Tool or Takeover? AI in Landscape Architecture

Host Forrestt Williams sat down with Land F/X CEO Jeremiah Farmer to discuss the present and potential roles played by A.I. in the landscape architecture world. While acknowledging the fears surrounding A.I. – some valid and some admittedly far-fetched – the conversation juxtaposed A.I. with the automation of tedious tasks that served as the impetus for the founding of Land F/X. We learned what some power Land F/X users are doing with A.I. in their daily lives, while also hearing about the prospects of workforce reduction and other possible drawbacks expressed by others. Jeremiah made a reassuring case that tech has a long way to go in replacing human designers in the creation of permit-ready sites. Sit down for a nuanced and realistic look at where A.I. might take us, and where it might not. 

Transcript

Copy

SPEAKER_01 0:12

Welcome to Overgrown Ideas. This is a LandFX podcast. My name is Forrest Williams. And with us today, we're very, very special guest. This is none other than Jeremiah Farmer, the CEO of Land Effects. What's up? Oh, hello, everybody. Yeah, this is quite fun. We're so stoked to have you. We brought you in specifically because the topic of today is AI and landscape architecture, and where those two things meet. Being that you're an expert software developer and extremely and intimately familiar with the industry of landscape architecture, we figured you're the best person here to talk about this.

SPEAKER_00 0:57

Maybe you're right, maybe you're wrong, but uh we'll find out. For sure. I definitely have some thoughts and things. Um definitely. I did I prepped for this, and I think um I think it's gonna be a lot of fun. Um you know, um and the one struggle I had with preparing for this is just trying to not go down the rabbit hole of buzzwords and and um black clouds in the future, right? And that's what I'm seeing. Like for instance, um the first the first thing I I looked up and it just and and I admit I used AI for this. But um it but I remembered offhand some things, and it wasn't even able to find me some good quotes, but you know, some generalized things. The telephone, right? Okay. Telephone is gonna destroy society, right? You know, numerous letters of the editors and newspapers were like, you know, decrying how bad telephones were. Um and then there's the internet. The internet is just a fad. It's it's just a fad. It's gonna warp the minds of the young. Warp the minds. Going to destroy our ability to meet people in person. And and the species will die out. I remember reading that. Um how about autocorrect, right? Autocorrect was gonna make us unable to communicate. You know, now it's AI that's gonna have all this. And meanwhile, I just thought it was funny that I used the Internet, Autocorrect, and AI to research that exact paragraph. Um I enjoyed, I enjoyed the the uh assistance that all three gave me. Do you feel like you're getting out of touch with society and your fellow humans? No, to me, I mean, I I just you know, and and I have multiple ways to frame this too. But I mean, fundamentally it's just to me, AI is just computers are now finally doing what they've been promised they would do since the 1960s.

SPEAKER_01 2:50

Since um since Star Trek, when you can just say, computer, make me a cup of tea and clean the room or whatever it is. Exactly. Play some classical music. Trevor Burrus, Jr. Exactly.

SPEAKER_00 2:59

And so it's finally there. I mean, I look back, yeah, when I started on computers, you know, it was with a Commodore 64, and oh, I wish I had the internet then, right? And and you know, I mean, to f find information then was so difficult. And now this is just insane how quickly I was able to get all of this information, and it was by using AI. So I I love the irony of we're sitting here talking about what is AI gonna do for the industry. Ever all my prep all my prep was assisted by AI. I mean, because you know you start by Googling some things, but that Google search is impressive. And and that's probably the first thing to kind of mention. When they say, well, how is AI gonna impact landscape architecture? And and that was like the that um uh the the thing that you shared, which was also from AI. Yeah, it's already impacting landscape architecture, and one of the biggest ones is in searching the internet and researching, and and that's an immediate, immediate assistance that it offers. Absolutely.

SPEAKER_01 3:59

Absolutely. Um there have been a couple of studies, the um International Federation of Landscape Architects, as well as the Canadian Um Society of Landscape Architects, they did some peer-reviewed studies and surveys of the industry, and they're finding that over 70% of landscape architects express concerns about business ethics accuracy and like the erosion of human connection and design. Because everybody's scared of AI just doing everything for them. And that's concerning when it comes to design in particular. But you had something to say about that. Oh, multiple.

SPEAKER_00 4:41

I mean, well, it it you know I mean to put it blunt, you're you're worried about the wrong things. Um For instance, I was like, let me start with this. What do landscape architects spend their billable hours on? Because that to me is uh, you know, I can use that as a sort of a a hitch list, uh, you know, um to guide us. Because obviously, well, if you're spending a bunch of your billable time on manually counting, right? Right. That's that's a non-LANFX user. Yeah, right. Yeah. So can AI help with that? And and so anyway, so I thought this was neat. You know, what do LAs spend their billable hours on? Um concept and schematic, 25%, design develop, 20%, construction documentation, 30%, project management, 15%, construction admin five, and the rest is five. All real basic numbers, and I even had to massage those numbers a little bit because it was, you know, just all these ranges. There's no authoritative thing on that. And this is all right in line with going all the way back in time. Um, you know, these sorts of things. And I remember when we started Land FX, uh, you know, my dad had come up with these very similar numbers. Um and the issue was that these surveys were showing 30, 40% of a project was spent on construction documentation.

SPEAKER_01 6:10

Okay.

SPEAKER_00 6:10

Which is just a lot. Yeah. And and that was the whole original pitch of LandFX was let's shrink that, give you more time for the other phases. Um, and then if we even drill down into that, we're like, oh, so wait a minute, there's an obvious thing. Okay, is AI, computers, whatever going to help with construction documentation? Let's drill down into that. Hardscape, grading, drainage, planting, irrigation, details, and specs. And it's like, oh, well, landfacts, we're already assisting with four out of five of those. Um specs is an interesting point I want to revisit. But um it's like we're already there with engineered intelligence. And so why is there a worry that AI is gonna, you know, who's gonna say I click a button and it's gonna generate a construction document for me?

SPEAKER_01 7:02

That's what people are saying, and that's what people are concerned with. Now, I do want to I do want to follow that up though. I did have sales reach out to some of our Rockstar clients and ask about what they're doing right now and what their concerns are for the future. So we've got some really good feedback. Oh, I love this. So um, well, here's somebody that we both know, Will Howard over at StackRock. Okay. So he's using it for writing emails, researching code, planning concepts, assisting with renders. Obviously, he thinks it's not there yet. He feels like it's I like this quote. He says he feels like it's 2002 Wikipedia. There's a lot of great things about it, but you still have to fact check. Yes, 100%. But there's some other people. Um, who is it? Uh we were looking at Andrew Bolt was really concerned about, you know, um maybe teams will go down from 20 to 5 once a lot of these things get automated in offices with auto layout tools and things like that, and how that will actually have an effect on how many people you need in a landscape architect office. Um there's a lot of these people expressing concerns about like, you know, what if? Why can't it just click a button and give you a layout and then you just finesse it?

SPEAKER_00 8:20

Because I have a good answer for you. Okay. I love this answer. The fear that AI will suddenly start stamping out fully compliant, permit-ready landscape plans is wildly misplaced. Not because AI isn't powerful, but because the bottleneck has never been drawing production. It's liability, regulation, and professional accountability. AI can hallucinate a beautiful plaza, it cannot hallucinate compliance. And that is the finished plan. The finished plan. Needs to have a stamp, needs to get approved. You're gonna say generated by AI, that's not getting approved. That's the bottleneck. And by the way, that paragraph was written by AI. I was chatting with it, rapping with it about all this, and just got some really great paragraphs from it. But you know, that was, and that's the beauty of it's taking some of my words and my concerns and it's regurgitating a little bit of what you're saying, but just worded a little better. Um, but but even then, I mean, it goes right. The first things to be automated, repetitive drafting, plant schedules, quantity takeoffs, code lookups, red line incorporation, visualization, right? All the yeah, all these things that, first of all, we're already doing, and things just like what Will says that they're doing, you know, those rapid concept visualizations. Absolutely. So, you know, you take a firm that's not using land effects, um, yeah, there would be more of a hit. But since we're already addressing a lot of that, when we've seen that, that's actually something that's been not so much concerning, but just something we have noted quite heavily, is quite commonly uh a firm might buy a handful of licenses. And then two, three, even four years later, they drop half of them. You know, now they only need three. And you know, that could concern a software company. Oh my gosh, you know, you're dropping half your licenses. And but it's because what do you think it is? They're like, oh, turns out with the efficiency of the software, we don't need as many people. Now, even with LandFX, that usually takes multiple years to get to that point. So with AI, same thing. You know, but again, what is the AI going to do? Because you still have your professional stamp on that plan. And if an issue comes up during the certification permitting legal things later down the road, and you go, oh, I don't know, I clicked the AI button for that, you're not gonna get that stamp anymore now, right? And then won't approve. Yeah.

SPEAKER_01 10:52

Um but a lot, you know, there's a lot of concern in other creative industries where they're trying to build some sort of government regulations where AI-generated content has a sort of watermark, a digital watermark of some sort, so that in this case, say a landscape architect doesn't auto-lay out the plan with AI and then submit it and say they did it themselves, and it's difficult to trust that. So that's you know, that's another one of those many, many, many layers of issues and concerns about the future. Because at right now, if there is an auto-layout tool, who's to say they that people won't just phone it in and say they did it themselves?

SPEAKER_00 11:35

So they're okay using spell check, but they don't want AI. I mean, that that that's where it just you know boggles my mind. Is it's it's just the next step. Um the the next thing I had was just comparing it to BIM. Um and I I think that's a a nice angle to take. Uh first of all, you know, we've got to go down memory lane. You know, what was CAD pre pitched as and what was it offering us from really the late 80s all the way into the 2000s, right? Oh, it's gonna make you more efficient and revisions will be faster and all these things. For some industries, that was kind of true. For landscape architecture, um, I I I just distinctly remember, you know, uh my dad transitioning to CAD. And and he even said uh he was willing to accept, you know, a huge hit in initially moving to CAD. Um, but he way even underestimated it because it took so much effort to try to move to CAD. Because, you know, ink on paper, you just do it. Right. On CAD, you're just constantly struggling with the limitations of CAD. Um and so it was very, very challenging. And and you know, it was just very hard to kind of point to, oh yeah, it took me this long to do a project, and then with CAD, if anything, it took more, you know. And and then, especially those early days, plotting was a problem. Oh, that's a whole other rabbit hole. Yeah, so then BIM comes along, right? And then BIM, remember, you know, BIM, the basic the basic pitch of BIM was what if concept design and design develop? What if those phases were merged? What if you have the super software platform, you draw it in 3D, and it's just generating your construction document for you. That was always the pitch of BIM. To this day, it still doesn't do that. Um to this day, it doesn't necessarily save any more money or time, just like CAD. Um, and it's just this tool that you can kind of go, okay, some things are more efficient here or there. But you know, really it's it's the purpose-built software is where you get those huge benefits. And AI is just a little you know tool that could be added onto that, like what we're looking at, like the you know, the photo or the plant suggestion and the plant photos.

SPEAKER_01 13:58

Right.

SPEAKER_00 13:58

The way we're able to sprinkle it in there. Um, you want to talk a little about the plant photos? I think that's a really nice sidebar.

SPEAKER_01 14:05

So that's like, yeah. So the two things that LandFX have implemented AI in directly in our software is the chat bot assistant that's trained on all of our documentation and has been maybe 50, maybe 40% success rate with clients coming in with questions. They get what they need from the chat bot, and then they're done. And that's killer. And then, of course, the second one is the plant photo generation that we put in for we have over 50,000 plants in our in our database, and not all of them have photos. So a lot of times our users they get in, they need to find a photo for a plant, we don't have one, so we put in the uh Google Nano Banana, which I love saying. Uh we put that in as the image generator um platform inside the plant database dialog. So user can click a plant, doesn't have a photo, auto-generates one from AI, and it's really, really good. We it's getting used um roughly a dozen times a day right now. We've only had it live for two months, and those dozen times a day are specifically clients that select that photo to use in their plan as the representative photo for that plant. That's not tracking how many times they've rendered the image and refreshed it, but a dozen times a day getting selected.

SPEAKER_00 15:28

Yeah.

SPEAKER_01 15:29

So that's pretty stellar. And it's filling a uh gap in our database where we don't have all the photos yet. So it's a win-win for everybody. And I think it's a really cool integration of AI. And but it's also not one of those, like we were talking about, it's not an auto layout finish the job for you. It's a purpose-built tool to fill uh a niche, uh, just like you were saying, it's which is a great use of AI as long as you look at it from those angles.

SPEAKER_00 16:00

Another one I was looking at is, you know, like uh yeah, you mentioned Will mentioned about you know for like writing emails and researching things. Those are right now where AI is really good at. Uh some other things um it's really could could be really good at is contracts, paperwork. I found some really good because those are laborious and tedious. Exactly. Um, you know, specs is something, you know, we had always talked about having something in LandFX for writing specifications and just never got into it. And there but there's some really good uh options out there. And I was looking into it, I believe there's about three main ones. The one that I felt seemed like it was embracing AI the most, it you know, may and again, maybe that's just in how they're wording it on their website. I didn't actually try them all out, but um, but it was still the most highly rated was the Dell Tech spec point. Um and so that was interesting because you know every once in a while I kind of look up to see where are we with those spec writing? Because ideally, what we've been talking about for ages is what if we could kick out all of those CSI numbers right into a spec writing software and it just writes those specs. You just give it the numbers and it writes all those specs. And that would be a huge time saving. And that would be huge. Like that's something that, again, AI can do right now today. Now are you gonna say, oh no, that's gonna mean this person over here I hired to do our spec writing, I don't need anymore? Like again, it's just like implementing LandFX, implementing anything is well, you know, maybe they can do something else. You know, you get more value out of that employee. I never understood. Why would I be worried that? I mean, my gosh, when it's come up anytime running a business where I don't need an employee, I'm not sad about that necessarily. I'm like, oh, this is good. We're gonna get a, you know, this is and I don't immediately go, great, let's fire set employee. It's like, well, what what should we do? Do we transfer them? Do we make a new project for them or whatever? But it's never like to think that the sky is falling because I'm my team is more efficient is a very strange angle.

SPEAKER_01 18:10

Yeah, efficiency is the is the goal. It's not something to be afraid of. Right.

SPEAKER_00 18:15

Yeah.

SPEAKER_01 18:15

But you know, it does have to get used wisely. Um let's see, we have some other uh fun client stuff here. RRM and Slow, they have formed an AI committee to figure out how to integrate AI responsibly. Um Trevor at Designs by Sundown in Lake Bluff, Illinois. Uh definitely using it for conceptual stuff, you know, inspirational image generation. That seems to be a recurring theme with a lot of these comments. Um but again, they all express skepticism about using it in anything more serious than that. Sure. Um But the can the the concern is expressed with a couple of people still. They're just like, uh, what's gonna happen in five to ten years?

SPEAKER_00 19:06

It is also, I mean, I'll throw it out there because it's coming up, everyone. I want you to think about this. This is a good little adventure for you. And this may go ahead, use ChatGPT to write that Mother's Day card for you. Um and and I know at first you're like, oh my gosh, I could never do that. Yeah, it's it's hard. It's harder, you know, every year writing that Mother's Day card and come up with something. And there's been times where I'm like, hey, can you write me a quick little poem about it's so good. It is good.

SPEAKER_01 19:37

It's so good at that. Does it feel less like more impersonal when you're doing that?

SPEAKER_00 19:42

I don't know. I mean, it it's you know, I don't know. I mean, because I'll admit, and maybe I was already primed for that because my hack before AI is when I would buy a Mother's Day card, I would look at other Mother Mother's Day cards until I saw, oh, that's a nice little thing to say. And I'll write that in this other card. Right. You know, and so I was kind of already sort of doing that. You know, I still love my mom. You know, I have no problem of telling her.

SPEAKER_01 20:11

You're not outsourcing that to AI.

SPEAKER_00 20:12

No, it it just you know, it just you get kind of that brain lock. You're like, oh geez, I need to write a Mother's Day card like today, it's Saturday, and Mother's Day is tomorrow. And and you're like, oh boy. And you and I have to write something original that I've never said to her before. That's hard.

SPEAKER_01 20:27

Well, you're hitting on something because that essentially is the creative process where you learn from other existing things and you combine it yourself and output a fresh idea, where conceptually AI is doing something similar. It's trained on all of human knowledge and it's combining things to output not necessarily a fresh idea, but a consolidation of other ideas. Uh so it's a very similar in some ways. Trevor Burrus, Jr.

SPEAKER_00 20:54

It is to be yeah, and that's a great way to word it. It really, to me, that's the way I think of it. It really is just kind of you know, you have the internet, and it's now kind of this just summarization system of all human knowledge. Yeah. Um, I found another another couple great examples uh available today, which again are are incorporating AI into the into these ones, uh into these tools. Pandoc. Panda, P-A-N-D-A, Doc, D-L C pandadoc.com.

SPEAKER_03 21:26

Okay.

SPEAKER_00 21:27

Um for doing you know all your contracts and all your legal paperwork and all that, all AI assisted. I'm like, wow, I I was blown away, by the way, at that website. I was like, that you know, if I had to you know write more contracts and things, I'd be doing that in a heartbeat. Um and the other one was monograph, M-O-N-O-G-R-A-P-H, monograph. Um, but that one is project management and budgeting, you know, and so especially uh doing a construction project, you know, all the different phases and everything. Um there are these tools here, and again, they're just starting to sprinkle. AI in there. And again, is that going to evolve over and again, how many years? To where you can just talk to it as if it's your own in-house lawyer or something? Hey, AI assistant, write a contract for this job. And that's all the only prompt you have to give it, and it just knows everything because it's on your company server and it does it. And is that a bad thing? I don't know. Are we going to evolve to where you are using literal multiple AI engines to cross-check each other? Right? I mean, probably.

SPEAKER_01 22:34

Well, that's that's a bit of what Agentic AI is doing currently with some of like the cloud bots, claude co-work, um, cloud code, things like that. What they are kind of doing is you tell it to do something, it'll do it, and then it'll review it, and then it'll refine it, and it'll kind of like auto-go on its own until it gets to something usable. Um but also the guardrails and security issues on those agentic platforms right now are chaotic. So it's a little bit dangerous.

SPEAKER_00 23:07

It's it's still early, and that's you know, and that's the and that's the other thing to always keep in mind. I mean, look at the things people said about air travel in the 1930s. You know? Yeah, exactly. It's dangerous, it's expensive. And it's like, okay, hold your horses, you know, wait 50, 75 years, and you know, you can get a $100 ticket to Ireland. Um, hello. That's pretty amazing. Aaron Powell Yeah, it's uh one of the most impactful things in society is air travel. Yeah. Um I thought another uh one that I felt was uh worth repeating here. Um you know you'd mentioned that the AI chatbot um adding on to that, um I I check those tickets every day. That AI chatbot is pretty consistently handling 10 percent of our tech support system. Aaron Ross Powell Okay.

SPEAKER_01 23:56

It's handling 10 percent. Uh so people that get to it, roughly 10 percent, are getting what they need and moving on. Aaron Ross Powell Correct. Yep. Oh, I thought it was higher than that.

SPEAKER_00 24:06

Um, it's uh no, it's just uh of the total of the total tech well no, it's uh it's um of the people who go to it, you're right, about half of them are happy.

SPEAKER_01 24:16

Okay, okay.

SPEAKER_00 24:16

But not everyone's going to that first. Correct. So of our total tech support load, okay. Which is why we're you know setting that up for Leah to be kind of our tech support concierge. That's right. Right? Because this it's only gonna get better. Yeah. Ten percent of our tickets, and like uh to me that's I mean impressive because that's that's effectively um uh another a full technician salary. That's the value. Yes. The value immediately is one more technician. But also since it's 24 hours a day, I I would offer it's almost the value is even more like two to three technicians already. Yeah. And it's only gonna get better. Um but the other one, Ian, you mentioned the the Claude and the writing of code. That's that's been the big impact of AI so far everywhere, is in code writing. Um what would you say? I would say my development time is easily 25% faster.

SPEAKER_01 25:07

I would say, well, you've seen my integration.

SPEAKER_00 25:10

You're just you are this guy, he doesn't even type code anymore. He just whispers to himself, hey God, can you just spitting out code? Like, how'd you do that?

SPEAKER_01 25:22

So I um I use cursor which implements agentic coding, so it can iterate over itself until it gets to something really good. Um but what that I'm I'm really skeptical of it. So I review everything it does. I'm not shipping AI generated entire code base stuff. But even being said that, it's still I would say 40% faster. Yeah, that's impressive. Yeah. Yeah. Well, here's a good example. The um the chat interface that I'm working on. Yes. I I would not have been able to even do that as a side project before AI, period. So I mean, uh that's that's a lot.

SPEAKER_00 26:14

Oh, yeah, yeah, just one little project. Exactly. Um and yeah, we're gonna see more of that. That um it's impressive. This was an interesting one. As I was researching this, I basically got curious about India.

SPEAKER_01 26:26

Okay. Right?

SPEAKER_00 26:27

You know, it's all this outsourcing of code writing to India, yeah. And outsourcing of call centers to India. And I'm like, huh. I wonder how this AI takeover is impacting that. Because I would imagine if I had outsourced anything to India, I would now be considering using AI instead. So it could be curious. But so how did I find the answer to that question? Again. Ask AI. Way faster. You're telling me I'm I'm gonna limit to only Google searches and read article after article and everything. And I could the AI summary was right there. Now, so but again, this so this did come from an AI summary. Okay. So I don't know how much of this was hallucinated. But you know, but yeah, like think of that. And so AI said, well, offshoring has definitely slowed down, but companies are not pulling out of India. Meanwhile, India, both at the government level and the private level, they're I wouldn't say full-blown panic mode, um, but very importantly pivoting to AI. They're investing massive amounts of money to retrain people so that they can be on top of it. And they are, and so now like one of the things they're doing is they're, you know, all these engineering companies in India are retraining their people on AI prompt engineering. And so they're just gonna use AI tools to still do the same thing they're doing. And so, and probably better than people here, because now they're trained in it. Right. And so it's a great example. So, like, well, if India is a country or these large firms in India are not necessarily panicking, why would a landscape architecture firm panic?

SPEAKER_01 28:05

Aaron Ross Powell And you know, to be honest, the landscape architecture industry, especially when you compare it to the big software companies, have always been a little bit of the black sheep. AutoCAD's not thinking about us, right? Revit's not thinking about us. So we're going to probably be the last industry to be fully automated if that even does happen. Trevor Burrus, Jr.

SPEAKER_00 28:27

Well, it does. So yeah, as Autodesk rolls out these AI assistant tools in CAD, they're not necessarily going to apply to placing plants, placing irrigation or it's going to be on us to to um create some of that. But then I think it's probably good to kind of mention what are you know what are our sort of priorities and our roadmap for that was the next thing on the list.

SPEAKER_01 28:50

How did you know? Beautiful. So what is the future of land effects and AI? Like what are we looking at in the future? Well, I liked this quote.

SPEAKER_00 29:01

This did not come from AI. Okay, good. Oh, this his whole thing came from AI. Um I love that this came from Aviation Week, which uh was is very poignant because it was an Aviation Week article um way back in the day that convinced my dad that he had to move to CAD. So he saw this article, just this whole room of, I think it was Lockheed, if I remember right, but um and and it was obviously came from a press release from Lockheed, you know, as happens, but um, but just saying how look at here's this entire room of engineers, and all the drafting tables have been removed. And it's all just simply computers. And this was about 1988. Wow.

SPEAKER_01 29:46

You know, okay.

SPEAKER_00 29:47

And so my dad sees that and he goes, Oh my God, because it was barely a year earlier, you know, they had an article about, oh, here's the future of CAD. And it's like, uh-huh, sure. And they're like, it's these $30,000 machines and blah, blah, blah. And it's like, yeah, okay, that's gonna that's and then all of a sudden a year later, boom, they've completely swapped out with just basic consumer grade computers. And he goes, Oh my god, this is happening way faster than anyone could have thought. And but he was he was a little over the top with you know, being scared of the future, which good enough, that's why he was one of the first LAs to move to CAD. Uh, and then it born landfacts. Yeah. So this Aviation Week article I read just last night, um, SPIA, the Society of Professional Engineer Employees in Aerospace, so it's the largest uh union for engineering employees, most of the Boeing engineers are SPIA members. Um, their official statement, we're likely to see AI-assisted tools used by our members long before we see our members replaced with AI.

SPEAKER_01 30:56

Nice.

SPEAKER_00 30:56

And so in the same way that, you know, that the aerospace engineering, of course, is going to be at the forefront of technology. They're always gonna be two, three, four steps ahead of us lowly landscape architects. They're even saying it's not replacing anyone, it's only augmenting them.

SPEAKER_01 31:12

Comes back to that.

SPEAKER_00 31:13

Efficiency gains is the goal here.

SPEAKER_01 31:14

It's not something to be afraid of, just like we said. Yeah.

SPEAKER_00 31:17

Yeah. But um I I think the the other the where I see it is there is truth to it is more like that sort of 20-year time frame. Yeah. And and and that's true, but then again, who knows what's going to happen in 20 years. Um but but certainly, you know, would I, for instance, recommend someone go to college for computer engineering? I would say no. So there definitely are some careers. Um to go to get an MBA as a financial analyst, I would say no. Don't do that. Right? That's a bad time to just be an entry-level person. And like those are some of the careers. Yeah. I mean, you know, Wall Street, you know, so to me, there's something, not our landscape architects can be replaced. Those are very obvious. Are AI trading tools gonna monopolize Wall Street and probably create a crash, like when the computer trading tools create a crash in what, the early 90s, mid-90s? What was that? When there was the blip. Yeah. Yeah. So without fail. Oh yeah, it's it's it's coming. Um and you know, I was just thinking, like, you know, is every interaction with every company going to be with an AI bot? Probably. And in fact, it actually already is now. You know, and but soon enough. Here's a fun one. What about the government? You know, I have this whole personal struggle with the uh the DMV in Nevada. I don't want to go into the story. But um yeah, you know how nice it would be to just instantly talk to an AI bot at the DMV?

SPEAKER_01 33:01

That would be so much nicer. I wish that would happen tomorrow, right? Aaron Ross Powell Well, on the flip side, they are integrating AI in the Department of War right now with Palantir. That sounds a little scarier to me. Like, why not? Please do the DMV first.

SPEAKER_00 33:16

Yes. No, and that's yeah, that's that's a fun sidebar. But yeah, that that's the one that boggles my mind because everyone's falling over themselves trying to say what sort of safeguards they're they're building into AI when one of your main consumers is the military. Yeah. And I'm like, so no, there's you cannot have, for instance, a safeguard saying I will not ever harm a human, right? Yeah. Is that the prime Isaac Asimov's Isaac Amunov's prime thing. Yeah. Right? No, we've already thrown that one out. Yep. Because we want cheaper wars across the board. Every country does.

SPEAKER_01 33:50

Yeah, yeah. So it is a little bit scary. But you know, the future, like looking at the future is so funny because it's it's scary now from our perspective, but then when you look back in time and you think, oh, well, what were we thinking? Of course. Like I just watched that movie BlackBerry the other night.

SPEAKER_00 34:08

Oh, that's a great movie.

SPEAKER_01 34:09

It was a great movie, right? And it was what, uh 2000, the years were roughly 2001 to 2009 was the time frame of this movie, or 2012, something like that. And at the end of the movie, it was like the birth of the iPhone and how it took over. And it's kind of funny because looking back, you're like, that doesn't seem like it was that long ago, but I can't imagine a time without the iPhone now. Yeah. And so you put that in this perspective. In 10 years, it's gonna be difficult to imagine a time without AI. Oh, yeah. It'll be integrated in so many things and aspects of our life that we won't things are gonna evolve and we're we have no idea where it's going.

SPEAKER_00 34:51

And and you and you look at these, it it just in just modern history. I mean, the first was the uh, you know, the Steno pool um, you know, was wiped out with the computer revolution. And, you know, I uh it's so fun to see those you know movies from the 50s and 60s, you know, the room full of that. Claggedy, clockgity, claggity, clock. Yeah, the wonderful Harrison Ford movie from the 80s. Yeah. And it's just this giant room of secretaries. Yep. Those all got wiped out over the 80s and into the 90s. And you know, and it's like um, yes, some people lost their job, but it evolved. And then with CAD, it evolved. With BIM, it evolved. With AI, it's evolving. Yeah. And it's just simply um evolving is all. And yes, you can always go, oh, it was better just rooting around in the mud. Why'd we have to grow legs? Like, well, okay, fair.

SPEAKER_01 35:41

But I would say the the 20th century was uh was prime society. It was really nice looking back.

SPEAKER_00 35:49

It's always things always look better looking back. They always do. I I think you know the the the the thing to hold on to looking forward is you know, take away the idea of AI. Pretend you don't even know AI exists, just go back in time, barely what, two years. Yeah. What were you complaining about then?

unknown 36:07

That's it.

SPEAKER_00 36:08

Just go hold on to that. What took a bunch of time? What aspects of your job sucked? What aspects of just running a business sucked? And and go, oh, wait a minute, what can AI offer for that? You know? Exactly. Um, like you mentioned, we're we're having a whole, we're revamping a bunch of HR procedures and the handbook and all these things. It's a giant pain in the butt. AI is super helpful. Yeah. Because we're like, okay, we need because we've employees in what, 14 states, I think. And it's like, so we need w what labor laws apply across all 14 states. I mean, it would be a massive cost, and we're able to just get that for free.

SPEAKER_01 36:53

Right.

SPEAKER_00 36:53

Um, you know, like you said, the contract writing and the emailing, all these things. I'm waiting for I want Microsoft Outlook to be completely replaced with just a an interface, a little person that you can talk to.

SPEAKER_01 37:12

Uh well like maybe in the shape of a paperclip. Well, yeah, maybe not paper clip. No. I learned that lesson. This one will be a staple gun.

SPEAKER_00 37:22

Um But it just, yeah, but like as if it's your own assistant. And you could just say, hey, assistant, what meetings do I have this week? You know, sort of a thing. Instead of constantly you scheduling every meeting, replying to every email, reading every email. You know, what if you you know, how would it be different if you actually had an assistant? And you'd be the 20th century. It's an old 1950s sense, right? Yep. You know, you dictate your memo to your secretary, you know. And they do all the busy work. It's kind of bringing that back. And you're like, and there is a nostalgia for that. Hey, there we go. Start wearing suits again? Yeah. Smoking in the office.

SPEAKER_01 38:01

Jeremiah, thanks for joining us. Um, this has been really fun. Um and I think the point of our conversation is that the future's looking bright, and there's not too much in the immediate future to be concerned about. In the long term, who knows?

SPEAKER_00 38:18

Yeah, I think I it's it's it's definitely good things. I mean, it's it's concerning. Um I mean, the thing that's ironic is it's literally to me, it's more concerning to us. Um because AI has the promise that you know any teenager in in a basement could, you know, vibe code an alternative to land effects, right? You know, I mean that's that's what it's offering. And any competitor could could rapidly spool up and reverse engineer some aspect of our software that they don't have. You know, those are scary. You know, we have to embrace AI with code writing because it's so much more efficient. So we have to do that immediately.

SPEAKER_01 39:01

Well, fortunately, we have so much data. We have the plant database, we have code. Irrigation manufacturers, site manufacturers. You can't vibe code that.

SPEAKER_00 39:08

No, correct. No, we we've got we've we we're we're very secure. Um, but I'm just saying that you definitely it's impacting our industry far more extremely um than certainly a professional certified engineering discipline. Right. Right. But yeah, but it's it's it's still coming. And but it's just exciting. Like I said, I'm just been blown away at how quickly it's offering these really finely usable things. Like I said, I was promised in the 80s and they never happened.

SPEAKER_01 39:39

And they're finally here. They're finally here. I love it. Okay, thank you for joining uh the Overgrown Ideas. This is a podcast by Land FX. We come out roughly once a month. And you know the deal. Smash that subscribe button. You gotta do it, you gotta do it. We're slaves to the algorithm and slaves to the AI algorithm. So it's just a part of reality now. Please, please subscribe. Thank you very much.