#353 - Chris Powers & Jason Baxter - Foster: How Fort Is Harnessing The Power of Advanced Data & Artificial Intelligence

Chris Powers: OK, exciting episode today, a lot going on, and I'll preface this by saying if you go to the show notes, you're going to see a link to a Wall Street Journal article that came out today around how Fort has thought about and is using AI and not just talking about it, but actual practical uses. And today's episode is going to talk a lot about how we got here.

And then what we're actually doing. But before we talk about what we're actually doing, it's essential to know how we got here. If anybody is listening for the first time, you probably know about Fort, or you can find out more about us. But to date, we are known to the world as a commercial real estate investment company based in Fort Worth, Texas.

We have been focused on buying industrial real estate throughout Texas and the broader Southeast, the Sunbelt; we've done over 2 billion transactions. But what is most interesting for today is where we're going and how we're using technology, data, and AI to advance the company into the future.

So, let's start with a look back to 2016, when Jason and I partnered. We met in 2014 and really started working together, and later that year, early 2015, was one of the things that was really important to us. It was technology, and I don't think we fully knew what it was going to be, but we just knew where the world was headed, not just in 2017 and 18, but we were really looking out at where is the world going to be a decade from now.

And a few things came to mind. One, we were really fortunate to go to a place out in Silicon Valley called Singularity University. At the time, we thought we were going to like tech nerd camp, but there were a few things that came.

Jason Baxter: Which was true.

Chris Powers: It was true. We were not going there looking for the idea.

This was early 2018, but things were being talked about there around AI, how AI would be used, and the need for data—not just data, but structured data. I remember we came back and thought, if this is even halfway true, it's going to be world-changing. And we weren't the first people to believe this, but to be a real estate company in Fort Worth, Texas, especially in an industry that's been archaic.

We thought, man, if we could do this, we could leapfrog a lot of the competition, a lot of the folks that we compete against. And so, in 2018, we said we needed to start planning for how this might work. The second thing is interesting. Something that we had talked about, irrespective of AI, which was, we always thought it was bizarre that let's take a property, for example, anybody listening to this in the real estate industry, we'll call it an address, one, two, three main streets in any company that address lives in so many places, and I'm going to leave out.

A few, but they live in your email. It lives in your Slack channel. It lives in your accounting software. It lives in your property management software. It lives in your investor management software. Your analysts are doing models on that property. Marketing has flyers and things going on. You have a whole folder system where you save stuff, and nobody really knows what was saved.

You can go on and on. My point is that address lives in so many different places and most companies. And when I say most, I mean virtually all; they don't talk to each other. So what you have is this mess of data that isn't being aggregated into one place. Different people in the company use different data sets and make decisions based on other numbers.

The leadership team hopes everybody's advancing the data at the same pace so it all lines up. And spoiler alert, that never happens, at least not across the company. It might be for one data set, but as a company grows, there are tons. And so we also said, okay, we need to either find something or build something that aggregates data into one single source of truth.

Jason will talk more about what one source of truth means and how we're using AI. But I wanted to set the context that we're a real estate company that started having these visions six to eight years ago. As we sit here today in 2024, it's all coming to a head, and it's all starting to work after six years of not just financial investment but thousands of hours of Thought, making mistakes, learning from them, iterating, and growing.

Today, it is essential to start releasing to the community that follows us the tech and investment management worlds and how we've thought about these things. Jason, I'll start with you. Let's start with what one single source of truth means so that we can talk about how AI is going to interact with that source of truth.

Jason Baxter: Well, you just teed it up. It's what it is. It's all those different pieces that live in those other places. That same thing is living in everybody else's different world inside a company, like the address one, two, or three Main Streets. It is the thought, how do you get that before you ever act on one, two, or three Main Street, right?

You know, that is an opportunity in the world. How do you first identify it and put it in place in a system? Right, and in a database or an operating system where everything that happens from that moment forward, the day you decide that that's an opportunity in the world and you bring that into that data into your system, and that can be anything.

We're talking about property, but it could be a client, a business, or anything else —any opportunity in the world that you want to bring into a system. Can you start with it? In a place where everything happens from that moment forward inside the company, every process, document, communication, and action is taken on that investment idea through reporting.

Through communication, exiting, and selling it potentially in the future, whatever action happens, can you collect all that information in one single place? Can you connect that data in a way that everything is in one place so that you have a single source of truth? A concept that a lot of people do understand and do dream of, and hope of. The hard part is you have to start way before you even think about doing something.

So, it starts from the external world. What are you going to do? Then, you think about where and how you are going to get that information. What are you going to do with it? And then, once you get it, how am I going to connect everything inside of a company to that thing before it ever gets there? So that's what we've done.

So, we've created a robust database that is very structured and designed to do just that. So, we started with that concept of how we structure our data in a way that we know we are going to have a single source of truth from this point forward. And it's a painful thing if you're starting over.

We were fortunate enough to start it early enough. In the life cycle of our company, we were already trying to find solutions, so we had to build our way into it. Our advantage is that we started so long ago, and we didn't know any better, right? We didn't have to switch from a legacy system.

We didn't have to reinvent the wheel. What we got to do was start fresh. And we started at a time when we were lucky. We were able to start at a time when the technology was advanced enough for us to have the tools and systems available to us to build the right thing and be prepared for what is today, which is what we're going to talk more about as AI.

Because a few years prior to that, the tools weren't there to set your data up, which is why other companies didn't have it. You couldn't set it up in the proper way to take advantage of what AI is today. And that's what we learned at Singularity. The one thing that I will add to the singularity concept, or the thing that we went to, is that Silicon Valley was the singularity, which is the idea that, at some point in the future, artificial intelligence equals that of a human. That is the singularity. That means that AGI is as intelligent as anything in the world, including humans. And that's a math equation that people have come up with.

We're on track to do that, and so I only knew that after going to Singularity. At Singularity, we learned about the future of the world, whether it's robotics or food production, but the one common theme was data, how you're going to organize it, and how you're going to use AI to take advantage of it.

That was the common theme. That's what we left there with. That's how we thought about this single source of truth: That is what data is going to be in the future. You've got to get it all in one place and have an apparent view of it. 

Chris Powers: So, I am still determining the exact date, but let's say early 2023.

Chat GPT there came out in late 22. Early 22, that's when the kind of.

Jason Baxter: Yeah, it had been going for a long time. And it hit an inflexion point. 

Chris Powers: Yeah. It hit the inflexion point where it hit the mainstream. Let's start with how you began to build the thesis for how we would use AI.

Jason Baxter: So I think there are two phases of it, correct? I think the early phase, which is what we talked about, that single source of truth, we knew that there was going to be a moment in the future when we would have our data in a place to take advantage of it differently. Right, and we knew large language models existed back then.

They were just very different than they are today. They had yet to make any real progress. But we knew something was going to happen. So there was that phase of saying, we know this is going to happen. Then, the second phase was as we started to get closer to things like chat and GPT; we started in, this is more our internal technology team, and Greg, who's on our team, began to have conversations around.

Things in the technology world, especially in terms of large language models, are starting to shift. We started having conversations internally about what we could be doing now to make sure that we could start preparing for that. And, you know, we were doing.

Some things, but the inflection point really hit when chat GPT launched; what time was the newer version, the chat GPT 3. 0, right? Or chat GPT 3, which allowed us to test some things with our data to show very quickly that we're prepared, right? And in the moment that happened, we put a plan in place, and this was, yeah.

End of 22. So over a year ago, a year ago now, we said we were going to have what we're going to talk about today: Foster. We're going to have ours ready to go and working and tapped into all of our data before anyone else in the country does, or in the world for that matter, in terms of an internal company having all their data accessible and usable through natural language.

We can talk to anything in our company. Any piece of data, any note, anything, and that was the goal. So we started that a year and a half ago through that process. The whole world was trying to discover. How do you do this? How are you going to connect the data? How do you isolate the data? How do you organize the data?

How do you? Unstructured the data. How do you restructure the data to get the natural language model to speak to it? How do you protect the data so that it doesn't get out to the world? You can't necessarily use chat GPT at the time, and you can't because people want to protect their data. So, we realized all these things were challenges.

And so we spent the better part of a year figuring out the way this is going to happen, not just for us, but how people are going to do this? And, of course, we didn't do it in a silo. We were studying the entire landscape of what AI was and what people were doing. And what we realized was that it was because of the foundation we had built.

We had a unique advantage in that we were able to move very quickly and use some of the thoughts and ideas that were out there in the market about how this would happen. We were able to innovate and create what we believe are unique structured ways to do it that other people may take a long time to figure out or may never figure out.

Whether they like it or it works for them, we know it works, and it works very well. We've been able to move very quickly, and we can talk about some of that more technical stuff, but it definitely gave us an advantage. So, that inflexion point was that leap off the language models to prove that you can take advantage of your data if you have it structured correctly.

Chris Powers: Okay. Let's talk about how most people are using chat GPT and other tools and then how we're doing it. 

Jason Baxter: Yeah. So there's a lot of tools now. I mean, I think the mainstream is still chat GPT, but you know, Google Meta, all the companies are working very hard, and they've all made unique, crazy progress, and it's happening so quickly.

It's such a rapid evolution. And so there's a substantial competitive landscape out there for mass data. They're going to go through all sorts of regulatory things over the next few years. But there's this whole race, right? The mass data thing is a race for the general use of AI and all the things that it can do on a broad scale.

We focused on the fact that it was great. Those tools are excellent. We will use them in some capacity. Everybody will use them in some capacity, but the value inside a company. We talked a little bit about this last time: the value of a chat GPT is not what is the value inside a company when you have your structured data; the use of large language models is, but the way that you think about that is what is critical.

And so, what we have chosen to do, why, and how we do it differently is most people. I'll talk about chat GPT for a second. Most people log into chat GPT, ask questions, and drop in documents. You can now create custom GPTs and do some incredible things. You can only trust and basically organize some of your company's data in a way that's going to give you actual, accurate results and actionable insights and guidance.

As an internal person on your team, by just uploading all your information to chat, GPT is going to try to normalize that you're going to have to train a model on it. And it's going to try to figure out what you're what you want based on everything you've said. So it's going to use it just like general knowledge.

There's value there, but less value the way that we do it. So what we do or what we've done. What we've created is basically the reverse framework of what a chat is. GPT is.

Chris Powers: Was that an ensemble? 

Jason Baxter: Yes. So what we've said is that there are many great models out there right now—open-source, large language models that, in many ways, are more advanced than chat and GPT in some areas, right?

It's chat. Chat GPT is the top and the most dominant one, and it works very well, but other ones have their unique advantage, and they're open source. And so what we've done is we've first created an ensemble, meaning that we've picked at any moment in time five to seven of the best open source large language models that we have structured so that we can play them almost like an instrument.

Right, so they're all located and established in our system in a way that makes them accessible and usable for certain things. And they're interchangeable. So we monitor them, test progress on them, and validate which ones work better for what. As we're using our AI, we're able to basically use whatever's best in the market at any time.

As those improve, we will continue to involve them. So, it's great that we can tap into the best large language models. But we have considered our data instead of training it to become a large language model, and I think that's how a lot of people have talked about it.

We're going to take our data, build it into a large language model, and train that model to understand us. You can't do that. It's very complicated, and it's tough not to have the AI hallucinate. And come up with other things about your company or what could be because it's taking general knowledge.

And so it's tough to isolate and get truth facts, right? And so what we thought about was, well, we already know that the information inside of our system, our Fort Capitol. And everything we've ever done, the financials, the teams, the reporting, the history, every podcast, all the things that we've done.

We know those are facts, though. That's real. That's history. We want to keep that the same. We don't want any new thoughts about what happened in the past. We want it to be facts. And so what we did is said, let's first find a way to structure what is in our system. We call it a corpus.

It's another version of our database that we first unstructured by doing a reverse ETL with a semantic search or retrieval process to create what we call facts. So facts are essentially a series of information that we know are factual, and they're tied together, and it makes sense.

So, think like a property. Or, a history of communication about something or financial data, right? We're able to create facts, and I'm doing the thing. Cause when you see them on the screen, they're literally like factual notes. I want a summary. What we found is that by creating this process of this reverse ETL with the semantic,

We've been able to do that very rapidly in the retrieval process. Once we solved how to do it through our entire database, we've been able to basically unstructure and restructure the data so that it is now organized in a form that we connect to those open-source language models.

Out that we have control over so that they're not connected to the outside world, we can interface those facts through Foster. So Foster is our AI that uses both our corpus, which is facts and all the best, highest-performing large language models that exist today. What you get with that is truth, and you get really.

You get valuable insights, immaculate answers, and, essentially, another team member who is more intelligent than any other team member. That's what you get because it knows everything about Fort, everything about their job, everything about them, everything about the history of the company, and everything about what we're trying to accomplish.

It can do all that very rapidly. So when you ask it questions, like essentially what I'm telling you right now, it would definitely answer them better than I can. It would, but it doesn't. We have many, many tests of it doing it, so that's how we do it.

Chris Powers: Okay, I hope everybody was able to follow along with that.

Let's take it back a step and give some practical examples of what you mean by that. How is the company using it and answering it in a way? For example, when I talk to you, you're sometimes like, look. You're hearing a lot about AI, but for the average company, none of this is actually practical. It's mostly hype.

And then my answer to you is, okay, what are we doing that's not? You'll tell me these things are definitely a practical way to go about your business day to day, but they're different from the way other companies have thought about it. 

Jason Baxter: Yeah, because it's like an evolution, right?

If you sit down and try to imagine what all it can do, or if I'm talking to somebody and they're trying to say, should we do this? And they're like, well, tell me how what it could do. And you give them a few examples. It's hard to imagine what value that really creates for somebody who hasn't even first thought about why that data is essential and how it's structured in that single source of truth, right?

The result of these examples in a one-off standalone, like a use case, is less impactful to telling someone than the whole thing. Right, but there are many, many millions of examples that we can use. We can start going through some of them three or four times. So, I'll speak to the ones that I feel are most valuable to us right now.

And then there are the less valuable ones, the ones that we have latched onto and started using most rapidly. It doesn't necessarily mean that they have the highest value in terms of what we can do with them or what we're doing with them. But we run our executive meetings every week, and we've talked a lot about those structured meetings on this podcast. Those meeting structures are recorded, and they have transcripts, and all that stuff is documented very well in our system.

And so what happens now that we have fostered that? Information is ingested into our database and then fosters immediate learning. When I say immediate, as soon as the meetings are over and we save those meeting notes, Foster now has that as a part of its facts, right? When I leave the meeting, I can ask Foster if you can give me the summary of the meeting.

Other systems can do that, but this is where it's more valuable to us. It knows the people in the meeting, their roles, what everyone is working on in the company, their goals, and their objectives.

So when I say give me the meeting summary, what were the biggest takeaways from the meeting? It knows the biggest takeaways because it knows what we're working on as a company, and it knows which ones are connected to the most significant impact on the company. I could then ask what the most actionable follow-ups from the meeting are.

And it will give me that. Those little ones are like little efficiency hacks. But when you think about that over time and the history of that conversation as it continues to build week over week, and you start to ask Foster, instead of one meeting note, you say, Foster, what has been the most common theme that we've talked about over the past three months.

And what should we be thinking about? It's not that you're going to get an answer that you didn't already know. What you're getting is a good feedback loop, like a super assistant, that can help someone like me in my position think about the whole thing more regularly and consistently and see the world more clearly.

And so that's at my level, at the individual team level or the person level. You can ask even more simple things. What should I work on next? Based on that meeting, what's the most important thing I should do next? You could walk back to your desk and ask that right then. What should I do right now?

Foster will say, based on what just happened in the meeting, this is what was said. It would help if you did this next. It's so accurate because it knows what they're working on. Remember that single source of truth. It knows every task they have in their world right now because it's all inside of our system.

Every job that they're working on knows that. It will allow us to align what we said in the meeting and what we said was most important with what they have in their world today, what we need to accomplish as a company, and the objectives and key results. So when you tie all those things together, it's like a super hack just through a meeting note structure. There are so many examples, but we can take it from the highest level example of financials as a CEO of a company if you want quickly. Information and say, Hey, give me a quick overview of the budget versus actuals last month on expenses across the whole company.

And it sounds wild, but you just asked Foster that and Foster says, here's the overview of the budget versus actual expense over the last month. Fantastic, were there any anomalies? Yes, there were three anomalies. What were they? Well, there was this and this. Please check and see what caused that.

And it will check and see what caused it. Well, how does it do that? It does that because our accounting system is also a part of our single source of truth. Our property management actions every day, as well as the notes, are all part of that single source of truth. So it's only going to one place where all the facts about that asset have been created.

Remember the one, two, three main street, it's all there. And so every expense, every cost, everything, is all in one place. It will go there and say, this is what was different. The crazy thing is that because it is an AI, it can do it across the whole company in milliseconds. As a CEO or even a financial person in the company, you can sit there and check on everything in the company without even having to go to a dashboard.

Chris Powers: Okay. Tell me then. From the more qualitative side, how does it know who our people are? 

Jason Baxter: That's where we started. We talked about this when we were leaking what we were working on over the past couple of podcasts. But I thought, and I still believe this. And this, every company, you know, if I was, if somebody came to me and said, please give us some advice, whether I'm the right person to provide the advice or not, it's another question.

But if they said, what should we think about first? Inside companies, their AI models should first know what the company is, who it is like and why it exists. What is the purpose? What is the mission? What are they doing? What does this company do? And I mean, very specific. What does this company do?

And then the next thing, once you have established that inside your AI, who are the people? If you don't know who those people are, then everything you're trying to get the AI to tell you is irrelevant because it all depends on who the people who are going to be using the AI are.

Chris Powers: And what do you mean by who the people are? What does that mean to you? 

Jason Baxter: Everything that you could imagine about a person, and I'll be specific, so every person has a role, right? Every person has responsibilities. Every person has a level of where they are in their career, right? So a position, a title, all those things. Every person has a profile or some testing metric, right?

Whether it's an Enneagram, a Myers Briggs, or what we use as a culture index, any of those things give insight into who this person is. That person is also a part of their responsibility, and the sort of work, experience, and things they have to do every day are all known. And so you want the AI to know all of that as that person asks questions about their world or that meeting.

Objectives and key results in the company. It is speaking to them, knowing their strengths and weaknesses, what they typically work on every day, what they're trying to accomplish, and where they fit in the company. If you don't establish that, you get a more general AI. And what we're trying to do is create a super teammate for every person.

And it's the same teammate, meaning it's the same person, the same fosters, just one teammate, but it can be everything to everyone, specifically to them. And so that's what I mean. You have to know all those things about the person first.

Chris Powers: And I want to go back real quick. You were talking about facts. How are facts created?

Jason Baxter: The facts are that it's reverse ETL. Essentially, what we've done is build a process. For those who need to learn what ETL means, it's more of a technical term. Anybody in the technical world will see that it's not that special, but it's extract, transform, and load, meaning you extract the data from one place.

It could be from the outside world. In this case, it's from inside our system. You extract the data, transform it into another form or structure of data, and then reload it to where you want it to be. We use that process in a particular way.

This semantic retrieval is where we first program to try to get out of this information that we already know is facts. Since it is already structured, we know and imagine I'm doing it in a dumbed-down way, but I create a box around facts, right?

So we're saying that's a fact. But instead of going into our system and figuring out what all those facts are, we built a way to do it. That reverse extract transform load process allows us to very quickly say, okay, now we want to create facts around this area of the company.

Right? So we go into the financials, and we say, we're going to create facts. Well, we can do it very rapidly now. So, in a very, very short period. As opposed to spending months and months and months taking one mass amount of data, putting it in a system and then trying to train it, right? We want to get it out, call it authentic, turn it into facts and then communicate with it.

Chris Powers: So essentially, the front work we did with FOS building the single source of truth database, if even if you said, I have Jason's brain right now, I know what he's doing without all the data being already structured and years of it in history, you couldn't just start, you'd have to start by getting all the data right first.

Jason Baxter: Yeah. Or you'll have to pay somebody to get your data. That's an excruciating process. And it's tough. And the hard part is, what data are you even talking about? Most people's data could be more structured. It's in files, folders, papers, and mismanaged document management systems, where things are everywhere.

Nobody even knows where anything is. That process most people will go; it's too much. Right. And they stop right there because the level of detail that you have to go through is Pretty in-depth if you want to do it right, right? There's a way to hack it and get a Hodgepodge version of this that works well but is different from the level that I'm talking about.

Chris Powers: If you think that Foster is so brilliant, why don't you ask Foster what Foster believes it is and read it back to me? 

Jason Baxter: So you want me to ask Foster what it thinks it is? And FOS or just Foster? 

Chris Powers: Yeah. And FOS. 

Jason Baxter: Okay. Because I think that is an important, I guess, tie together for those two things is that Foster exists because of FOS.

So, think of Foster as the natural language interface of FOS. So it's all one thing. And now, the way that we think about it, believe it or not, is that FOS is actually fostered. So everything is really fostered because a brain oversees the whole thing. And so all that action that happens inside the processes, streamlining of processes, and all that occurs in FOS is really being observed by Foster.

And so we think of foster as everything. Then, FOS is how we do things day by day. So you want me to hold on, Kate? Let me ask.

Okay, it's going. That's given me a lot. And this is what's excellent about AI. Even when people get AI or establish how to use it, there are still ways to have an extreme advantage. And that extreme advantage, believe it or not, is going to be how good people can prompt the AI to do the thing it wants them to do.

The training of prompting and communicating with AI to get what you want, at least for some time here, may help them figure out a better way to do it. But at least for now, what I have discovered is that the better I can get myself, the better I can show someone else how to prompt engineers with everything they ask, and you get a fantastic result, which is what I just did.

I've gotten good at communicating with foster. So here's a good one or an excellent example that it gave. So I added, can you give me an analogy? So it would make more sense. But here's an analogy of what FOS is: FOS is an operating system, right? And foster advanced AI software that runs on top of it.

So, FOS is our operating system. On top of that, an advanced AI sees everything. It's like an operating system, a general operating system. Like Microsoft or Apple, those are hardware resources. They provide a platform for running different applications, and they can scale that and run as many things on it as they want.

They're all interconnected. That's what FOS is to us. It's a platform for running applications. FOS is in our system, and FOS manages ours. I'm reading this. It is what it wrote: FOS manages our operational resources and provides a platform for running our business processes, just as advanced AI software leverages machine learning and natural language processing to provide intelligent insights and automation.

Our foster leverages these technologies to optimize our operations and drive value to our business. So, we basically have those same general things that are known in the technology world: operating systems and AI. That is what our entire system is. It is one ecosystem of one giant operating system that controls everything, that single source of truth.

And on top of it sits an AI that's fully interconnected. So, and then it kept going. In essence, FOS and Foster represent the next generation of technology, infrastructure, and design, combining the power of AI with the scalability and flexibility of a cloud-based platform to transform the way we work and drive value and success to the business. It's a good answer. 

Chris Powers: It sounds wise to me. 

Jason Baxter: Yeah, and so it's fun to do these types of things because, again, what it allows you to do as an owner or a person in business, that's responsible, right? You can clear your thoughts. You can reframe your thoughts. You can get better insights. You can think about what the company actually is because what it's saying is what it actually knows.

It can examine the system, understand its structure, and describe FOS technically and literally better than I can. So when it talks about it, it always tells the story better than I can. And so that's what you're seeing here. 

Chris Powers: We've talked a lot about this. Let's bring it home by explaining why all this matters and how you see this playing out as we move forward.

Jason Baxter: Well, it's going to be different for everybody for what they're trying to accomplish. For us, what we've built and how we've structured everything, it's apparent that it's going, and it is giving us an opportunity to think about the world differently as we move forward. And it's what we talked about in the very beginning: a single source of truth.

And, but what does that mean for us in terms of how's it going to help us? And what's it going to do is it's being able to spot those opportunities that are out there in the world, any investment thesis, or Opportunity that exists, whether it's class B industrial or anything else, we will be able to see those choose to see them and then see them and then ingest them into what we're doing and then act on them very quickly by either.

The current processes and things that we use or because our system is built so flexibly adapt new processes into our system immediately, very rapidly execute those investment opportunities, and then continue that single source of truth. When you start to do that at scale, where you're doing multiple investment theses across different platforms and collecting all that data in one single place, you get a view of the world that allows you to make excellent decisions but also see opportunities that others can't.

So, it does go back to the single source of truth. In this episode, we talked a lot about little process things and cool things that can be done and the structure behind them. But really, when it boils down to it, it's about how you can make good decisions, the best decisions.

How can you have insight that others can't see? How can you get a worldview that gives you an advantage? And how can you then act on that and execute it at a level that is higher than most? In the market, including workflows, the process is the system. How do you ingest it? What do you do next?

How does it communicate? How does the data communicate with each other? How does it become actionable? How do the decisions get better? How does your team get more efficient? How do the profit margins rise? And how does that continue to? Push the company in a direction that gives them an advantage. And that's what we're doing.

And so it's all of it combined. It's not one single thing. It's the fact that the foundation is there. We will do any of that. But it all starts with the foundation and that single source of truth. Now, we can see an opportunity. Differently than most, but where the real power comes is when you know that opportunity; it's so clear how to do it.

It's so clear how to act on it. It's so clear how it's going to become successful. It's so clear how it needs to be structured. It's so clear the people that need to be involved. It is so straightforward how to set the goals around it. It's so clear because it's all in one place, right? And it just gives you a huge advantage. That's where we're headed with this. 

Chris Powers: The perfect way to tie it up is that it, yeah. Everything you just said culminates in our mission to be the best operator in the world and the world today. It's not just because of interest rates. Things are just more complicated. There's more regulation.

There are more people involved, more technology, and more competition. Operating well has never been more critical. And today proves that we're just one step closer to becoming the best we can be. 

Jason Baxter: Yeah, I agree with that. The reason why it's essential to do these things is the world's going to be more challenging, perceived hard or perceived easy at any moment in time; the truth is it doesn't matter because we can't control any of that.

What we can do is build a framework in a way we operate where the team is more efficient, they're happier because you make it easier on them, and you get smarter every day because of the way that you collect everything. And your world internally gets better, no matter what the outside world is doing.

So the outside world could be crumbling, and things could be getting worse from a perceived, like macroeconomics or interest rates. But we can't control that. But what we can control is how we move inside this machine and how this machine moves through those markets and is the most efficient.

And when you have something like that, you can put a blinder on, and you're, because we're also collecting all that macro data. So we know what's going on. It's not like we're sitting there not aware of it; we're actually getting smarter at that as well. So we're moving through the market very fluidly with zero fear.

Zero fear means, yeah, there's going to be hurdles and risks and all that stuff, but you get better and better at it until you're just moving with confidence every day. 

Chris Powers: All right, man, this was awesome. As always, sometimes I come out of here the luckiest of anybody. If you're listening to this, there is a link to a Wall Street Journal article in the show notes.

That came out today, and I've also included a link to a couple of videos that we made that help visually explain more of what we talked about today. If that's interesting to you, Jason, thanks again for joining me. 

Jason Baxter: I appreciate it.