KP Unpacked

Software Is Getting Hard to Invest In

• 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 | 50:41

What happens when non-technical people at a hackathon build in one day what software startups have been pitching for years?

In this episode of KP Unpacked, KP Reddy and Nick unpack why sitting in a hackathon full of non-technical AEC people building working prototypes in eight hours is making software feel uninvestable. Incumbents are building features they've wanted for five years. Nobody needed corporate approval. Nobody needed a startup. They just needed a day and a keyboard. If that's the new baseline, what exactly is a software company selling?

The conversation covers why PE firms are overpaying for AEC companies at 14x EBITDA and losing their best people 18 months after close (almost clockwork), why Procore's AI agents announcement landed with a thud, why vibe coding is a rabbit hole that creates individual value but rarely scales to the company, and why token pricing is heading toward the same bundled unlimited model as internet bandwidth in the 90s. KP also reveals his new LinkedIn rule: send a 10-page resume, not a one-page highlight reel. If you control what I see, I can't assess what matters. And a hospital system owner asked Zero RFI to build a real-time team qualification tool because they're tired of getting the B team swapped in mid-project without notice.

Key questions answered:

  • Why is software getting harder to invest in after every hackathon?
  • What did non-technical AEC people build in one day that startups have been pitching for years?
  • Why are PE firms losing their best AEC people 18 months after acquisition, almost to the day?
  • What's wrong with paying 14x EBITDA for a firm that doesn't grow like that?
  • Why did Procore's AI agents announcement land with nobody caring?
  • Is vibe coding a rabbit hole or a real productivity tool?
  • Why should you send a 10-page resume instead of a one-page highlight reel?
  • How is KP cross-referencing resumes against his LinkedIn connections using skills files?
  • Why do hospital owners want real-time team qualification tools mid-project?
  • What's the AI diffusion crossing the chasm moment for the broader economy?
  • Why is the Bay Area network effect getting stronger, not weaker?
  • Why are tech company balance sheets suddenly going CapEx heavy for the first time ever?

If you're building software for AEC and wondering why incumbents are building your features in-house, a PE firm trying to understand why cultural fit isn't transferring post-acquisition, or a founder debating whether to raise venture or bootstrap, this episode will force you to ask whether the software playbook still applies when anyone can build anything in a day.

Listen now.

Tech Bubbles And AI Diffusion

SPEAKER_00

Alright. It's the most interesting topic to get open. It's the best cold open topic for us. Um Adam? Adam is pretty pretty relevant. Uh I wonder how many of our l what percentage of our listeners know about you guys. Very small. I agree. Very small.

SPEAKER_02

I mean, I think we there's a couple things that happen. I think one, we kind of live in a bubble. Um, and then our social media tends to reinforce that bubble. I mean just seems that way. Yeah.

SPEAKER_00

I mean the social this the social media bubble is essentially the algorithm feed that me and you are just obviously on the same one because we're interested in the same thing. But I think it's like I I feel like most people that are in and around tech and business, you probably probably see the same 500 key tweets a day.

SPEAKER_02

Like that's just what the algorithm serves you. Also, I found out from some friends or like outside of like New York and the Bay Area, if you walk into a coffee shop and ask people like what AI tool they're using, yeah, they may not have an answer. Likely true.

SPEAKER_00

You know, the the diffusion the diffusion to most of the rest of the working class outside of tech is just hasn't it hasn't happened yet. So uh which is an interesting question. Like, when do you when do you think that really when what's the crossing the chasm moment for diffusion for most of the economy?

SPEAKER_02

I think devices, right? I mean, I think where Apple's kind of messed up. You know, the thing with Apple is they're they're not innovators. They have to like they're coming out with their flip phone, their foldable phone. That's been around from other providers for a while. I think they let other people prove the market because they have such great distribution, right? Distribution's amazing, stickiness is amazing. They don't have to be first mover, they can wait for everyone else to iterate their products, and then they can decide whether people care about it and what the form factor should be, and then they can mimic it and push it through distribution. It's the same knock with Siri, that Siri is horrible, right? Which it is. They'll get there eventually. So I think the minute it is on the device and people can use it, and it's you know, it's Siri, it's not Chat GPT or whatever. Yeah, I think I think that's when you have consumption.

SPEAKER_00

They just have not been good at software.

unknown

Yeah.

SPEAKER_02

Pretty good at hardware. Yeah. Not good at software. Well, good enough. Right, good enough. I don't think we know I I tend to listen to Asian news a lot of times, Asian business news, because they just have a much better lens. They talk a lot of a lot about hardware, a lot about materials, right? Batteries and that's been their game for a long time. Yeah. So when you listen to the Asian business news, pretty much all they're talking about is hardware. And there was a conversation yesterday I was listening to where they said it's pretty interesting, Samsung has a foldable.

SPEAKER_03

Yeah.

SPEAKER_02

But Samsung electronics displays, Samsung displays actually makes those displays, and they're likely the supplier to Apple. So like Samsung. I mean, Samsung's just winning in general.

SPEAKER_03

Yeah.

SPEAKER_02

They're winning the memory game, yeah, they're winning the battery game, they're winning a lot of games out there. They're doing well. No, I think that's when he when we think about when does AI really start to affect people's lives. I think that's when it happens. I was at this thing yesterday, and a well-known journalist was there, and he gave kind of a keynote, and he said that he just got back from China. And China surveyed their people about a simple question: how do you think AI is going to affect your future, positively or negatively? 85% of the Chinese population said positively. That same survey was done in the US, and they said 70% of the US population thinks AI has a negative effect. So I think also like you can sit here, it's kind of like social media, right? We'll sit here and complain. Oh, our kids didn't get off social media while you're scrolling through Facebook. Right? So, how many adults have gotten off of Facebook or Instagram or anything else to prove to their kids it's a bad idea? Not that many. Not many. Yeah. Not many. So I think there's an idea like, I don't know if people are boycotting AI or they're just trying to stay disinterested in it. But if there really is that much of a negative outlook, you know, will people not do it on principle? Kind of like the people that like drive Teslas and they have the sticker. I bought this before Elon.

SPEAKER_00

I don't like Elon.

SPEAKER_02

Yeah, I got this before Elon went crazy or something. So how far are you gonna boycott? Yeah, you know, when when something has utility, I don't so I don't know how like how long you can boycott.

SPEAKER_00

I mean yeah, in some what in some ways this is just like the age-old question of how does a a Luddite interact with the world and like how long will they avoid the best tools simply out of principle? Yeah, and out of some, you know, maybe a philosophical or political stance. But at the end of the day, like it has a tool.

unknown

Yeah.

SPEAKER_00

I was talking to my sister-in-law last night who's staying with us in the city, and which is a whole other topic. There uh, they are my sister-in-law, my brother-in-law, and their two kids, and we uh thought it would be a good idea to have them all stay with us in our rented house here. And they have a five five-year-old and a one-year-old, so it's been fun. Oh, that's been good. But anyway, we have these like fun kitchen conversations in between our kids yelling at each other. And she is looking, she was looking at buying a house where they live in Colorado, and she knew nothing about seller financing.

The Moment AI Becomes Personal

SPEAKER_00

She's like trying to, she's trying to put together a creative financing way for them to purchase a house that's a little out of budget. And she went down this path of seller financing. She knew nothing about it. And she's like, I spent two hours of my kids fell asleep and I had to stay in bed with them for a bit. And I just went down a rabbit hole in Chat GPT. And she's not like, you know, a power user of AI, right? But I think like the more you get exposed to, oh wow, I went from zero to one on, you know, a pretty obscure topic like seller financing, right? I mean, you're probably I mean, even with Google, you're probably still combing through like what quality and non-quality answers look like. It just it's a it's expensive from a time standpoint to get up to speed. I mean, I talked to her this morning, she's like near like the back of her hand. Yeah. And so those types of experiences where you are leveraging intelligence in ways that affect your personal life pretty significantly, like buying a house, pretty significant personal experience. I think it's gonna like like you don't come back from that experience and not use the tool again.

SPEAKER_02

Yeah.

unknown

Yeah.

SPEAKER_02

I mean, I think it's interesting, but I but I think there's always that early, like early days of internet, the most ongoing daily use of email was people like mass emailing jokes. I would get jokes from people like, what are you doing? Yeah, I think we're using it for work, then we're sharing attachments, like any of that stuff, right? It was forwarding jokes, forwarding jokes, yeah, forwarding jokes around, right? So I so I think there, you know, I think there's, you know, I'm trying to always size up people when you know, because how do you tell people ask people like, hey, like, how good are you at AI? Like, what's your proficiency, or where you where are you? You know, people don't have any sense of like, hey, I'm at a third grade level of AI or whatever, right? People don't know. I usually ask people simple things like, hey, what are your what are the your three favorite skills files you developed? And when they go, what's that? Then I kind of know. Yeah, right. You're using it to create your recipe, you're using it as a glorify, a better version of Google that can help you, you know, find things. Yeah, right. It's a certain search aspect. Yeah. But but also I think fascinating enough, I mean, Google's numbers came out yesterday, crushed. I think what they're projected was $2 your EPS. They ended up at nine. Something like that. It was like wild, right?

SPEAKER_00

It was interesting. The stat that I saw circulating was first negative free cash flow quarter in five years, maybe longer. Probably longer, probably a decade.

SPEAKER_03

Yeah.

SPEAKER_00

It's just like Google prints money. Yeah, they have free cash flow, and they didn't, you know, they're spending so much on CapEx that finally have a negative free cash flow print. But yeah, the actual like

How To Gauge AI Proficiency

SPEAKER_00

the EPS was super strong.

SPEAKER_02

They also own like five to seven percent of both OpenAI and Anthropic and double percent of SpaceX, of SpaceX and Waymo. Yeah, so I was I was talking to this guy yesterday, and this is a little, we hadn't really talked about it, but I was talking to this investment banker yesterday. And I was like, look, you know, seven years ago, if you were a tech company, how did you use your balance sheet? And you mostly use your balance sheet to manage cash flow, which most of them reach a point where they're heavily cash flow producing, so that. And then really RD, right? Outside of that, how are you if you were at Facebook, what were you using the balance sheet for? The only utility it had was to fund internal things. And then they some of them got into venture, right? They're investing off the cash flow. Salesforce ventures has done pretty good, right? Not bad. Probably better than Salesforce has, quite honestly. So that's the only mechanism for leveraging for doing anything with the balance sheet. But now you're unlocking these massive balance sheets in tech that are actually going into cap heavy capex expenditures, data centers, chips, whatever, whatever, which is pretty fascinating. And now they're even taking on debt. So if you think about tech companies that capital light, capital efficient, all these things, they've now kind of gone to the dark side of like they are leveraging their balance sheet. Which I mean, when did they ever leverage their balance sheet? Maybe for some financial engineering or something.

SPEAKER_00

But their balance sheets were so large that what would you actually spend money on? Is the question. Yeah.

unknown

Yeah.

SPEAKER_00

Like you can't do a hundred billion dollar acquisition every day.

SPEAKER_02

So it's it's pretty fascinating. I mean, so when you think about like capital flows, a lot of capital flows is unlocking capital that's been maybe stagnant, you know, sitting in treasuries or sitting wherever, right? Which is, you know, probably where some of it was sitting. You know, where was Facebook storing all their cash 10 years ago? Probably just sitting in treasuries. Yeah. And then of course he's that's that's why he goes off and does metaverse, right? Right. Gotta find something somewhere to put it, seriously. And tech companies don't like doing dividends. If you think about that, like it's anti-growth. Yeah, like when have they ever done dividends?

SPEAKER_00

Yeah, that's like anti everything.

SPEAKER_02

Yeah. So I think it's interesting.

SPEAKER_00

So the one investing thread that I think is interesting on that topic is well, one, you reach a point. So like cap the capital markets reach a point in a cycle where you know they squeeze all they can out of a potenti out of a potential trade, right? You can't get 20% anymore on the mag on the mag 7 without really interesting future growth prospects. Like they've already acquired all the customers, right? Yep. So how do you grow? So there's this really interesting thing that happened where there's like an opening for large amounts of capital. I'm talking like billion like tens of billions of dollars to

Big Tech Goes CapEx Heavy

SPEAKER_00

go reinvest in in winners that are super blue chip and super familiar, that have strong balance sheets, and accelerate growth, which is through AI and through other business models and you know, horizontal versus vertical, going into robotics, metaverse, whatever it might be. And so there's a place for capital to go now. And that's that's been the entire trade that you know that we've seen in the AI CapEx sphere. Yeah, because there is a new way for them to acquire scale, like increased scale and and and increased, you know, monetization of their of their users for the first time in a long time. Like basically since the you know, since the web 2.0 era where they all were first acquiring their users, they can now get massive growth and scale.

SPEAKER_02

Yeah, and it's what what's really interesting, like the psychology of markets, right? A lot of it has to do with you know investing now, anticipating growth, right? So you might overpay. So some of the the math is built in, so to speak. But I found fascinating, I Blackstone is down, I think like 20 something percent on the year. They just sold QTS, the data center business they had bought for a 7x return, like maybe over two years, massive return. Stocks down. Because I think the psychology is if you're taking chips off the table, what are you gonna do with those chips? Yeah, right. And so in their mind, they're like, hey, like book the profit, right? Seven X. Take them, yeah, yeah, take the win and move on and do the next thing. But I think you're not getting rewarded in the public markets. And by the way, none of this is stock advice, whatever disclaimer that I'm supposed to say.

SPEAKER_00

But I just don't think you're in research, not financial advice.

SPEAKER_02

I think I mean, I think it's the psychology right now is you're not gonna get rewarded for anything other than growth moves, you know. So taking chips off the table with QTS, and I think they had a couple, an energy business that they sold off. Yeah. Everybody's like, why? Like, why, why we all know energy is a problem. Why are you taking chips off the table? Their response is like, hey, we want to distribute cash back to our investors, we want to, we want to book the win. Right? So I I think it's interesting, you know. You would think Blackstone should just be rocking it. But the question now is this they're gonna issue dividends and liquidity to all their investors. Do those investors come back to them, or does that cash now go somewhere else? Like, there's no guarantee the cash rolls back to them. Sure. Then they go and do uh something with Blue Owl or whoever, right? It's maybe more edgy. Yeah, but I thought but I think that's one of the things that you really as investors, it's always like well, even at a personal investment level. My brother's been in NVIDIA for like eight years. Like he just bought, I think he bought it because his son was in the game. He's like, Oh, observationally, right? Like, oh, all his graphic cards are NVIDIA or whatever.

SPEAKER_00

And a little bit he's just like, I'm so up on it, but like sell and do what still feels like there's room to run. I mean, compared to what?

SPEAKER_02

Yeah, right? Sell it and put it in into where, right? And I think that that's where I think even at a personal investment perspective, you know, unless you do it professionally, why why sit here and make trades? Yeah. Now, unless you're trading, why would you sit here and make trades? Still a good company, still good stock. Same thing with Google. I mean, Google's been crushing it.

SPEAKER_00

There's the uh this Peter Thiel commentary on so he was an early investor in Facebook on the board for over a decade, and they IPO'd, and I think he shortly thereafter sold most of his position after they IPO'd, which was a massive return, like thousand plus X, right?

SPEAKER_02

By the way, the rumor of that is he invested through his IRA, yeah, yeah, exactly. Which is like it's like the not non-taxable event.

SPEAKER_00

The biggest IRA event in history. Non-taxable event. Incredible. And then so yeah, and so looking back when he was reflecting on the sale of that of his position, and you look at all of his other major investments during that period of time, he would have been better off doing absolutely nothing, just sitting in Facebook, yeah, then make then starting 10 different funds, then making any new startup investments, but which by the way have been successful. Right. It's not like he's he's losing it. He's not doing poorly, you know, in investing, whether it's through founders fund or any other major trade he's in, but it was just like the the concentration of of capital and you know the the compounded growth that they've had since that period of time, it would have like dwarfed everything else he could have done. And so his like mistake is you know, I like you you you have to ride your winners longer than you typically think.

SPEAKER_02

Yeah, but I think also like I don't think that's why people do it. I mean, at some point he's got plenty of money, right? So what the point you're saying is like he didn't get he could have been extra rich versus just really rich, right? And I don't know that these guys spend in a way. I mean it's pretty interesting in the Bay Area, like people don't spend. I don't know if it's because they work all the time, which people do work really hard. You know, very few vacations. I and I but I don't I think it's like they just don't like to show off like their money, right? It's like I just company just went public, I'm gonna go buy a new Prius kind of thing. It's not I'm gonna go buy a Ferrari or anything like that. I mean I'm sure there's some of that. But but I think for people like him, it's like what can I go do next? You know, in the world of like money doesn't make you happy, but if you can put money into companies like SpaceX, etc., I'm sure he gets some joy out of that versus just camping out on Facebook. Yeah. So I don't I don't know that all of it's a hundred percent about money. Sure. Still doing fun. There's the still doing okay.

SPEAKER_00

I think like a more difficult, like a really difficult question as an investor is is even if you had success, your next investment, there's there's a question of like, do you want to do you want to be clever and smart or do you want to make money? And I think it's it's a really simple question that is probably difficult to answer for most investors because after you do really well to your point, like what's your goal? Yeah, do you really care about making more money if you're worth $100 billion?

unknown

Yeah, probably not.

SPEAKER_00

Do you want to do, you know, do you do you want to go down a philosophical bent and start a lecture series on the Antichrist? Yeah, probably right. Who wouldn't want to do that? So yeah, anyway, but but I think like my point is if you want to make money, yeah, just objectively, like you know, not knowing anyone's personal finance position, sometimes letting your your like the compounding winners ride is just the best move, especially in tech.

SPEAKER_02

Yeah, no, I mean I think there's a lot of facts around that stuff. I think if you look at like Gersner at Altimeter, I mean Sequoia. I mean, if you think about Sequoia, like a lot of times they did not, they just kept their shares in companies, they never looked for liquidity. They they distributed shares to their LPs, right? And they kept their shares, and LPs could do whatever they wanted with it. So no, I think that like I think some of this stuff is super interesting right now. But I do think getting back to some of our kind of core topics, I thought it was interesting. Procore just announced this morning they're AI agents, and like nobody cares, nobody cares other than them. It's pretty interesting. Like, I think there is this weird like AI backlash. It's almost like, you know, like we always say, like, remember when Uber rides were five dollars, right? And I think right now you're getting so much token subsidies. So I was

Hold Winners Or Get Clever

SPEAKER_02

talking to someone this morning, they're like, Yeah, like they have AI agents in Procore. I'm like, how do they charge for that? Like, oh no, it's included. So they're paying for the tokens, right? So they're like, Yeah, like you know, we spent five thousand dollars on tokens last map last month at our company, and so we're gonna be highly encouraging people to instead of asking Claude, ask Procorp because it's like a free resource. Sure. So that's gonna be super interesting because that that's not sustainable, right? It's not sustainable if they're not gonna try charge for tokens or any of it, right? You have to have some recoup.

SPEAKER_00

Did they raise their pricing at all, or is it just purely?

SPEAKER_02

No, I think they're still doing the same.

SPEAKER_00

I mean, it's too early to say.

SPEAKER_02

I mean, I think eventually they do, right?

SPEAKER_00

I was talking to a startup this week that's that's including the anticipated token usage in the pricing of the product. And so it looks and feels like a similar buying experience and price experience to what a SaaS product would look like, but they're just bundling the token cost in just based on what they see in their data. And I think that's like, you know, for now, if they can accurately predict what the token spend would be, I think that's like a pretty good strategy because the last thing you want to do is have a conversation with someone who doesn't understand token spin well, right in can in contract stage in a sales process, and have to explain, like, oh, here's how the token like we're gonna we're gonna pass that through. We don't have any control of the pricing of tokens. Yeah. So like right now it might be good, but in you know, a couple of years our the model changes and it's you know five to 10x the price, or you want to use a different model, whatever. That is like you know, lawyers and finance people hate that conversation because they have zero predictability around it.

SPEAKER_02

Yeah, it's kind of interesting. Like during the internet days when I was selling internet services, right? We would charge for bandwidth, right? So basically packets, right? Like, oh, you get this much bandwidth, so to speak. And then where everything went was unlimited. Here's internet access, it's unlimited. Like you don't get charged for for bandwidth for packets, essentially, right? Packets or tokens, whatever you want to call it, right? And what the caveat was it's unlimited, but they would throttle you back. Okay, you can get throttled back. It would be I mean, I think even like some of the cell phone providers still do that. Unlimited data doesn't mean unlimited, it means like you'll always have access to it, but they'll throttle you back if you start streaming all kinds of crazy stuff, right? And so a lot of that is does become interesting financial models. I'll tell you. Like during the internet days, like when we were going to unlimited and we could throttle people, we actually our margins improved. So it was actually a benefit to doing those things. So you can you can build the right models. I mean, you have to charge enough, obviously, right? But then you also start to look at patterns and and and so I think you're starting to see that now. With I mean, I played around with Fable, I joke around with our CTO. I'm like, I'm about to use Fable. You know, like get off the checkbook. Yeah, it's good, yeah. It's great. Most people have no business using it. Sure. They can get away with on it. So anyway, I think I think that's been interesting. I think the other thing that's become super interesting around token spend, I think vibe coding has is interesting. I think that most people are wasting their time on it. It's a rabbit hole they're going down. They built some, you know, because getting started. So this is a new opinion for you. Yeah, I mean, get that what's happening is like the citizen coder builds something, they get started fast, right? And the way I look at vibe coding is it's like you code for yourself, you code for your team, or you code for your company, right? And I think a lot of people are just like vibe coding

Token Costs And AI Backlash

SPEAKER_02

stuff, and they somehow and they get individual value right away. They somehow think that like, oh, I should like to deploy this across my company. Well, they don't they don't know about security and like they they can't do any of that stuff, right? So it's it's so easy to get started, it's very hard to finish. And so I think what's happening is I saw something the other day where someone was like vibe coding something. I'm like, you can use Excel, like Excel does that well too, you know, and they're like, and it's I think it's this weird like if you're if your default app used to be email in the browser, and now it's like email, Slack, and Anthro and Claude, you're just gonna do everything in Claude, right? And I think you go down this rabbit hole. So I think people are building a lot of stuff that's really not necessary to do their job. It's just it's like become their default application that they play in, and then they get this, then you get distracted, right? And in a country that has lots of ADHD. I went out to build like some scheduling tool, and then some ended up doing something else, right? Yeah, so I think it's like the distracted nature of it. I still believe in it. And we just had our hackathon, and I had someone there like 25 plus year structural engineer, probably anti-AI at some level, um, was like totally changed his tune. Once he got on the keyboard, built something that was helpful to him with our help, obviously. But he was like, I'm a thousand X behind. His texting was like, I'm a thousand X behind AI now. And before that, he was like, I don't know about this AI thing.

SPEAKER_00

Yeah. To me, the the value right now of vibe coding is it introduces a new perspective on what AI can do. So it's purely like a it's like a what it's like a white pill on AI. Yeah. If you see, you know, if you go from non non-technical to building a software product, like a you know, prototype, it's just like something that if you're not a coder, you've never been able to do. So that changes your perspective on AI. The second is I think for per like for personal applications, it can be valuable for you to like build, yeah, like a personal suite of tools. Yeah. If you know, if you're so inclined, but I do think the the distraction potential is quite high. So you do have to, I think measure just the value and how much time you're spending on it and be pretty disappointed.

SPEAKER_02

And I think it's different if you're doing it for your own, like for fun.

SPEAKER_00

Like are you building like a health app for yourself to track your diet and stuff like that? Like one of the guys, yeah.

SPEAKER_02

One of the guys on our team got a puppy and he built the puppy app. And basically, to he's never had a pet, and you're like, like, when do I need to walk it? How much do I need to feel like just basically like track its progress, right? And have a better understanding how he could care better for it.

SPEAKER_00

It's like a real life Tamagachi.

SPEAKER_02

Yeah, okay. Yeah, exactly. The opposite.

SPEAKER_03

That's fine.

SPEAKER_00

Yeah, so there's like all these small micro apps that you can build for your custom PNF situation, and I think it's great for that. Yeah, no, 100%.

SPEAKER_02

I do think in the world in the world of things happen slowly and then happen quickly. I'm feeling it for AEC a little bit. That a lot of the naysaying and everything else, there's there's a couple dynamics. One, I think the people that are getting it are going deep and doing it really well. Two, I think private equity is kind of ruining AEC. And what's happening is people's companies get bought, the employees quickly get disenfranchised and they

Vibe Coding Benefits And Traps

SPEAKER_02

spin off and start their own thing. And if they spin off and start their own thing and they've been playing in AI for the last year or two, they're starting new firms from scratch and thinking about it differently, and thinking about they hire differently and all of it, right? There's, you know, so I think, you know, I'd always said my most the most exciting thing to me for AI is entrepreneurship. And I think you are gonna see a lot more entrepreneurship in AEC because of AI. But I think partly it's because, you know, I was explaining to someone the other day. Like, you go to you don't go into civil engineering because you're gonna get rich. And if you quickly fit, you know, if if you thought you were, you quickly get the realization that you're not gonna get rich being a civil engineer, and then you go work for a contractor or a real estate developer. That's what a lot of my friends did. They're like, oh, this is terrible. Like, I'm never gonna make any money. They went and worked for a real estate developer, and then they made money, right? So I think like the these companies that are getting bought out, the people aren't there because they wanted to create, they wanted to understand the business side of the business, or that they wanted to build that vocabulary and justify their existence. There was a little bit like, hey, I've got my clients, I want to do good work, I get paid good enough, you know, whatever. And I think private, I was talking to someone about like they said a P firm came in and deployed like an incentive plan for people, and they were so excited about it. And like no one cared, like no one cared at all.

SPEAKER_00

In terms of like their just a basic bonus structure, if they're yeah, they were like, Yeah, okay.

SPEAKER_02

Like it like it wasn't a behavioral change. It's like, oh, that'll be nice. Like, I think I can hit it. Yeah, it wasn't like, oh, I'm gonna work more. I might be in here this weekend. Is there any incentive structure that would change that? I think generally, most engineers that are sticking with engineering, the incentive is additional capability, new, new opportunities to learn new things. I don't think it's money. I mean, people want to get paid well, but I don't think it's like, you know, oh, I want to double my salary next year. I don't think they think like that. I don't think I think even the equity structures and esops and all that, I mean, they're not they're not great. It's all it's all on paper. Yeah. You know, it's almost worse than startup stock. It's all on paper. There's no liquidity to it.

SPEAKER_03

Yeah.

SPEAKER_02

And, you know, so I don't I don't think that that's what drives a lot of people. But I think when the PE people show up and, you know, are trying to sell this view of the world and they don't come from the industry, it's just gonna be bad news. And I I think there's enough warror stories now about like cultural change and you know, and all that, there's like a lot of disinterest around PE. There's there's one firm I know, the CEO, their closely held employee in non-Esop, and he said it's like clockwork. 18 months after a PE firm buys one of his competitors, he gets like email, emails and text messages from people there saying, Hey, can we chat? Hey, can we have coffee? He's like 18 months almost like through the day. That's the time horizon where people say, Hey, I might have had some stay incentives, whatever, but this is not what I signed up for.

SPEAKER_00

Yeah, yeah. So when you say you feel for them, what what's your recommendation given that that's the state of play, that PE firms are continually interested in the space? What's the yeah, what's their what's your core recommendation if you talk talk to someone in the so I think there's some PE firms they're gonna lose their shirt, they overpaid, they overpaid for stuff.

SPEAKER_02

There's less people around to recap these deals where they can make money.

SPEAKER_00

What's the what's the core assumption mistake that the PE firms made?

SPEAKER_02

I think they're overspending. And that the idea that you know people, it's it's kind of there's there's like a simple psychology, right? If you if you talk to some of these firms, we've been around for 60 years. Like, if you look at like this the proof points they even sell in the market is one of longevity, right? Oh, we're building something for the future, we're building something for generations to come. Most PE firms operate on a three to five year flip, right? To recap. I was talking to someone the other day. Their company just got bought six months ago and they're going through a recap. And they're being asked by management, everybody else, oh, we need to put all this stuff together for the new for the new buyers. Like, this isn't engineering work. Me putting something together for the new buyers, right? And so I think, you know, if the PE firms continue to kind of keep flipping these companies, I think they're they're gonna continue, and and by the way, they're flipping them, many cases overpaying. And then, you know, the employees are like, why am I here? Like, you guys aren't in it for the long term. Why should I be in it for the long term? Very interesting.

SPEAKER_00

I mean, we saw a deal the other day. So it almost just it destroys in some ways the mission and the culture of the company. Right. Because it changes the entire, yeah, the mission alignment is gone, and we're building something with longevity and with durability, and we're we we care, we care about generationally where this where this firm goes, where this institution goes. And as soon as you opt out to sell it, and there's now a five-year liquidity cycle tied to it, it destroys that that culture and that mission.

SPEAKER_02

Yeah, man, think about this. I mean, we saw a company the other day that traded for 14 times EBITDA. So if the company stays the same, your payback is 14 years. If it grows a bit, it's short. But 14 times EBITDA, these companies don't grow like that.

SPEAKER_03

Yeah.

SPEAKER_02

And if you're not gonna make it up on growth, where are you gonna make it up?

SPEAKER_00

I mean, to get that sort of price, wouldn't there need to be some historical growth that shows that they're trending in that direction or they can return faster?

SPEAKER_02

The ego of some of these operating partners, they believe they can drive more growth. More growth than maybe is there. But my point being is like if you're gonna pay that kind of money, there's not some, I mean, the P firm's not just looking at growth, they have to be looking at costs. They're gonna look at businesses like, oh, well, that business based on the other business or portfolio of services, this one's underperforming. Get rid of it, get rid of that one. They'll they'll financial engineer it, right?

SPEAKER_00

How much how much is AI playing into the pricing? Not at all. They're not even factoring in it.

SPEAKER_02

No, they have no idea what they're doing about that. I don't I don't think they're not at all. But I mean, look, I think there's you know, not all PE is bad, but I think um there's a lot of folks that are starting to realize it's not the move for their company.

SPEAKER_00

Yeah. Really interesting. So I got to attend your hackathon this week. Decent segue to where we started, which was what's the timeline for AI diffusion across the real economy? And it's interesting to be in those rooms because it's mostly non-technical people. Not, yeah, I mean not mostly, but there's a lot of non-technical people there that are building and prompting the ideas for the solutions. The like one observation I had from it was I think the level of the the the like essentially what you can get to in a day is pretty

Private Equity Collides With AEC

SPEAKER_00

remarkable. Yes. Like the timelines are sh are shorter than people believe. I think everyone goes into a hackathon and they they feel stressed and pressed by the amount of time, like they see you know, taking time. We have eight hours to build a product. But like universally, everyone got to some sort of pretty yeah, like a serious threshold with addition. Yeah, yeah, they had demos, right? Like no one was no pitch felt half-baked, and the product that was demoed was never really half-baked. I mean, and you know, I'm sitting there having, you know, I get pitches from from software companies that have a lot of the same sort of velocity that you'd see in a in a hackathon and even similar features that I've been pitched across the AC space. And it's a really interesting framework. And to me, like I go in those, I go into that sort of event and I'm like, I can't invest in software. It's just it's just so hard. Like there is if if a a group of people that are half into AI and half out of AI can concept and build this feature. And by the way, many of them are are incumbents that are doing this in-house. Yeah, they're not they're not dept shops or they're not startups. Not at all. So, like, if if an if an incumbent with no software capabilities is getting a proof of concept or any in a in a pretty strong product demo in a single day that they feel excited and good about. Like I read their reactions on their face, they're like, wow, this is gonna legitimately help the company. We've been trying to build this for five years. That was like another piece of sentiment. It's like we've been wanting to build this for so long, it's been on the radar of everyone. So, like, we don't need corporate buying to do this, we just need to go out and finally do it. And so, yeah, just if increasingly it feels like software is is uninvestable when I sit in those sorts of events.

SPEAKER_02

Yeah, which is tough. Just absolutely. I mean, I I think the other thing is just my general observation is a lot of the stuff that people are building are features that they expect from the pro cores and autodesks and everyone else, trembles of the world, and they're not getting. And and they're building stuff that's gonna make them productive tomorrow. So it's not even, I mean, I don't I think you know, in a in the world of like the you know, startups have to worry about certain things, right? Oh, is it a big enough TAM? Is it big enough pain and all that, right? For these hackathons, they're not trying to build something for a mass market or for a TAM. They're just trying to solve like this is something I deal with every day. Yeah, wow. I just in a in a day, I just got back 10 hours a week, 20 hours a week, right? Which is interesting. So you look at some of these things and you say, well, that you know, if people can do this to solve their own problems, why are they ever gonna buy a piece of software that they have to learn, that's not built the way that they want to build, that doesn't do exactly what they want, they have to work or do a workaround every time.

SPEAKER_01

Why are they gonna buy another piece of software?

SPEAKER_02

And I don't see it.

SPEAKER_03

Yeah.

SPEAKER_02

And it's even wild too. Like, you know, I have this thing that once a week I read like academic publications around AI, right? Because it's really easy to get caught into the consumer world and the Twitter world of AI, like, oh, this claw this anthropic engineers publish this PDF on automating your agents or what there's plenty of that stuff, right? So I've been really spending it once a week, I try to read like another like kind of deep white paper, right? With lots of calculations on it and formulas and stuff like that. And what I'm kind

Hackathons Make Software Feel Cheap

SPEAKER_02

of starting to contemplate, even the next layer of like, you know, this is all like crud stuff, but even like some of the computational things, the people or the heavy, you know, whether it's FEA or any kind of flow calculations and hydraulic calculations, I mean, all these formulas are very well known. The problem with a lot of these formulas, so yeah, I think we we we had talked about this back when um we were dealing with the flood in in Asheville, right? And it's like if one guy designs a road and one guy designs a stormwater sewer system, and then the other person designs a detention pond, right? They do it independently. Nobody's taking a systems approach to say, like, oh, if this breaks, that'll go. Yeah, it's all individual components, right? And so then you start to think about people building these things and being able to interoperate and really start to simulate. And so I was reading this white paper, I was reading this publication about like the idea of like time domains and how to think about time domains, both like quantitatively or a little bit more qualitatively, like and like sort of, like in the last in the last three days versus 36 hours ago, right? Kind of things, right? And it was pretty fascinating because I could see how someone that's designing a structure and then they're thinking about hydrology and hundred-year floods. I don't know if you saw there was like a flood in downtown Manhattan this week. I didn't see that. Yeah, there's like this flash flood, and Soho got flooded. Right. And so as you start thinking about how crowded cities are, how connected our infrastructure is, everybody's working in little silos. So I think it's gonna be super interesting how you extend these higher order computing to think about things as a system, right? Versus just everybody thinking, you know, even in a building, right? We think about structures operates, right? Kind of independently of HVAC in a way, other than like space, right? Fighting for space.

SPEAKER_03

Yeah.

SPEAKER_02

But nobody's thinking about how they work as a system.

SPEAKER_00

I think that's one of the really interesting things about the AI era is that data can be easily consolidated, seemingly for the first time. I mean, like theoretically, you could get there with you know a massive comprehensive software product, but like the resources you would need to get you know, to do that well. It's like you're on the level of Google, right? Where you just, I mean your your infrastructure and your balance sheet is so like needs to be so strong to resource that.

SPEAKER_02

It's kind of fun. I was talking to a hospital system owner, and what they were telling me is like, hey, one of the biggest things, like, you know, great, you guys can help us with understanding the artifacts of the project and all that stuff. And they brought up this idea. They said, look, you know, when we hire XYZ big firm, it doesn't mean anything. What matters is who's on the project. And so we were basically kind of on the fly on this call, like kind of designing out, like, okay, let me build out, let me go get the resumes of all the people that are actually on the project. And let's

AI For Systems Level Engineering

SPEAKER_02

gain an understanding of their level of capability to serve the project, right? Like, are they good enough to work on this project? Because their point was like, these firms don't ask us. In fact, we'll start with the, you know, we get pitched a group, a team. Here's your team. And then by the time the project gets put out, like, oh, that team's on another project. We have this other team, and we get some level of authority to okay, that's the team. Well, if that team changes, we're just stuck. We're stuck with this firm, and they can keep swapping out people, they can put in the B team, whatever it is. And so they asked me, like, hey, I'd like to have a simple way of re-qualifying people that might get added to my project team. Yeah, except to reject them because I don't have time to read through the resumes and blah, blah, blah. Right. And they were telling me that this big project in the middle of the big project, someone left and they got a new team added to that. One, they're starting in the middle of a project.

SPEAKER_01

Two, they just started with the firm a week earlier, and they were like, we need more about their firm than they did and who to go to and who to talk to.

SPEAKER_02

And I thought it was fascinating. Like, I mean, it's like one of those, like, you know, we're the people business, right? And the biggest variable is the people part of the business. And this person was just super fixated on this idea, like, I need to understand who's on my project. Yeah. And I need to understand what the risk associated with them being on the project is, which I thought was like pretty fascinating.

SPEAKER_00

Which before this year, you know, last year could call it before 2024. There's really no leverage to ask that question. Right. Because, like, how are you going to deliver the information in a way that that person's going to be able to actually assess the quality or the lack of experience of the individual? Yeah. Like read through their resume. Yeah, here's their resume. Like, I'm not going to sift through the whole project team's resume. Yeah.

SPEAKER_02

I think I I think I wrote this post a few months ago about like I've been telling people to send me their long resumes. Because when you send me a one-page resume, you're absolutely optimizing to get the job, right? Which means you're only going to talk to me about the things that you want to pop. Um instead, I'm like, send me your 10-page resume. I want to know everything. And then, and by the way, you're creating the filter of how you want to present yourself. Whereas if you give me the 10-pager, I get to decide how I extract data from, right? You don't you don't get to present your best self if I'm doing it, right? And and so it's so it's pretty interesting. Like the idea of like long form data, the you know, long form information, I think it's it's a great thing, right? Because I want to know someone's entire like work history.

SPEAKER_00

If you can process and compress the information quickly, yeah, which that's the game that we're in now, yeah, then the more information, the merrier.

SPEAKER_02

Yeah, it's kind of interesting. too like I have this thing that in my skills file that understands all my LinkedIn LinkedIn connections and so I can suck in a resume and ask it like who do I know that worked at that firm during that period of time?

SPEAKER_00

Interesting. Which is pretty awesome. That's cool. You would not, yeah, there's no there would have been no way to do that.

SPEAKER_02

Cross-reference that yeah and so now it's like oh you're a Palantir during these dates do you know so and so no it's like that's weird. I talked to them too they didn't know you either so you weren't that important. No really it's weird that you were the head of product and the CEO doesn't remember you.

SPEAKER_00

One other observation I had from the hackathon was the the user there's something with the user experience of vibe coded products that doesn't feel quite right in terms of like it feels like the user experience for how we're using new software today that's AI native needs to evolve and change. And

Resumes, Risk, And Project Teams

SPEAKER_00

so right now you're typically seeing like you know with the vibe coded product you just you just it looks and feels like normal software like it looks like your existing products that you know the the software you buy off the shelf. And so like you know there's a there was one company that you know was was doing they were doing sheets that uh presented at the hackathon and there was the question of like so they they were they were uploading and enhancing timesheets handwritten time sheets and there was just the question of like why do you need to handwrite the timesheet you know in in to in today's parlance all we really need is the data and you know some sort of of you know record of identification to prove that that person submitted you know their their their time card as they you know to meet compliance standards. And the answer was like oh well we just haven't had an evolution of compliance standards. So like we still have to do this for legal purposes and for compliance we still have to fill out the manual sheet. They have to sign it but it just feels it's like can we do like a you know face recognition and you know audio speak into like you know voice to dictate into you know an application that essentially submits the the time like there it doesn't feel like we've quite caught up from a user experience standpoint and we're still doing we're like the the models are serving an outdated mode of user experience.

SPEAKER_02

Maybe but here's the other thing you have to remember is like when you deal with labor forces one you're assuming connectivity you're assuming they have you know number one app used on a construction site is WhatsApp you because you can speak any language you can you know you you can have an Android you can have a cheap phone right so one you don't have continuity of device and systems and and usability right those things may not have signal right so that's a whole nother issue. So I so I so I think some of this is all but but you know what's interesting about like when you talk about UI I was I was actually telling someone yesterday about this about the hackathon it's like when you cook for your when you cook a meal for you and your family and you over salt it a little bit you really say oh my god this is horrible like oh it's a little salty it's fine because you made it right so all these people that are making their stuff they're not going to be overcritical like if it's a third party software company why would you make the button that small like I mean there's all it is is critical about how this doesn't fit me. Yeah when you make it yourself your tolerance for UI and interface and even usability is so much lower because you did it. So I think even like I found it fascinating like the way people think like oh I mean that's not that important. It's like I hear you but I also know if I was a software startup coming in here selling you something you would say that is the most important thing I can't buy your product because you don't have yeah so I think there's this weird like you know fruit's a little salty but I'm not gonna actually admit it or send it back to the kitchen so to speak right but I I think that's that's fine. But but I also think when you're spending your own time on things I think people have great intuition on is that feature going to create real value or not I mean I think about like you know in our in our team right in Slack people ask for things and I'm sometimes I just like go in there and veto them and I'm like how much like you think it's worth like building this little tool or this little app or this little interface or whatever like how often are you doing this? And I I mean sometimes I was like

AI Native UX Meets Real Constraints

SPEAKER_02

well because you have a resource that can build stuff you're just gonna like ship feature requests right and so I think there is a little bit of like and partly it's a time constraint but I think that's what's cool about this stuff is people are building for themselves or not building for the masses.

SPEAKER_00

Yeah well um these in-person ones were a little bit odd to me still but I'm enjoying I'm enjoying the the increased fluidness in our adaptation versus having to look at your face on the screen for an hour