profit organizations monetizing, uh, monetizing their talks and panels, they kind of keep it in-house. So it was pretty nice. So I gave the keynote at that event around Zero RFI. So it was really the first time I have given like an industry talk on what we're doing at Zero RFI. It was pretty cool because there was like really good clarity on what I'm doing and why, and why the tech matters and like how it matters. And afterwards, I had everybody come up to me and say, Hey, come, can you come meet with my CEO? Can you come meet with my whoever? We want to understand how we can work with you. Because I think, you know, the best way to describe I think what we have built because we're delivering as a services organization, there's almost a correlation between the impact of these of the tech we've built and the lack of TAM, meaning construction management software, massive TAM. I call it the major muscle tissue. It's a big muscle, it's a major muscle. We do not need to build that, right? Design tools, Autodesk, Bentley, they've got that figured out. Major muscle group, right? What we have focused on is the connective tissue, that there is no TAM, but is likely the most critical aspect of where things break. Right. So you're gonna tear up your rotator cuff, but you're not gonna hurt your chest and your bicep, right? So it's it's this inner this interconnective stuff that's really the stuff that is most critical. That I don't care how good your major muscle groups are. Doesn't matter. Doesn't matter. Right. And so, and the problem with these things is that there's no TAM. So, as VCs, we would never fund this stuff. But as an owner's rep, it's so critical. And we've really hacked that stuff. So, a lot of the feedback I got was like, man, like I had an investor yesterday, I pitched, and they said something that was very nice to me. They said, of all the people in the construction industry in the tech industry, you may have the highest right to win of anyone we've ever met. And I was like, absolutely. That was very, very nice. In VC Lingo, that's like a compliment. Totally. Yeah. You know. So no, it was a really good talk. It was pretty interesting, too. I listened to some of the panels, and they were talking about citizen coders and how they have a lot of citizen coders, and now they're struggling to let them keep coding. Right back to the kind of our conversation about privacy and this and that. There seems to be a lot of fears that is kind of driving like regression of adoption of these AI of AI systems. And we're definitely seeing on some of the startups, too. Some of the startups that came out hot and heavy, AI, whatever. A lot of these firms that are die hard incumbent software companies, they're kicking out the AI tech startup, and they're just saying, like, you know, ProPore's AI tools, good enough. Yeah, it does it. I don't, I don't, I don't need the risk. Like, at least I know what ProPore is doing. They're a public company. Oh, so funny, yesterday I posted something. There was an analyst from Jeffries, which I found fascinating. He said that Micro, he's like, I don't know why people talk about Microsoft Copilot sucking so much, his words, when they just turned around and sold 700,000 licenses to Accenture. It can't suck that much. And I was like, is that true? Like, is that true that it can't suck that much? Because Accenture bought a bunch of it. You know, I don't see any startups running on Copilot, right? So I think we're I think what's happening is people kind of gone to the edge, they're not sure what's down there, and so they're pulling back and saying, you know, maybe copilot is good enough, right? Maybe at least it's safe, right? So people are, I think I think they're regressing to their wherever they feel safe.

SPEAKER_03

Interesting. Yeah, yeah. Was the indust industry reception at that event, given that you know it's mostly AUC people, not necessarily the people you're selling directly to, was their feedback was there a level of threat that they felt from you? Was it obvious how they were going to partner with you? Did they have like a ton of questions on yeah, just like where are you going long term and how it how it relates to their business? Like what was that dynamic like?

SPEAKER_00

Yeah, I mean, look, I I think I think what's interesting is these are all like stalwarts of of the industry, right? And so yeah, and they're friends, right? Like they're you know, Ty and Hensel Phelps is not worried about waking up tomorrow and me putting them out of business, right? Like so I so I think it's a balanced view. I think the one thing that someone said that I thought was pretty interesting that was somewhat fearful, they're like, as a as a general contractor, our job is to manage, mitigate, and transfer risk. That is our job, right? To see risk coming, mitigate against it. Sometimes we have to push that risk down to the subs, that's the job. And they were like, KP, based on what I just saw, you might be able to take the risk down to zero or close to, which means why does an owner need to buy a GC, hire a GC? If you're if if we're the insurance policy of risk and your tools from what we saw, it looks like you have the opportunity to drive it down to zero. Then why am I buying insurance if there's no risk?

SPEAKER_01

Fascinating. Yeah, fascinating.

SPEAKER_00

Yeah, and then someone also brought up on if you know what owner OSIPs are, those are basically like owner-controlled insurance policies. Okay, so you as an owner, instead of your G senior subs having insurance, you can actually self-insure.

SPEAKER_01

Okay.

SPEAKER_00

And the main reason people don't do it is it's like a lot of paperwork, it's a lot of pain. But you know, so someone also brought up like, well, KP, if you're working with an owner, might as well do that. Like, aren't you essentially the GC? I don't think so. Not near term, near term. But I I think you know, but I also feel like look, you know, you have a bunch of senior people, they're like, Yeah, like I'll probably be retired by the time KP, you know, that my retirement date and KP putting us out of business might be the same date, you know, kind of thing. So I I what I love about that group is they they are a little bit purist, they're they're not the business people, they're not trying to figure out, you know, they're they're they're that's not their job. It's really just kind of more of a pure intellectual technical view of the world.

SPEAKER_03

Okay, yeah, very cool. Yeah, I happened to catch most of your talk. I came in like five minutes after you started, but it's great, thought it was good, very clear, clear vision. And uh it was the first time I heard you give a talk on zero, so that was compelling.

SPEAKER_00

Yeah, I got some bumps up. We recorded it, I've I've rewatched it, got some things to clean up a little bit. I think for that, I I think for it landed well for that audience. Yeah.

SPEAKER_03

I think the I mean one thing that came came through, and I think you always do a pretty good job of this, is the disruptive potential was communicated well, of just like, hey, here's all this massive technology change that's happening in the background of the AC industry. Here's what's like here's my vision for what's likely to happen, just you know, with all of that in the background. And here's the most efficient way to effect the most change for the build environment. And I'm gonna build a company around that idea and shadow that thesis, too, right? But I think that most people haven't put all of those pieces together. Like, you know, you have the AI conversation, you have the robotics conversation, no one's talking about, no one's going, you know, to the level of depth of like how do the business model changes as a result? What is risk transfer? How are we ensuring the project? How are we using inference? Are we doing on-prem setup? Like it's so many questions to play out that full game theory. And I think what you did really well was like you actually played it out for people. And for some people, it was like, holy shit. And for others, it was okay, like tell me more. I'm I'm I'm intrigued. And I don't think everyone in the audience got exactly where you're headed, but I think they saw that you've put more pieces together than than anyone else cohesively with zero. And I think that that was like one of my takeaways.

SPEAKER_01

Yeah.

SPEAKER_00

Yep, you know, vantage point helps, right? All those folks leave that conference and they got to get back to work and run a project and be heads down, right? Uh totally. We I mean, even in investing, right? We have the benefit of having a great strategic vantage point. And not in the mix, you know, we get to see we get to see the we get to see the whole test work, right? They just kind of get to see their next bit.

SPEAKER_01

Yep. Let's go.

SPEAKER_00

All right, man.