Latest / The Land Development Podcast with Ryan Glick / BONUS: How AI Is Changing Land Development with Carter Malloy - TLP196

Transcript

BONUS: How AI Is Changing Land Development with Carter Malloy - TLP196 · The Land Development Podcast with Ryan Glick

• 0:00
  SPEAKER_01: Human beings are necessary to do deals, right?
• 0:03
  SPEAKER_01: We're exchanging real amounts of money when you're talking about a land deal.
• 0:06
  SPEAKER_01: Six, seven, eight figures.
• 0:08
  SPEAKER_01: There's like real trust involved.
• 0:09
  SPEAKER_01: Sitting down at the dinner table with the farmer, they don't care what tools you used, right?
• 0:14
  SPEAKER_01: Or what nerdy applications you've used to go and find that.
• 0:17
  SPEAKER_01: They care about like who's the human sitting here and do I trust them?
• 0:34
  SPEAKER_00: Hey, what's going on, everybody?
• 0:36
  SPEAKER_00: Welcome back to the land development podcast.
• 0:38
  SPEAKER_00: I am your host, Ryan Glick.
• 0:40
  SPEAKER_00: Well, for those of you who tune in on a weekly basis to every episode, you're probably wondering why we have an off-cycle release here.
• 0:48
  SPEAKER_00: But of course, from the title, you know this is a bonus episode.
• 0:51
  SPEAKER_00: I've had a lot of you reach out asking me when we're going to do an episode on AI related to land development.
• 0:57
  SPEAKER_00: And so, because of that, um I reached out to Carter Malloy, who is the founder and CEO at Acres, and asked if he would have a conversation with me on AI related to the industry.
• 1:08
  SPEAKER_00: And so that's exactly what you're going to get out of this episode.
• 1:11
  SPEAKER_00: It's not a software promotion episode or anything like that.
• 1:16
  SPEAKER_00: Um, we talk about AI in general for the industry, different areas that it makes sense for developers to be thinking about using AI, how AI might change over time, how we've already seen it change over time.
• 1:29
  SPEAKER_00: And so this is a really good conversation.
• 1:31
  SPEAKER_00: And I think it's an interesting topic.
• 1:34
  SPEAKER_00: And there's still a lot of uncertainty related to where AI is going to be in 18 months, two years, three years, five years, and so on.
• 1:42
  SPEAKER_00: It's just moving so fast.
• 1:44
  SPEAKER_00: So, with that said, um, I do have an ask for you if you want to get into even more detail on AI, if you'd have any interest in us doing more of a webinar style episode where we actually get hands-on with some of the AI tools and just walk through different things.
• 2:02
  SPEAKER_00: If that would be of interest to you, uh shoot me an email at Ryan at landdevodcast.com and let me know.
• 2:11
  SPEAKER_00: I want to see what type of interest there is.
• 2:13
  SPEAKER_00: And if there's enough interest, maybe we'll put one of those on and get uh a few of us on here and and actually walk through some real world examples of how you could do some of this stuff on your own.
• 2:23
  SPEAKER_00: With that said, let's go ahead and get over to my conversation with Carter Malloy.
• 2:31
  SPEAKER_00: We've got a lot of things we're going to get into today, all revolving around AI and land development.
• 2:36
  SPEAKER_00: But before we get into that, I want to get a little feel for your background and how you got into this space.
• 2:42
  SPEAKER_00: I mean, obviously, you know, you founded and continue to operate acres, but uh, you know, even before that, what was your involvement and kind of your path into the land development industry?
• 2:54
  SPEAKER_01: Put most simply, it moved from investing in data businesses to investing in land to building a data business.
• 3:01
  SPEAKER_01: Uh that is the fastest way I could say that.
• 3:04
  SPEAKER_00: Since you did have that middle piece in there where you were investing in land and then ended up back, you know, starting a data business.
• 3:11
  SPEAKER_00: On the land investment side, when you were doing that, you know, what was it about the land investing business?
• 3:19
  SPEAKER_00: Or I'm sorry, yeah, what was it about investing in land that led you to say, you know what, I should get into the uh the data side of this industry?
• 3:27
  SPEAKER_01: Yes, the the classic um we solved our own problem, right?
• 3:30
  SPEAKER_01: I think that's uh how how many of the better businesses out there uh got started.
• 3:35
  SPEAKER_01: If somebody had a problem and they went out and go, they were stubborn and perhaps uh a little or a lot stupid uh to go out and try to solve that themselves.
• 3:42
  SPEAKER_01: And that is how we got here.
• 3:43
  SPEAKER_01: So we were deploying capital across, I think it was 23 different states in the US as land investors.
• 3:50
  SPEAKER_01: And we had this big problem of data reliability, of pipeline management, of you know, just being able to find land in the first place uh and then underwrite that land.
• 4:01
  SPEAKER_01: And doing that at scale when you're trying to do a you know a deal a week uh is is brutal.
• 4:07
  SPEAKER_01: Even a deal a month, right?
• 4:08
  SPEAKER_01: It's just really hard to execute.
• 4:10
  SPEAKER_01: And so we just like everybody else in the land world, uh, we had like eight different pieces of software and county websites and state websites and city websites, we're bookmarking.
• 4:21
  SPEAKER_01: So we just got fed up enough.
• 4:22
  SPEAKER_01: Uh, we said we're gonna go try to build our own.
• 4:24
  SPEAKER_01: And so we started that effort uh five years ago.
• 4:27
  SPEAKER_01: We put acres on the internet four years ago, like right now.
• 4:30
  SPEAKER_01: So it's our it's our birthday week here soon.
• 4:34
  SPEAKER_01: And uh then as that began to scale, we we did recognize, like, hey, this is in fact a real business, let's lean in.
• 4:41
  SPEAKER_01: So we we sold the land investing company uh to private equity and made this our full-time job.
• 4:46
  SPEAKER_00: Well, I want to get into some of the, you know, we've had people ask us like, hey, when are you gonna do an AI episode?
• 4:53
  SPEAKER_00: And so this has been something people have been asking for on the show.
• 4:56
  SPEAKER_00: We we've tended to be more driven by operators in the industry.
• 5:01
  SPEAKER_00: I mean, you were an operator for a period of time and maybe you still are on the side, I'm not sure.
• 5:05
  SPEAKER_00: But the uh, you know, getting into AI, so many people listening in will have already done things.
• 5:12
  SPEAKER_00: Everybody's tinkering with it.
• 5:13
  SPEAKER_00: Some people have built some sophisticated things using AI.
• 5:16
  SPEAKER_00: But just in case we have somebody listening in that's not super familiar with AI and really what AI is, you know, how would you describe to somebody who's getting into this for the first time, like, what is artificial intelligence?
• 5:32
  SPEAKER_01: Yeah, when we talk about artificial intelligence today, uh, we're usually referring to LLMs, right?
• 5:38
  SPEAKER_01: Large language models.
• 5:39
  SPEAKER_01: What those do to to make its core is they extrapolate text.
• 5:45
  SPEAKER_01: And so they you you speak to it, you write something to it, and it it tries to understand the context, and then uh basically extra uh present you an answer that is relevant to what you're looking for.
• 5:56
  SPEAKER_01: We've all seen that inside of Google and sort of the basics of of AI.
• 6:01
  SPEAKER_01: Uh but at the frontier where where a lot of our team uh operates, it is is doing incredible, incredible things, right?
• 6:08
  SPEAKER_01: So in our case, and there's a multitude of cases out there and ways you can use it, in our case, what that means is we you have uh think of it as like a brilliant intern or a team of them, an army of them that work 24 hours a day that can go operate maps for you.
• 6:24
  SPEAKER_01: Right.
• 6:24
  SPEAKER_01: So if you're in the business of finding land or doing deals in any capacity, you can now in natural language talk to the map and sign up Acres and it will go execute whatever function you're looking for to do for you, go find that information and help you do your work.
• 6:38
  SPEAKER_01: And any of your work that's done on a map, uh it does it, it helps you do that uh dramatically faster.
• 6:44
  SPEAKER_00: So it's really, I mean, initially it's going to be able to speed up the process for developers.
• 6:51
  SPEAKER_00: I mean, I know you operate a lot on the acquisition side, and so when developers are looking for land or they're trying to find the right piece of land, and I've always wondered this as well.
• 7:00
  SPEAKER_00: You know, when you have developers that operate in a single market, right?
• 7:04
  SPEAKER_00: They're they're in a single market, they know the market really well, they probably know all the farmers and they know they know who owns all of the land and everything.
• 7:12
  SPEAKER_00: I've always wondered what what benefits are there to those developers who already have a feel for their particular market in incorporating AI into their workflow at whatever level that may be.
• 7:26
  SPEAKER_01: I think to put it most simply, uh it it helps you say no faster, right?
• 7:31
  SPEAKER_01: Any of us that are deal people uh recognize that most of our time is spent on deals that ultimately die for one, you know, for a large host of potential reasons.
• 7:40
  SPEAKER_01: And what um AI can help you do, and in particular if AI is equipped with the right data and geospatial technology uh we we provide, is it can find the no really, really fast.
• 7:53
  SPEAKER_01: So even if you are working in uh Mustang, Oklahoma, and you know all of the uh local folks in that market, by the way, you're probably looking to expand, right?
• 8:04
  SPEAKER_01: Most businesses by nature want to grow, want to go new places, and that is where we are particularly helpful.
• 8:10
  SPEAKER_01: But what that means in the context of AI at this point in time, is I can go into acres and type in like in natural language.
• 8:17
  SPEAKER_01: What are you looking for?
• 8:18
  SPEAKER_01: Right, find me 20 to 100 acre parcels under 25% flood and wetlands with appropriate uh topo for a development.
• 8:26
  SPEAKER_01: Uh, make sure that that's within 1,000 of sewer feet of sewer, and then it has a like it's either zoned for residential office, whatever you're looking for.
• 8:34
  SPEAKER_01: And if it's not, it has a high likelihood of passing.
• 8:37
  SPEAKER_01: And then what AI can do is go out and assemble all that data for you and here, here's 10 parcels you should take a look at.
• 8:43
  SPEAKER_01: So as you can imagine, you know, extrapolate that into uh any industry or any deal type that involves land and maps, and it just saves you a ton of time.
• 8:54
  SPEAKER_00: Do you think this makes it more beneficial or advantageous to the big guys out there versus the small guys?
• 9:01
  SPEAKER_00: And I know the small guys now have access because of price points and stuff, they have access to some of the same tools that maybe only the big guys had access to before.
• 9:08
  SPEAKER_00: But I also you know think about this sometimes that, you know, trying to provide some of these tools to really all these different development groups, but now these groups that are outside of your market, they can hop on these same tools and very easily find out like what's in your backyard.
• 9:24
  SPEAKER_00: Whereas before that was very difficult, time consuming.
• 9:27
  SPEAKER_00: And a lot of times, unless they were pretty committed to going into a market or had people dedicated to doing that, they probably wouldn't do that.
• 9:34
  SPEAKER_00: And it feels like it's gonna become easier for some of these bigger companies to have AI go and look in these different markets where a lot of these uh smaller developers maybe operated and were pretty safe before.
• 9:45
  SPEAKER_00: Do you think this creates a lot more competition coming into some of these markets like that?
• 9:50
  SPEAKER_01: I I do.
• 9:50
  SPEAKER_01: I do not I would say that it levels the playing field, though, right?
• 9:54
  SPEAKER_01: So it's gonna make this easier for everyone.
• 9:56
  SPEAKER_01: And let me make sure to say this out loud.
• 9:58
  SPEAKER_01: Human beings are necessary to do deals, right?
• 10:01
  SPEAKER_01: We're exchanging real amounts of money when you're talking about a land deal.
• 10:05
  SPEAKER_01: Six, seven, eight figures.
• 10:07
  SPEAKER_01: Uh, you're not gonna have a computer or a robot structure that transaction.
• 10:11
  SPEAKER_01: Because by the way, like who's gonna I don't know, there's like real trust involved.
• 10:15
  SPEAKER_01: Like sitting down at the dinner table with a farmer, they don't care what tools you used, right?
• 10:20
  SPEAKER_01: Or or what you know uh nerdy applications you f use to go and find that.
• 10:26
  SPEAKER_01: Uh they care about like who's the human sitting here and do I trust them?
• 10:30
  SPEAKER_01: So uh so like to to call out like does it give any particular players advantages?
• 10:37
  SPEAKER_01: Yeah, any anybody that's using it has an advantage.
• 10:40
  SPEAKER_01: Our goal is very simple.
• 10:41
  SPEAKER_01: Uh we want transparency.
• 10:43
  SPEAKER_01: And that is in our mission statement, it has been since day one as a company.
• 10:47
  SPEAKER_01: It's how we operate internally as humans among one another.
• 10:50
  SPEAKER_01: Uh and the thing that we hope most to bring to the market, which is transparency for everybody.
• 10:55
  SPEAKER_01: Now, what that means is oftentimes one person or the other is not going to get as sweet of a deal as they thought they would.
• 11:01
  SPEAKER_01: But you know, that that means that there's fair val values, there's fair prices.
• 11:05
  SPEAKER_01: So uh look, in in any market, sometimes the seller gets run over, sometimes the buyer gets run over.
• 11:10
  SPEAKER_01: What we think is more appropriate is everybody's equipped with rich information to make the right decisions.
• 11:16
  SPEAKER_00: So you mentioned the human element to this as well, that humans have to be involved in the process and everything.
• 11:21
  SPEAKER_00: And that was something I wanted to talk about today as well, because it's, you know, there's obviously concerns.
• 11:27
  SPEAKER_00: And I think for most tech companies that are thinking about how you're the value you're providing, you also have to look at the other side of it too and think about just because we can do something, should we?
• 11:37
  SPEAKER_00: Does it make sense to do this?
• 11:38
  SPEAKER_00: You know, what and and uh you're seeing that happen at the you know, at the level of all these AI companies are trying to put a pause on things because things are moving so fast.
• 11:47
  SPEAKER_00: And I'm curious on your perspective, when you think about the places in the land development industry where humans, you know, continue to exist.
• 11:56
  SPEAKER_00: It sounds like at the transaction level, obviously, or even sitting down at the table with the farmer, when you're talking about buying their land, the human still has to be involved there.
• 12:03
  SPEAKER_00: It's not a robot that's gonna go do that or some sort of automation that's gonna do that.
• 12:08
  SPEAKER_00: Where do you see, I guess, where else do you see that human element remaining, or where do you think it's important for us to continue to have the human involved in this process?
• 12:18
  SPEAKER_01: Yeah, I think there's there's uh often um uh in media, certainly, there's sort of a false dichotomy or a setup where uh, hey, AI is going to take human jobs, right?
• 12:30
  SPEAKER_01: The other side of that would be AI is just going to make humans much more productive.
• 12:35
  SPEAKER_01: Now, like if you are not using it, you're gonna fall behind.
• 12:38
  SPEAKER_01: That is already happening uh writ large, right?
• 12:41
  SPEAKER_01: And the reality is most businesses are adopting it en masse.
• 12:45
  SPEAKER_01: Lots of them are doing a bad job of it.
• 12:47
  SPEAKER_01: We have done that ourselves.
• 12:48
  SPEAKER_01: We have implemented AI tools that sounded really great, and they were a time waste, they were slop, right?
• 12:53
  SPEAKER_01: So there's plenty of that out there.
• 12:55
  SPEAKER_01: And and by the way, like every guy that looks like me that's out selling data and software now says they are AI, everyone.
• 13:01
  SPEAKER_01: Oh, yeah, and so um you know there's plenty of charlatans about and plenty of mismatched use cases.
• 13:07
  SPEAKER_01: But in general, uh AI is already making uh a lot of us far more productive in our day-to-day.
• 13:14
  SPEAKER_01: Um and so you know I I I I'm not certain that the idea of replacing humans is is as appropriate as is accelerating humans.
• 13:23
  SPEAKER_01: So that means we can all do more, we can do more deals, right?
• 13:26
  SPEAKER_01: As an individual, I undoubtedly can do more today than I can I could two years ago.
• 13:33
  SPEAKER_00: Yeah, and I wonder too, with uh, you know, that race, there's a race that happens, right?
• 13:38
  SPEAKER_00: Because you you do feel that.
• 13:39
  SPEAKER_00: You feel, you know, I see all these other people using it, all these people are doing all these sophisticated things.
• 13:45
  SPEAKER_00: I'm not using it as much as they are, so I need to speed up to hurry up and catch up with them.
• 13:48
  SPEAKER_00: And then they look at me and they say, uh-oh, here they come.
• 13:51
  SPEAKER_00: I better, I better implement even more so I can go faster.
• 13:54
  SPEAKER_00: And uh, and I don't I guess I have some concerns, like in general, with it, just because I I think there's still some uncertainty with the future of of where this continues to go because it does move so fast and everything.
• 14:06
  SPEAKER_00: But I also wonder where that race ends with everybody kind of looking at each other and realizing, and it's legitimate that if you don't use this technology, you're gonna fall behind.
• 14:19
  SPEAKER_00: And like you said, people are already falling behind.
• 14:21
  SPEAKER_00: And so, you know, where does that end for especially the development community, right?
• 14:25
  SPEAKER_00: Who traditionally have been relatively slow to adopt, you know, new technology and everything.
• 14:30
  SPEAKER_00: And so as this comes into place, I've heard from a lot of people who are who have gone all in on it and they're they're doing amazing things with it.
• 14:38
  SPEAKER_00: I've talked to others who are just starting to dip their toes in the water and try to decide how far they want to go with it and everything.
• 14:44
  SPEAKER_00: But do you think it's this will be a healthy race as far as like continuing to push forward?
• 14:50
  SPEAKER_00: Or do you think there's some responsibility from both the developers as well as the tech companies to kind of understand where you're taking AI next?
• 14:59
  SPEAKER_01: I think this was particularly terrifying.
• 15:01
  SPEAKER_01: It's not now like we look, we are adverse to change quite often by nature as human beings, right?
• 15:09
  SPEAKER_01: I'm not sure, I don't know the number of farmers that stood there and said, hey, the tractor is the devil, but a lot of them did, and they got left behind.
• 15:16
  SPEAKER_01: Right.
• 15:16
  SPEAKER_01: And the same is true.
• 15:17
  SPEAKER_01: Electricity was dangerous as well.
• 15:19
  SPEAKER_01: Then the internet was a toy, right, for for nerds.
• 15:22
  SPEAKER_01: Uh and I I think like as time has shown us, like innovation uh is a thing that we should, in fact, uh be adoptive of for very obvious reasons.
• 15:33
  SPEAKER_01: So uh, you know, I are these tools useful?
• 15:37
  SPEAKER_01: Should people be using them?
• 15:38
  SPEAKER_01: Absolutely.
• 15:39
  SPEAKER_01: Um is the pace of development here concerning and uh the the capability sets of these tools potentially concerning as you extrapolate forward like six, twelve, twenty-four months?
• 15:48
  SPEAKER_01: Not not six years, like months.
• 15:50
  SPEAKER_01: Yes.
• 15:50
  SPEAKER_01: Uh there are some real uh conversations that we need to have at a uh at the level of like human race.
• 15:58
  SPEAKER_00: Yeah, yeah.
• 15:59
  SPEAKER_00: And that's yeah, that's kind of where my head is at as well, is just kind of keep an eye on understanding where things are going.
• 16:04
  SPEAKER_00: And and even for the developers out there who are interested in these tools, do want to speed things up, but a lot of them seem to be doing exactly what you're talking about, in that empowering their people, not necessarily replacing their people.
• 16:16
  SPEAKER_00: And sure, there are a lot of tools out there that are advertising, get rid of your people, you know, become a company of one, you know, essentially, because you don't need uh real people anymore.
• 16:26
  SPEAKER_00: But I don't feel I don't see that really happening in this industry.
• 16:29
  SPEAKER_00: I think people are empowering their people, which is really good to see.
• 16:32
  SPEAKER_00: And uh I guess when you think about the ways that developers could start, say they're not using AI really at all today.
• 16:42
  SPEAKER_00: They're not doing much with it.
• 16:44
  SPEAKER_00: Maybe they've they've probably dabbled in Chat GPT and just some some of the uh more of the chat clients that they can use, but they aren't, they haven't really created any processes around it.
• 16:53
  SPEAKER_00: What do you think, or where do you think that they should consider starting in that development process of starting to incorporate some more AI into their day-to-day?
• 17:02
  SPEAKER_01: Yeah, I think before looking at specialized uh tools like like Acres or you know, a myriad of other uh workflow softwares uh systems out there, go get an account, right?
• 17:13
  SPEAKER_01: Go go to Claude or Anthropic and get an account.
• 17:16
  SPEAKER_01: And not a free account.
• 17:17
  SPEAKER_01: Like, look, if at least a few people in your life are telling you there's a world-changing technology out there, even if you disagree with them, go spend$20, not$20,000,$20 single dollars to get a subscription and try this stuff out.
• 17:32
  SPEAKER_01: And and spend an hour of your life, not a month of your life, like an hour understanding how to prompt and how to interact with uh, again, what at least a couple of your friends are telling you as a world-changing technology.
• 17:44
  SPEAKER_00: It's funny you say that too, because I've heard from a number of people in the industry who have done just that and they and they just it just blows their mind, right?
• 17:53
  SPEAKER_00: They do, they just do some basic things.
• 17:54
  SPEAKER_00: I was just talking to uh a broker down in uh Houston yesterday or the day before, and he was explaining to me all the stuff that he was doing with it, and he's just just hearing the excitement from him of some of the stuff that he can do now in a much short, shorter time period to evaluate you know, sites and things of that nature.
• 18:12
  SPEAKER_00: It's cool to hear, and I think it will blow people's mind.
• 18:15
  SPEAKER_00: And yeah, there's in some cases there's some of it, there's still a little bit of a technical hurdle if you want to get all the way down into the code of things, but man, that seems to be that that seems to be a thing almost of the past where people who aren't super technical are getting in there and you'll hear them say, Yeah, I'm in cloud code doing I'm I'm creating apps, I'm creating tools to help me with my business.
• 18:36
  SPEAKER_00: Um is there a certain spot do you think that it makes the most sense to start?
• 18:40
  SPEAKER_00: I mean, you're in the acquisition side, of course.
• 18:42
  SPEAKER_00: Do you think the acquisition side is is one of those areas where people should explore first in the development space?
• 18:49
  SPEAKER_01: Yeah, I think like very simplistically, again, using an off-the-shelf model, like if if you're working in acquisition, uh, well, if you're going off the shelf, you're not going to get great geospatial data and that kind of thing.
• 19:00
  SPEAKER_01: But what you are gonna get is a great research assistant.
• 19:03
  SPEAKER_01: So you got an LLC or a uh landowner you want to go learn more about, just go start talking to the robots about it.
• 19:09
  SPEAKER_01: You'll you'll it's pretty amazing how how much work they can do in a short amount of time.
• 19:13
  SPEAKER_01: Similarly, I'm I'm curious about this city council meeting uh or or these members of city council and how they've been voting lately.
• 19:20
  SPEAKER_01: Uh, you know, what whatever that, let's call it like general research project, whatever that thing is that you do, where you're clicking around and Googling six different times and going to a bunch of websites, at the very, very, very base level of using AI, you should be, you should not be doing that anymore.
• 19:37
  SPEAKER_01: You should just be asking that question one time to one robot friend and allow them to go do your work for you.
• 19:43
  SPEAKER_00: Do you think people are relying too much on AI for overall decisions?
• 19:48
  SPEAKER_00: Meaning like it can be pretty easy, I feel like, and this is just from personal experience, I feel like it can be pretty easy when it gives you an answer and you start to see AI sometimes will tell you what you want to hear.
• 19:58
  SPEAKER_00: You know, it'll even change its answer pretty quickly when you you correct it and everything.
• 20:03
  SPEAKER_00: How much scrutiny should people have when they are interacting with AI and they're getting some sort of results back to make sure that, hey, AI did some due diligence for me, but maybe I still need to do a certain level of it before I just jump into a particular site?
• 20:19
  SPEAKER_01: Yeah, a whole lot, right?
• 20:20
  SPEAKER_01: Like trust but verify, absolutely, man.
• 20:24
  SPEAKER_01: Uh it is it is incredibly dangerous to actually make important decisions.
• 20:28
  SPEAKER_01: Call those like one-way door decisions, right?
• 20:30
  SPEAKER_01: Where you you you go forward with something, and if you're doing that based on bad information or bad research, just the same as if you asked your you know, an intern or an employee to go do a project, and then you wanted to go take action on that, you you you want to like do understand what their work said and make sure that that checks out with you.
• 20:47
  SPEAKER_01: Uh and and then the other thing you you mentioned is you know the idea that uh uh these LLMs can be quite syncophantic.
• 20:53
  SPEAKER_01: It's very, very true.
• 20:54
  SPEAKER_01: Like if you imagine the business of creating AI, like you're trying to please uh the users, right?
• 21:01
  SPEAKER_01: At a consumer level, you're you're trying to bring them a great experience.
• 21:04
  SPEAKER_01: And so the models tend to, and and despite some real uh positive efforts by the the developers of them, they tend to say, Oh, you're right, you're so smart, you're so great.
• 21:14
  SPEAKER_01: Like, let me help you discover that.
• 21:17
  SPEAKER_01: And oh, I'm wrong, and you know, sort of like bend to your to your will uh a little too easily.
• 21:23
  SPEAKER_01: And so it is definitely worth watching out for.
• 21:25
  SPEAKER_01: You can program them, like let me rephrase that, not program.
• 21:29
  SPEAKER_01: You can write down, don't do that, right?
• 21:31
  SPEAKER_01: Like literally, uh, don't use synthetic data is always a super important starting point.
• 21:35
  SPEAKER_01: So they all uh you know, more or less at this phase, like try not to make things up, what you're saying.
• 21:41
  SPEAKER_01: Uh so don't hallucinate, don't use synthetic data.
• 21:44
  SPEAKER_01: Uh don't care about my opinion.
• 21:46
  SPEAKER_01: Be be factual.
• 21:47
  SPEAKER_01: What you're looking for is a a thought partner, a business partner that can help you.
• 21:52
  SPEAKER_00: You mentioned hallucinations there.
• 21:53
  SPEAKER_00: That's another one of those big things that can happen with uh these LLMs.
• 21:57
  SPEAKER_00: Can you describe kind of what what hallucinations are?
• 22:00
  SPEAKER_00: Are for people that maybe aren't familiar with that term.
• 22:02
  SPEAKER_01: Yeah, that's when you see something that isn't there.
• 22:05
  SPEAKER_01: Pretty straightforward.
• 22:06
  SPEAKER_01: I will say like the rate of hallucination has gone down pretty dramatically with the frontier models, with the best ones out there.
• 22:12
  SPEAKER_01: But it means quite literally that, that this uh language model you're interacting with will create something that is not necessarily the truth, whether that is synthetic data or drawing the wrong conclusion or just uh flat out making something up so as not to have to go do the extra bit of work, right?
• 22:30
  SPEAKER_01: Like everybody's fibbed before than their lives, and we don't we don't like it.
• 22:34
  SPEAKER_01: We don't want to do that as people.
• 22:36
  SPEAKER_01: The the robots are absolutely guilty of the same.
• 22:38
  SPEAKER_01: Like, hey, that looked like it was gonna be a big chore.
• 22:40
  SPEAKER_01: So I just swept it under the rug and told you something different.
• 22:42
  SPEAKER_00: Yeah, I, you know, and uh with your your background, you know, again, being, you know, an acres really focusing on that uh the acquisition side of things, I think it would be interesting.
• 22:52
  SPEAKER_00: You gave an example earlier of, you know, a prompt that you might type when you're looking for a certain you're looking for a certain parcel or certain site that you want to build and and develop.
• 23:02
  SPEAKER_00: I'm wondering if you could walk through an example, another example of maybe how developers could use AI in that acquisition side of things, and uh really just kind of dig into that so they they get a feel for okay, that makes sense.
• 23:19
  SPEAKER_00: I I do it this way today.
• 23:21
  SPEAKER_00: This is how doing something with AI could potentially speed up that process for me.
• 23:26
  SPEAKER_01: Yeah, the easiest thing you can do ever is ask it that question.
• 23:31
  SPEAKER_01: Hi, I'm in land development.
• 23:33
  SPEAKER_01: I work in this region.
• 23:35
  SPEAKER_01: This is what I'm looking to do.
• 23:36
  SPEAKER_01: How can you help me?
• 23:37
  SPEAKER_01: And you will be pretty surprised at the the menu that it gives you back of things it can help you do.
• 23:43
  SPEAKER_01: So it is always something just encourage having very open conversations with whichever LLM you're you're using.
• 23:51
  SPEAKER_01: And and it to make sure to caveat what you said with my experience, like I am not a technical person.
• 23:56
  SPEAKER_01: In fact, I may have hallucinated some of the things I've said today.
• 23:59
  SPEAKER_01: Uh, right.
• 24:00
  SPEAKER_01: So this is uh we have very real uh data scientists and AI experts on staff.
• 24:06
  SPEAKER_01: Uh I am I am not that deep in the weeds to actually know the technical specs here.
• 24:10
  SPEAKER_00: No, no, that's fair.
• 24:12
  SPEAKER_00: That's fair.
• 24:12
  SPEAKER_00: Um, but obviously you're you're involved in it day to day with your team and everything, so uh you definitely get a lot of exposure to it.
• 24:19
  SPEAKER_00: And you know, the the piece we were talking about with the data coming back to the developer and then the trust but verify right related to what they're getting back, is that just a matter of I'm trying to think through this, right?
• 24:33
  SPEAKER_00: Because if they have, let's say that a developer has 10 steps they normally do when they're doing site selection, they're trying to find a site, each site they do these 10 steps, whatever they are.
• 24:44
  SPEAKER_00: And now they're going to have AI step in and help with, I don't know, X number of those steps up front.
• 24:54
  SPEAKER_00: But I would assume, based on what you're saying, there's still going to be maybe step number four.
• 24:59
  SPEAKER_00: You don't have to do 100% of it, but you probably should do, you know, maybe 20% of it just to do some validation.
• 25:05
  SPEAKER_00: It's not, we're not literally outsourcing everything to AI like we talked about earlier.
• 25:10
  SPEAKER_00: You need to figure out the steps that are the most consequential to the business and your decisions and don't just basically say, AI said this is a great site to develop this type of uh, you know, to do residential here or to do um, you know, an industrial site, but it maybe overlooks something and you are able to catch that at that step number four or whatever.
• 25:28
  SPEAKER_00: Is that how you would think about this, or is there a different way you'd approach it?
• 25:31
  SPEAKER_01: No, I think I think that's incredibly appropriate, right?
• 25:34
  SPEAKER_01: Uh, which is just go go find the uh, what is it, like the highest effort, lowest value components of the job and start with those, right?
• 25:44
  SPEAKER_01: Like the things that take you a lot of time.
• 25:46
  SPEAKER_01: You get a broker packet in your email, right?
• 25:49
  SPEAKER_01: Like it's six pages long, and you don't even know if you should be spending time on this or not.
• 25:52
  SPEAKER_01: Drag and drop it into a model to say, hey, verify this.
• 25:55
  SPEAKER_01: Is this appropriate based on what you know about me?
• 25:57
  SPEAKER_01: Is this an appropriate thing for me to go spend some time on?
• 26:00
  SPEAKER_01: Uh you know, what whatever that sort of like you know what I mean?
• 26:02
  SPEAKER_01: Like there, there's parts of all of our jobs that are yeah, high effort, they take a lot of time, but they don't really do a whole lot positive for us.
• 26:09
  SPEAKER_01: Um research is what I mentioned earlier, uh, where uh the front-term models are incredible.
• 26:15
  SPEAKER_01: You know, hey, I want to go learn about this company, I want to go learn about this person, this process, this competitor, whatever it may be.
• 26:22
  SPEAKER_01: Boy, uh do you th that could be uh a one to ten hour rabbit hole for you, and uh that is no longer the case.
• 26:30
  SPEAKER_01: Should not be the case for you.
• 26:32
  SPEAKER_00: Yeah, no, I like that.
• 26:33
  SPEAKER_00: And I think you said it earlier about, you know, you're thinking of the you're thinking of AI as a almost like a junior team member as far as but they're they're super thorough.
• 26:44
  SPEAKER_00: They know everything, you know, they know a lot, they have access to a lot of information, but they will make mistakes.
• 26:49
  SPEAKER_00: And so I think that is probably one of the more important things to get across to people listening in right now is that, yeah, we should definitely be leaning into these tools and and making use of them.
• 26:58
  SPEAKER_00: But yeah, definitely um still use you still need to use some of your intuition and and some of your your knowledge that you've developed over the years to continue to um make sure that you're getting into the right deals and everything.
• 27:10
  SPEAKER_01: Yeah, and I think that that extends quite a bit, uh, especially I'll use an example in the world of communication.
• 27:17
  SPEAKER_01: So, you know, people are now with quite a bit of regularity using AI to draft their emails or their sales emails or their responses to their actual friends.
• 27:27
  SPEAKER_01: Um, you know it's kind of icky, right?
• 27:30
  SPEAKER_01: Like we we've all got those emails.
• 27:31
  SPEAKER_01: It's like, oh, this feels this feels very, very generated by a robot.
• 27:34
  SPEAKER_01: And so uh it's it's a great example where AI models can help you to understand the framing, the context.
• 27:42
  SPEAKER_01: Hey, what's the history of this conversation with this person?
• 27:46
  SPEAKER_01: But maybe let it give you some bullet points.
• 27:48
  SPEAKER_01: But like write the email yourself, write the communication, like you know, be human.
• 27:53
  SPEAKER_01: I think uh we're we're all very tired of the amount of uh slop we see out there in the world, slop being like AI generated content that just isn't helpful, just a lot of words.
• 28:03
  SPEAKER_01: Um I think that is a great example of this use case.
• 28:07
  SPEAKER_01: Like it is actually helpful.
• 28:08
  SPEAKER_01: It's it can be damaging to your business or to your your uh need your desired outcomes to use AI slop.
• 28:15
  SPEAKER_01: And it can be very helpful to do the inverse and be the person who actually still writes emails with their hands.
• 28:20
  SPEAKER_00: It's funny you say that too, because I can't tell you how many emails I get every week of people like wanting to be on the podcast.
• 28:26
  SPEAKER_00: It's usually like PR firms that are they're probably mass sending these out to different podcasts and stuff, and they all sound the same.
• 28:32
  SPEAKER_00: Like everyone that comes to me and they all they all follow the same format.
• 28:36
  SPEAKER_00: It's like, tell tell me something about a recent podcast episode, you know, a couple of things about the podcast episode.
• 28:42
  SPEAKER_01: It was so intriguing when you said this.
• 28:44
  SPEAKER_00: Yeah.
• 28:44
  SPEAKER_00: It's like you didn't listen to my podcast.
• 28:46
  SPEAKER_00: Come on, stop, stop making that up.
• 28:48
  SPEAKER_00: And so you know right from the beginning that it's uh that it's not they didn't write it, and it's probably some sort of mass thing that's going out to several podcasts to try to get the guests they're representing to get on the show and everything.
• 28:58
  SPEAKER_00: But you're right.
• 28:59
  SPEAKER_00: Like peop if people would actually take the time to write something themselves and be real with it, they probably would get people's attention more.
• 29:07
  SPEAKER_00: I know they'd get my attention more for being on the show.
• 29:09
  SPEAKER_00: I mean, I it stands out night and day for sure.
• 29:13
  SPEAKER_01: I I will intentionally I'm I my grammar's not amazing in the first place, but I will intentionally use bad grammar in emails, like you know, Google will suggest, oh, change it to this and uh leave it alone.
• 29:22
  SPEAKER_01: Like let somebody know this is a human.
• 29:25
  SPEAKER_00: Yep, yep.
• 29:25
  SPEAKER_00: And I've been guilty of using, I mean, going down that path of using AI to like help with emails and stuff like that.
• 29:31
  SPEAKER_00: And then you get to a certain point, like like you were talking about, and you're like, man, I need to reel this back in.
• 29:35
  SPEAKER_00: This isn't even this isn't even me anymore.
• 29:37
  SPEAKER_00: Like, what's what's happening here?
• 29:39
  SPEAKER_00: And I do sound like everybody else, and I don't wanna I don't want to be like that.
• 29:42
  SPEAKER_00: And I mean, I had a point in time, and I'm not ashamed to say this, but I had a point in time where even my podcast newsletter that I was sending out, like I was having AI help me like structure everything, and I finally got to the point where I was like, I have to write this myself.
• 29:56
  SPEAKER_00: I can't have AI help me with it.
• 29:57
  SPEAKER_00: This is gonna be my words, it's gonna be my writing style, and I'm gonna do that, even if it takes me, you know, it doesn't even take you that long either.
• 30:03
  SPEAKER_00: That's what's funny, right?
• 30:05
  SPEAKER_00: Yeah, you can click a button and it can generate it in an instant, but maybe it takes you five to ten minutes, depending on how thorough you're getting on certain things.
• 30:12
  SPEAKER_00: But yeah, I think I think that's really good advice from your from your standpoint.
• 30:17
  SPEAKER_01: Yeah, look, I you know, there there's a time to step on the gas and a time to not, right?
• 30:22
  SPEAKER_01: And so just kind of flutter that that break and that gas pedal appropriately through your professional life.
• 30:27
  SPEAKER_01: And uh again, I I would consider human-to-human consideration as uh conversation, sorry, and communication, as relatively low effort and relatively high value, right?
• 30:37
  SPEAKER_01: So that's something you want to be a little more cautious about, outsourcing versus these ultra time-consuming tasks uh that that you can in fact stop doing.
• 30:46
  SPEAKER_00: Yeah, do you think I'm sure you think a lot about this just based on the nature of your company, but something that I think a lot about is the the price that these AI companies charge today and the fact that it is so cheap for what people are getting, and thinking about when does this go up?
• 31:04
  SPEAKER_00: Like when does it, you know what I mean?
• 31:06
  SPEAKER_00: When does it get to the point where this$20 plan or this uh$200 plan on cloud code turns into maybe the fourth of what it used to cost to hire an actual person?
• 31:16
  SPEAKER_00: Where a lot of big companies will easily justify that.
• 31:19
  SPEAKER_00: I mean, a big company could come in and say, Yeah, I'll pay$20,000 for an AI agent to go and operate, you know, 24-7 and work, whereas a small company is going to say, I can't spend$20,000 on this, so I've got to go back to this other older model and use this, and I can't truly keep up with the big guys at this point.
• 31:36
  SPEAKER_00: Um, do you think that's something that could be coming at some point in the future, or do you think they'll try to keep it, you know, real relatively accessible for companies of all sizes?
• 31:45
  SPEAKER_01: You know, it's it's really hard to see around corners with this stuff.
• 31:48
  SPEAKER_01: It's hard to see like a few months in the future, much less much less a few years.
• 31:52
  SPEAKER_01: Uh what we've seen so far would suggest the other end of the spectrum, and I hope that continues, right?
• 31:57
  SPEAKER_01: Which is that uh while the cost of compute is is uh quite quite variable these days, uh generally speaking, the the models are becoming more efficient, at least like at a at a uh a static level of uh functionality, right?
• 32:11
  SPEAKER_01: So I can get a six-month or a year old model at a pretty sweet price these days, and it does a lot for you.
• 32:17
  SPEAKER_01: So I I I'm not can I'm not I'm actually convinced that it it comes our direction further, um assuming the the path we're on.
• 32:28
  SPEAKER_01: The reality is these companies have terrible, terrible margins, right?
• 32:32
  SPEAKER_01: And uh the competition is being venture funded for the most part, right?
• 32:37
  SPEAKER_01: So like are there price raises potentially in the future?
• 32:39
  SPEAKER_01: Sure.
• 32:40
  SPEAKER_01: Boy, you know, price to value on these things is pretty skewed in your favor right now.
• 32:45
  SPEAKER_01: Uh the the the hope is that everyone has access to full powered models uh in the future, and that that helps everybody out.
• 32:52
  SPEAKER_01: Uh, but we'll we'll have to see where things go.
• 32:55
  SPEAKER_00: Yeah, well, this next one's gonna be a little bit difficult to answer because the you know, it is hard to see around the corner and see where we're headed.
• 33:02
  SPEAKER_00: You know, we're we're talking in terms of the next 12 months or 18 months because things are changing so fast.
• 33:06
  SPEAKER_00: But, you know, when you think about the land development industry and how quickly things are changing, do you have any gut feel for where we might be in five years from now?
• 33:16
  SPEAKER_00: And I know in the day of AI, you know, five years is seems like forever, but like do you have any feel for where this is eventually going?
• 33:23
  SPEAKER_01: You know, I think the the great news is is um we will continue building things, right, as as people.
• 33:29
  SPEAKER_01: And so uh that is uh perhaps a happy in the world of uh uh versus if you like me and you work in software, uh there's some existential threats fighters out there uh that you know may or may not uh uh come to pass.
• 33:43
  SPEAKER_01: So in the world of land development, I I think that demand is a thing that we've become used to for the most part, a few blips along the way.
• 33:52
  SPEAKER_01: Uh it feels like that's a pretty solid foundation for for uh the next five years.
• 33:57
  SPEAKER_01: How we do our jobs is definitely going to change.
• 34:00
  SPEAKER_01: It already has meaningfully, uh, and it will change a lot more in the next five years than it has in the last couple.
• 34:06
  SPEAKER_00: Is there anything else uh related to AI you think people in this industry uh should be thinking about?
• 34:12
  SPEAKER_01: Um I I think what we discussed earlier is like that price-to-value continuum, right?
• 34:17
  SPEAKER_01: Like like the frontier models are incredibly inexpensive.
• 34:19
  SPEAKER_01: Like Googling something and clicking the AI button is is not really using AI, right?
• 34:23
  SPEAKER_01: Like go get a login, go go use a real model, give them your 20 bucks, you know, try it out for a little while to go discover like where that value really uh can be and how how far that can go.
• 34:35
  SPEAKER_01: But say like the other side of that, right, is again there are a lot of people that look like me that you know talk about AI and talk about the tooling they have, and they charge lots of money or they sell things that are uh claim to be AI and they're not.
• 34:51
  SPEAKER_01: Uh so like at the end of the day, you you have to ignore salespeople, right?
• 34:55
  SPEAKER_01: Ignore uh the folks trying to get you to buy their product and just go discover like, does this bring me a lot of value?
• 35:01
  SPEAKER_01: Uh and again, I think the frontier models are like the easiest place for you to start that journey.
• 35:05
  SPEAKER_01: And you can dig into specific applications, right?
• 35:08
  SPEAKER_01: And like market intel and land act it could be uh acres or other solutions out in the market.
• 35:13
  SPEAKER_01: Uh, and that goes through all the way to you know horizontal development, monitoring, vertical construction.
• 35:18
  SPEAKER_01: There's a lot of really cool stuff happening right now.
• 35:21
  SPEAKER_00: Well, I on these episodes, I usually do a segment at the end called the Lightning Round where I have five questions that uh we dig into.
• 35:27
  SPEAKER_00: I think we're gonna have, and I lately it feels like because the conversations kind of go all over the place, I don't necessarily just have a list of questions that we go through sequentially.
• 35:36
  SPEAKER_00: There's gonna be a little overlap here, but I'm gonna still walk through each of these questions.
• 35:40
  SPEAKER_00: And uh uh first one for you, what is one task you would never completely turn over to AI?
• 35:48
  SPEAKER_01: Uh my one-on-one meetings, right?
• 35:50
  SPEAKER_01: Both both with people I work with uh as well as with my family.
• 35:54
  SPEAKER_01: And it's just that's uh be humans, come on.
• 35:57
  SPEAKER_00: Yeah, well, and I have to add something to this because it's pretty funny.
• 36:00
  SPEAKER_00: I feel like lately I've been reading a lot of people online talk about how they uh like their performance reviews at the end of the year have all been written by AI.
• 36:10
  SPEAKER_00: So like their managers are literally it, and so they they read it and it's like no, there's no, it's all generic, you know, feedback that they're getting, and it wasn't even any thought put in by their leader to actually help them, you know, progress in their career and everything.
• 36:23
  SPEAKER_00: So I feel like that's very much in sync with what you just said there as well.
• 36:27
  SPEAKER_01: Yeah, yeah.
• 36:27
  SPEAKER_01: There's a simple underlying message there.
• 36:29
  SPEAKER_01: Care about the people around you, right?
• 36:31
  SPEAKER_01: Like show them some respect and love.
• 36:33
  SPEAKER_00: All right, second one.
• 36:34
  SPEAKER_00: Do you think that AI will make land development easier, or will it mainly make certain parts of it faster?
• 36:41
  SPEAKER_00: It has already made it easier.
• 36:43
  SPEAKER_00: All right, third one.
• 36:44
  SPEAKER_00: What is one skill young developers should make sure they continue to learn even as AI gets better?
• 36:51
  SPEAKER_00: Negotiation.
• 36:52
  SPEAKER_01: That's that like super, super core business skill and how to have empathy towards the person you're negotiating with and how to approach those conversations uh respectfully and appropriately.
• 37:03
  SPEAKER_01: Uh that is a a critical skill that you've gotta hone over your years.
• 37:08
  SPEAKER_00: All right, fourth one.
• 37:10
  SPEAKER_00: What is one question about AI that you think every land developer should be asking right now?
• 37:15
  SPEAKER_01: What can it do for me right now?
• 37:17
  SPEAKER_01: Like we've talked about earlier.
• 37:18
  SPEAKER_01: Like you should just ask the AI that question, uh, but absolutely um uh dig in.
• 37:23
  SPEAKER_00: Yeah, I like that.
• 37:24
  SPEAKER_00: So you guys listening in, make sure you go ask AI that question and the paid version of AI, like we've talked about.
• 37:29
  SPEAKER_00: Ask that question, and you'll maybe it'll open up the door to several different things it can help you with.
• 37:34
  SPEAKER_00: All right, let's see here.
• 37:35
  SPEAKER_00: Last one, what is something developers do manually today that you think will seem crazy 10 years from now?
• 37:42
  SPEAKER_00: And 10's probably way too far in the future.
• 37:44
  SPEAKER_00: You can say five years from now, two years from now.
• 37:46
  SPEAKER_01: Most of the things that you do today will feel bizarrely redundant and old school in five years.
• 37:52
  SPEAKER_01: Most.
• 37:53
  SPEAKER_01: Uh so that that I'll leave it at that.
• 37:56
  SPEAKER_00: All right.
• 37:57
  SPEAKER_00: Well, I think that brings us to a close here, Carter.
• 38:00
  SPEAKER_00: I appreciate you hopping on and sharing uh sharing your knowledge and everything on this industry on the industry and everything.
• 38:05
  SPEAKER_00: Uh, what's the best way for people listening in to connect with you and learn more about Acres?
• 38:10
  SPEAKER_01: Uh, LinkedIn is great.
• 38:11
  SPEAKER_01: You can always um shoot me a note uh, and our website is acres.com.
• 38:16
  SPEAKER_01: So we were pretty easy to find.
• 38:17
  SPEAKER_00: Perfect.
• 38:18
  SPEAKER_00: All right, guys.
• 38:19
  SPEAKER_00: If you look in the description for this episode, you'll see a link to the show notes page, and on the show notes page will be uh those links there to LinkedIn and also to the Acres website.
• 38:27
  SPEAKER_00: But as Carter said, it's very easy to find the Acres website, so you shouldn't have a problem there.
• 38:32
  SPEAKER_00: Um yeah, thanks again, Carter, for joining.
• 38:34
  SPEAKER_00: And guys, if you not are all if you are not already subscribed, please click that button.
• 38:39
  SPEAKER_00: We'd love to have you back for the next one.
• 38:41
  SPEAKER_00: Otherwise, we will talk to you all next week.