Seth: Hey everybody, how's it going? Welcome to the REtipster podcast. I'm Seth Williams and today I'm talking with Brandon Cobb. So Brandon has been on the show before. Last time we talked back in episode 237, the conversation was mainly about land development and the business that Brandon was building around that. Since then, Brandon has gone pretty deep into the AI space. And when I say deep, I don't just mean using ChatGPT to write emails or summarize documents. He has built a much more advanced setup involving Claude Code and VS Code and cloud servers and custom agents and automations and a whole system designed to make AI a much bigger part of how a business operates. And I've actually had a front row seat to some of this because Brandon invited me into his community where he's been teaching people how to build and use this type of system. So today we're going to dig into what Brandon has built and why he chose to build it this way and who he thinks should be doing this and where the biggest advantages are to this kind of setup. So Brandon, welcome back to the show. How are you doing?

Brandon: I'm fantastic. Excited to be here. Yeah, it's been fun hanging out with you in the community.

Seth: Last time you were on the show, as I mentioned, we talked mostly about land development. Are you still actively doing that or what has changed in your business and life since mid 2025?

Brandon: Yeah, I know. It's crazy. It's been a year. We're still doing land development. That's our main business that we do. We've got a construction company that, we do some stuff for some other people, but primarily what we do is we take land, we get it entitled, we sell it to big national builders. We've got about an $8 million development we're doing where we're actually putting the infrastructure in the ground right now in a 74 home community. We actually finished it a year ago. So it's been done. We're just selling the lots off to the builders who are buying the lots. So nothing's really changed a whole lot, except, as you know, there's been a ton of change with AI. A year ago when we talked, there were no AI agents. And now, holy crap, you got Muse and GrokBot and Dots and dude, you can't keep up with all the different applications that are coming out.

Seth: It's crazy. Like literally week to week. It's like, oh, that stuff you thought was so great. That was last week. This week is the new thing. I know a lot of people were using AI in some way, shape or form. But for you, I think it was around early this year, early 2026, when I started seeing you post stuff on Facebook, talking about your different agents you were building, all the setup. And I just thought like, whoa, this guy, he's like way ahead of me. Like he's doing stuff. I don't even understand what's going on. I got to learn more about this. So like, when did you decide, okay, I got to double down on this and not just chat with ChatGPT, but like really build something serious. Did you see somebody else doing it or did you like hire somebody to teach you how this stuff works or tell me the origin of how you got into this new realm?

Brandon: We had somebody in our mastermind community that had presented it. And I was like, huh, that's different. And so I started studying. Of course, he opened up everything, kind of showed what he was doing. And then I went on a rampage. I bought every course, every YouTube, followed every YouTube channel, every X account, paid every AI consultant that I could get my hands on. Because this was a time where you had these AI consultants. Everybody was coming out being an AI consultant that knew how to use Claude Code, which is kind of funny because now everybody uses it. But that was what got me introduced to it, and it became nothing less than an obsession because I was actually able to replace people in my business or make the people that were there way, way more efficient. And there's three buckets to this. And for most people, like the average consumer, let's say like my wife or my brother owns a business and he's not a good example. Somebody working in like corporate America, average everyday consumer. The advanced AI stuff, you probably don't have to fool around with it.

Brandon: Using GrokBot or some of the similar stuff is fine. So there's three buckets that I've noticed. Bucket number one is where I used to be pre-March of this year. You're kind of new to AI. You've been using ChatGPT. You've been using Claude. You're kind of learning prompt engineering. You're building your first skill. And that's where I think the majority of people are or were. Bucket number two is your AI aware. So you've probably got 11 different platforms up, Perplexity and Claude and Lovable and n8n and all this stuff. You're using GrokBot, and you're like, holy crap, this is amazing. But you're spread across all these different applications, and you're chasing the new thing that comes out, whether it's Muse or Dots or GrokBot, whatever it is, and you feel overwhelmed, and rightly so. That's the second bucket. Then there's the third bucket that is what I would call the advanced AI guys.

Brandon: And at first, a year ago when this first just introduced, it was only the super, super tech guys, the guys that had been programming, the guys that knew how to speak this programming language and architect things. But as things got better and better, and especially after the launch of Fable, specifically in June, because that specific model was able to do things on a whole other level, like it was just the leaps and bounds. That really was the game changer that most people didn't notice, and that's actually why the government banned it if you were paying attention. This bucket... We're running our own setup. Every new tool that comes out is just noise because we can easily build and recreate it. These guys are running their everything on local, their own computer. Most people, that's not going to work for because they don't have a souped up server or computer or a virtual private server, which is what I do. And that's the best model for most people. And you can get a virtual private server for 60 bucks a month and it's not expensive. All a server is, is a computer in the cloud.

Brandon: It's no different than your laptop that's sitting right there, except when you close your laptop, everything on your laptop shuts down. Whereas a server can run 24 hours a day, seven days a week. And so all of your operations continue running, all your AI agents continue running. So those are the three buckets that I've noticed. And I probably moved into the third bucket, kind of skipped the second. Well, no, I did do the second bucket for a little bit with Manus and everything. But as soon as you discover the third bucket, This is the way, in my opinion, if you were a business owner and you want to truly stay on the forefront of what AI can do. I'm not using Muse. I'm not using GrokBot. I don't have to use 11 different applications. The guys in the third bucket, the main thing that they've done that's different than the guys in the second bucket is they built their brain on the server that they own. So ChatGPT or Claude can't make a rule and change something, and all of a sudden, everything you've built on that platform is at risk and gone. This brain is trained on everything.

Brandon: It's read my emails, my text messages, my calendar events, my sales calls, all my Zoom team meetings. It has built an entire infrastructure and understanding around my business. So when I build an AI agent and I tell it to go and execute a high-level task, like, hey, go do this due diligence on this piece of land, if I were to tell GrokBot to do that, GrokBot might go and do it, but it's not going to get the same result because the context that the brain has when it commands an AI agent, it hits everything, and you're able to get a much higher level output out of it constantly. It's like if I give you my assistant and my assistant's been with me for 10 years, you might have to kind of train your assistant to do a few things, but like she's probably pretty good out of the box versus if you just hire someone 19 years old straight into the workforce out of college. They're going to screw up a lot of stuff. So that's the difference between kind of GrokBot and this.

Brandon: And that's the main difference between the advanced AI guys in bucket number three and the AI aware guys in bucket number two is they've built this brain that's actually controlling driving and even building things that the guys in bucket number two are having to pay to use. So if somebody could come up with something that wired together all 20 of the platforms people in bucket number two are having to use and like bridged it all together and then trained it on everything they were doing, that would solve the problem. It could get them in bucket number three a lot easier.

Seth: This business brain or second brain or the context, like knowing everything about you and your life and your business, like all the stuff. Are you using Obsidian for that or are you using something else?

Brandon: No, I'm not using Obsidian. So it lives on my server. And I think you're talking about like creating the visualization. The brain's there in the cloud no matter what as you're building it. But a lot of these Facebook ads you see of the cool brain that people see, that's just a representation. It's just a visualization of what is actually in the cloud that you're building. And it's really cool. You can kind of customize them and play around with them. I don't know what application I use specifically to build it, but here's the magic thing is like, I just ask my AI brain and it tells me how to build things. I get great ideas online in our community. I'll screenshot it, feed it to my server and say, Hey, let's do this. Do we need to do this? Is this something we should be implementing? And it'll say, yeah, like, this is a great idea. No, it's not. And here's the reasons why, or you want to build a really cool looking brand. Yeah. Here's how we'll do it. Go Give me access to XYZ and I'll make it happen.

Seth: Yeah, it's an interesting thing. Like I've got a few different conversations I'm having right now with Claude Code. For example, one of them is the server that retipster.com lives on. It's been having these like weird issues where it's bogged down. And to be honest, like, I don't understand what's happening flat out. Like I'm, I'm kind of clueless, but I know there's a problem because Claude Code pointed it out for me. And for several weeks now I have been, I'm, asking Claude Code to like diagnose it and then send a support ticket to Liquid Web where my server is to like tell them what the problem is and ask them to fix it. And then they will send some gobbledygook response back to me that I still don't understand. I feed it back to Claude Code, Claude Code tests it and figures out how to reply. For weeks, this conversation has been going back and forth. I have no idea what's going on, but I can see that progress is steadily being made towards fixing the problem. And it's just making me wonder, like, is this dangerous? They're like, I have no idea what's even going on here.

Seth: It's just AI talking to probably some other AI on Liquid Web's side. And because I know there are times when I do know very much what I'm talking about and I ask AI about it and I can see the answer it's given me is like 80 or 90 percent right, but there's a little bit of nuance loss. Like it's not totally getting it. There are places where I would insert different caveats and that kind of thing. So like, I know it's not flawless. I'm just wondering, like in these situations where I don't really need to know because AI just knows it for me. Could that ever come back to bite you? Does that ever happen to you where it's like you kind of thought things were fine and then it turned out, oh, wait, AI was actually wrong about that?

Brandon: I'm trying to think because a lot of the applications that I'm using for is not like, well, I'll ask it for advice on certain things for sure. And I think that's kind of the application you're talking about. These things do break and a lot of it does come back to like human error. So, for example, I built a podcast agent that automated the entire distribution of the podcast. I do the Zoom recording. It looks for the title. It grabs it. It puts the intro on. It inserts the ad. It cleans it up. It enhances the audio. It uploads it to Buzzsprout. It writes the title. It's trained to look at all my other podcasts. Oh, Brandon likes the guest's name and the title. I'm going to put that in there. Oh, in the description of the podcast, Brandon wants a quote from the guest. He wants some bullet points. He wants the guest's contact information. And it nails it. Like I hardly, that's fully automated. And then it'll text the guest once the episode is live. That part actually broke because when I built it, That last little part, hey, text the guest when their episode goes live. It trained it.

Brandon: I trained it, and it did everything it was supposed to. But I woke up to like 250 email responses like, hey, what's going on? What are you talking about? I'm like, what are these emails? It had texted all of the past guests that had already been on the show where their episode was already live from like two, three years ago.

Seth: Yeah, I got one of those emails, by the way.

Brandon: Oops. I was like, oh. So again, that was human error in the training. So there's some things you can do around that, but you're, I would have never caught that what I mean? And so I fixed it and, move forward with it. And you're going to have little things like that break where I... The way I avoid the big pitfalls is I still have to verify that everything is working once it's complete. So a little hack that'll prevent a lot of problems going wrong. Start in plan mode when you are setting things up. Plan mode is different than the automatic edit. When you give it something and say, I want to build this thing, it starts building it. Oh, you get an idea. You got to stop it. Oh, you get another idea. You got to stop it. That clogs up the flow. When you're in plan mode, build out your entire plan. And I'll spend maybe half an hour to 45 minutes if I'm architecting a new build, a new agent, or automating some process. I wanna look at the entire plan that it creates, the entire flow process. And I wanna see if I can find mistakes or things that I missed, quite frankly.

Brandon: That's usually what the mistake is. And I find if I spend more time in plan mode properly architecting, I get a better result. So because I do that, I'm getting much better results and I'm having things break less often or it causing mistakes. So like the other day, like I'm doing a webinar to go over like our whole backend infrastructure here pretty soon. And so I had it build me the entire plan. And I said, hey, go pull all of our Vimeo content, go pull all of our past emails, go pull all of our best performing social media posts. I want you to build me a 12 email campaign, an eight text message campaign. You need to build the drip campaigns if they register what that looks like. The drip campaign, if they don't register, The drip campaign, if they do register, I need you to add images, all of this stuff like this would have, I would have probably paid somebody five or six grand to do all this or. It'd probably take me 40, 50 hours to probably build all this out the right way, create the automations in ActiveCampaign, make sure the triggers and the tags. You know how it is.

Brandon: Did I just trust it to go off and boom, click, yep, everything looks good? No, I had to audit it. I read all the emails. I made sure the text looked good. I looked at the flows and the drip campaigns. It took me about an hour and a half to audit everything. I made a few adjustments. I said, go push it to ActiveCampaign. It built all of the automations out. I used to pay somebody eight grand to build me a new pipeline in ActiveCampaign. I've built two pipelines so far, complete with the drip campaigns, the automation, 16 grand out the door. It did what he would do in 20, 30 minutes. I was absolutely blown away. And when you introduce a team of models working together on a project... So like, I have Fable usually build the code, Astra then checks the code, and for something like a webinar, it's got to tap Nano Banana for the image generation. If it needs to go scrape some stuff, it might use Gemini 2.5 Flash or something like that. And so I'm not using just one platform, Claude, to do things.

Brandon: Your server has access to all of these models if you set it up the right way, so you get the best of both worlds. And when you have a team of these models working together, they're all good at specific things. It's like if you tell the framer to build the whole house instead of the drywall guy and then the plumbing guy and then the HVAC guy. That's how these models work, and there's like 80 other models out there people don't know about. Once you have them working together, you'll have a lot less issues on the back end. But to answer your question is, is this kind of scary? Yeah, it's scary and exciting at the same time, because everybody's idea of AI is what, like when you hear the word AI, like what immediately comes to mind AI robots? It's a question to you, what comes to mind when you hear like AI and robots? Like what, what visualizations do you get?

Seth: Terminator 2, Judgment Day.

Brandon: Terminator. Or iRobot, Will Smith. That's what's in everybody's head. So yeah, it can be scary, rightly so. But I also think that we're on the verge, Seth, of the greatest time period to be alive. We're about to go through a renaissance of cancer and Alzheimer's, healthcare, aging. Oh my gosh. It is going to be amazing of what's going to come out of this over the next 10 years. In the next 10 years, we're going to have more change than in our entire existence alive on this planet. It is going to be absolutely nuts, and nobody's ready for it. And it's not just job loss and all that kind of stuff. I think new industries will be created. But what's happening right now with AI? Astra solving the Navier-Stokes Millennium Prize problem. They gave, there's seven Millennium Prize problems where all the smartest people in the world basically got together and said, hey, we'll give you a million dollars if you can solve any one of these seven problems. AI solved it in two weeks. You've got, was it Google's DeepThink?

Brandon: I think it's DeepThink that solved the protein fold problem and then released all of it for free. The protein fold problem is what's going to create the ability for labs around the world. Now, when you have the shape of the protein, you can create the antagonist to fight it. So if you know what the shape of the protein is for Alzheimer's disease, you can create the antagonist to bind to it, solve that disease. And so like that protein fold problem, they won like the chemistry Nobel Peace Prize for it. Like this is an exciting time. So I'm more excited than I am scared of all this stuff. But that's probably a really, really long-winded answer to, are you scared of AI?

Seth: I think this is a couple of months ago. I think Sam Altman or one of those guys was talking about how a hundred years of medical advancement will be condensed into about five years. Again, that was a couple of months ago. It's probably more like five months now. But I've also heard people talking about how there is a 10% chance of AI killing every human on earth, that kind of thing. So it's like, where's that coming from? Like that can't be totally unfounded. Yeah.

Brandon: You calculate that too, like where'd you come up with that percentage?

Seth: Yeah, I have no idea. It was a little alarming to hear that, but it was on the headlines a bunch a few weeks back.

Brandon: And David Sinclair, he's the leading scientist on anti-aging in his lab. He wrote the book, Why We Age, Why We Don't Have To. Aging is a molecular process that we can reverse. He reversed it in mice. And the first clinical trials are happening right now in humans where they're testing this anti-aging technology on the eye. Now, what the heck? Why would they test it on the eye? Well, they're testing it in blind people. And the reason they're doing that is the same anti-aging technology that can reverse engineer and grow the genes that allow you to stay young are the same technology that can grow the genes regularly. In the optic nerve, regrow the optic nerve in blind people. And what's interesting about the eye is it's very localized. So if something goes wrong in the eye and a blind person, they just stay blind. There's no downside. But if something, if they were to test this technology on the whole body, now you've got more risk there and that's happening right now. He's already taken mice, blinded them, introduced this technology, and made them see again.

Brandon: So we've done it in mice, and it's happening in humans, and AI is the reason he's able to do it. He goes, the testing that we were able to do with AI would have taken us 115 years in my lab. He's like, and what took a year would have taken 115 years. And you see how fast this is going. This is, it's, it is taken off. So again, this is why I come back to the feeling of being overwhelmingly, tremendously excited about the whole thing.

Seth: We talk a lot about agents. And I know so far in this conversation, we've dropped a lot of different terminology that the average person who is not into this whole agentic AI thing yet, they may be very confused by. So in the most succinct way you can think of, what is an agent?

Brandon: This is a great question because most people have it wrong. And I even use the terminology wrong on purpose because if I used it for what most of this actually was, people would kind of get confused. So most people are building are not agents. They're building skills that run on a schedule. And the word for that schedule, if you're a programmer and you want to sound all fancy schmancy, is Cron, C-R-O-N. So whenever you see the person use the word cron schedule, it just means whatever the task is running on a schedule. So the podcast agent, I got to be honest with you and come clean, Seth. It's not an agent. It is just a task or a skill that has been built that runs on a schedule. It is do A, B, C, and D, and it'll lead to this result. An agent is something that actually requires thinking and decision-making to do it. So my podcast agent, it just goes, go to Zoom, grab video, go to Google Drive, grab intro, go to Buzzsprout, upload, take video or, audio software, edit, upload, read past podcasts, create title, read past descriptions, boom, schedule date. There's no thinking involved.

Brandon: It's a very clear workflow. My, I have an agent that has to think of responses on when people text in or they download one of our things and they want to talk to someone in the staff. We've got people that actually speak to people, but some of that, like the beginning stages are AI agents. When you get a response that comes in, it has to read that response. It has to go look at the sales strategy. It has to understand who they're potentially dealing with, and it has to come up with a response. That's an agent. It actually has to think and do the decision-making. And this is where the magic of having your server with your brain comes in. Because if you're using GrokBot to do this or Dots or Muse or whatever, there's no context. That's the problem. My brain has every single coaching call I've ever done recorded. Every single one. So when someone asks a question, it goes and hits that brain, all those calls, and it goes, oh, here's how to answer this question. There's some specialized knowledge here that I need to introduce.

Brandon: Just being plugged into GrokBot or Dots or whatever can't do that. And so that's where the magic happens for the people that are in that third advanced AI bucket is it can actually do things on a level that the, people in bucket number two, AI aware people, can't do because it has that context and that training.

Seth: Are these agents continuously running in the background or do you have to manually trigger them to say, hey, agent, go and do this for me now?

Brandon: So that's where the cron comes in. So that cron, when you put it on a cron, you put it on a schedule. So one of the agents that I built is a fixer upper agent. So its job is just to scan all of them every five minutes. If something happens to break or there's an issue or something goes down, it automatically fixes it. And if for some reason it can't fix it, it'll send me a text message. It'll be like, hey, look, you got to jump in. You need to fix some things. So you can put controls in place where they kind of manage themselves. Like my podcast agent is just scanning every day to see if there's an upload. It's on a cron. It's on a schedule. My land due diligence agent, that's not. I have to actually send it an address and a parcel number and say, hey, I got this lead that came in. Run the analysis for us on this deal. We just put a lot in Brentwood under contract. We're going to build a $4 million home. And I needed to come up with analysis very quickly because it's speed to lead. Whoever can do the analysis and go see it first gets the deal.

Brandon: And it was able to produce a full-blown report. Literally in about 15 minutes. Now I had to call on it to do that. And I had to train this agent just like I have to train an employee. So I think what a lot of people don't realize is they might build an agent and tell it to go do something. And then they get frustrated if they don't get the result. These things are like employees. You have to train them. They're going to make mistakes. I have an email triage agent that triages all my emails and the correct inboxes. This was the one I was like, hey, Seth, we got to build you this one. And it saves me an enormous amount of time because I'm not distracted like I used to be sorting through emails, wondering when something important comes in. I just set a reminder on my email, hey, if you see an email for my son's medical records or an email from Sam, my attorney come in, text me. I need to know that that email came in and I'm not scanning anymore. And I'm way more productive and less distracted because of that. That email agent runs on a schedule.

Brandon: It's running every 15 minutes only from 8 a.m. To 6 p.m. That's another big thing that I see people, mistake people make is when they build them initially, they're running 24-7 and they're eating tokens. And I don't need my email agent running from 24-7. And so when I realized that, I was getting some expensive bills and I was like, oh, crap, they're running 24-7. Not good. I need to put them in the little worker eight to six box like everybody else. So there's a lot of little things. It's a whole different world when you move into bucket number three compared to the bucket number two.

Seth: So just make sure I'm following you. So earlier you were saying a skill that runs on a schedule, like a cron job. So when you say agent, is that what you're talking about? Or you're saying you're doing something different than that?

Brandon: The real word for agent is there has to be thinking and autonomous decision making. Okay. Most people are building a skill that runs on a cron. A skill is like, let me teach you how to go into my Zoom and grab this. Okay. That's a skill. And then let me teach you how to go and read all of my podcast titles and descriptions and recreate a new one based on the transcript from a podcast. That's a skill. And so you've got a group of skills that are working on a cron schedule. That's what most of these agents are. And because it's doing a task. So people People think, oh, It's completing a task for me. That must mean it's an agent. It's wrong. And so I'll even use the word incorrectly just because I want to fit in with everybody and what they're calling it. And I'm not going to be the guy being like, you're wrong. It's not. It's a task on a cron schedule. No, but I've built a lot of actual real agents as well. And those are the ones that require actual autonomous decision making and stuff like that.

Seth: So just for example, this email triage agent, what does that do?

Brandon: So it scans the email every five minutes. And initially I had to train it just like an employee. Inbox number one is, Brandon, you need to respond to this. Inbox number two is, you need to know about it. No response needed. Inbox number three is, hey, here's your bills that came in. Inbox number four is, hey, this is all your podcast stuff because I do a lot of podcasts like you. Email number, inbox number five is spam. And then inbox number six is my sales messages. So I have stuff linked to sales message. And whenever I need to take over a DM conversation, it notifies me and texts me. And it just sits there. So I stare at a clean inbox every single day. It's completely empty. And when I show people this, the first question I asked them was, hey, why does your inbox look like this? And they stare at this blank screen with no emails. They're like, never. That's because the agent, that's because the, it's not an agent. Because this is, there's no decision making. It's in there, and it's actually just triaging the emails into the correct inbox. So I'm not distracted.

Brandon: And if there's something important, I can just tell, hey, text me whenever an email comes in from these people. I need to know about it.

Seth: So this is not responding to emails on your behalf. It's just sorting and prioritizing them.

Brandon: Correct.

Seth: Okay. So the problem you were trying to solve was basically, and I don't want to put words in your mouth, correct me if I'm wrong, but I can see my full inbox every day. A lot of the emails are not important at all. Like they don't even need to be looked at or responded to, but some of them are super important. And you don't want to just like, miss the, that stuff through all the weeds. So what this does is it just helps you zero in on the stuff that really matters and the stuff that doesn't matter. It'll kind of push it to the side and you can ignore it for however long you want to. Right. But, but you still do have to look at it. Right. It still requires your involvement to like discern, is this really important? And then how do I respond to it? So it doesn't like handle email per se. It just organizes the mess.

Brandon: That's exactly right. Like I got an email today and I got the text, the lenders requiring more information. I wasn't going to check my email today. I got a lot of one-on-one meetings, but now I'm like, okay, in between this time period, I need to go. I need to send them that information. If that had fallen through the cracks and they followed up with me five days later. I'd be like, oh, crap, this is important. Insurance. I'm fighting insurance on a medical bill. Who'd have thunk? And so I've been waiting on some medical records to come back. And if you submit for medical records, dear God, it takes freaking forever. And again, they send it to you. I'm not going to check it every single day. Because I've got five different inboxes. But I had my little agent set up to notify me, hey, Brandon, the medical records came in, and I'm like, oh, perfect. So I grabbed them, and I could immediately send them off. It prevents the things that are important from falling through the cracks in email. That's the biggest. And then the second is something that I didn't know.

Brandon: I'm so much more productive because I don't have my inbox sitting there distracting me, looking for the important things that come through. I'm waiting on the lender to send something. Now I can just create that little cron schedule, have the agent go check. I don't have to think about it because I know my agent's checking it and sending me the text when that important email comes in. And the productivity that I've gotten from that alone, you don't realize how distracted you are with your email. At least me. I don't know. It could just be me and no one else feel this way. No, it's not you. But I was extremely distracted by my inbox every day.

Seth: So on that thing, email, it's very important. Like there are some people that will just email you assuming that they're going to get through to you. And if you don't reply, they'll take it as an offense or they'll take it as like, he doesn't care about me. Or you'll just miss a great opportunity if you miss these super important emails. It makes me wonder, how perfect of a job does your email triage agent do? Can you really trust it implicitly or does it ever happen where it's like, oh, that was a really important email that it said wasn't important? Has that ever happened since you've started doing this?

Brandon: It's going to happen in the first 60 to 90 days, 100%. And that's why you have to train it. Whenever your emails are coming in and you've said, hey, like there's no way you're going to sit there when you build this email triage agent and be like, OK, I know every single email and everywhere it goes and I'm going to explain it to you. There's no way it's going to do its best job guessing because you don't even know. And as things go into the correct into the incorrect inbox, you're going to just highlight it and you're going to relabel it. And when you relabel it, that tells it the correct inbox and it's learning. So that's teaching it. And so it very, very rarely makes mistakes. Like there's never any mistakes in my podcast folder. There's never any mistakes in my, my bills, uh, inbox, my sales message, DMS. There's never any mistakes in there. Only two inboxes where. Spam, same thing. The only two inboxes that might kind of flip flop is the FYI, you need to know about it, but not respond. And then the, hey, Brandon, this needs your response.

Brandon: Sometimes there might be some things in the, hey, Brandon, you need to respond to this. That's really just a FYI, you only need to know about it. But those are the main two inboxes that I really pay attention to or the, this needs your attention. And this is an FYI, you just need to know about it.

Seth: I'm sure we've both done this, gone down these rabbit holes where you're just spending hours building something. And you finally reach a point where you realize I either can't do what I'm trying to do or I can do what I'm trying to do, but it's not actually that helpful. It makes me wonder, like, how do you know when one of these agents is genuinely reducing work versus just reorganizing work? Or is there a way you can, like, keep tabs on this? Like, I spent X number of hours on this a month ago, and today I spend half that much time because of this agent. Like, how can you really know that you're making progress here?

Brandon: This is a great question. So the biggest issue that I see when people learn this, like, we've got a community and we teach this stuff. And every single time, two things happen. And number one is, I tell people, hey, it's going to be like swallowing the matrix pill. Like you're going to go in a different world. You're going to get obsessed. Text me at one in the morning when you're building things and you're up and it feels like the day you first started your business and you're excited again. Like it's causing me to re-fall in love with my business all over again, which is really, really cool and get excited about things that used to not get me excited. And that's the first thing that happens. The second is they start building things that are a waste of time. They think, oh, man, the biggest mistake I see people make. They, they're like, dude, I'm paying 300 bucks a month for QuickBooks. You can go just recreate QuickBooks. I'm paying 300 bucks a month for GoHighLevel. I'm just going to go recreate and build GoHighLevel. And they're like, they think, oh, software automations.

Brandon: Don't try to rebuild enterprise software. You can do it. And here's the key. Enterprise software like huge conglomerate because when you have software these big companies have a whole team of developers that help fix the bugs and things go down and keep things running you don't have that. But here's what you can do that's really cool. The enterprise level stuff that works really well for you, GoHighLevel, ActiveCampaign, that kind of stuff, just keep it. It's the Toyota that doesn't break. It's reliable. Just have your AI plug into it and use it. Don't try to go rebuild a bunch of software. But what you can do is you could build localized software. Like I built my own, Nashville, Tennessee land, you call it like land that's ripe for development portal or something. And because I have a very specific strategy, there was no software that existed that did what I want to do. Actually, there was two, but they're like $30,000 a year.

Brandon: And I basically recreated it in my market, where it went and scraped all the public data, the slope and the acreage and the front soil, all that kind of stuff, and put it on a map where I could actually see the parcels that I wanted to go after, and skip trace their phone numbers, all that kind of stuff. I could build that because it's very localized. It's not the nationwide data. The software is very, very simple and it works. You can build stuff like that. The other day, three weeks ago, one of my project managers came to me and he's like, okay, we got to do a change order. I'm like, okay, all right, send it to the bookkeeper. Here's how change orders used to work. And it made me want to pull my hair out and we built an app to fix it. So the old way was, okay, we're on job site. We need a change order. Okay. Send a message to the bookkeeper. Bookkeeper has to wait to get the message. Bookkeeper has to create the invoice and change order has to send to the customer has to sign the change order. And then bookkeeper has to notify PM that change order has been signed.

Brandon: Bookkeeper has to send invoice to accounts payable to be paid. And you just, stuff gets dropped all the time. And you get to the end of the project, and you're like, why is this change order not signed? Oh, I never got it. Blah, blah, blah. I built an app in 20 minutes that literally you pull it up on the phone. It's trained on all of our customer data. You pull up the customer, click the customer. You put the change order amount. You put the description. You hit submit. Pulls up the invoice so that you can read it. And if it's approved, the PM hits the approval button. It automatically texts the project manager on site to sign the change order right then and there. He signs it. It sends that change order to the client's accounts payable department. It goes into QuickBooks. It creates the bill all flawlessly. Saves an enormous amount of back and forth and headache and time. And in the app, the PM can see all the submitted change orders that haven't been signed. So he can call or manage that right then and there. Oh, my gosh. It's a game changer.

Brandon: And again, I could build something like that because that is a super easy piece of software to build. That's, my team and my team alone is only using it. So I don't have to have a huge dev team to do stuff like that. So that's the coolest thing. And that's why I push people. The one takeaway I want people to know from this is for the guys out there listening that want to hire somebody to do this, like so many people try to hire me to build stuff out. Like I don't want to, I'm not that guy. And you're missing the point if you are trying to hire people out, because this is the new skillset. You've got to learn to stay in business, how to understand not just how to use Claude or ChatGPT. You have to understand the entire ecosystem, how to build agents, how to architect agents, how to manage things, how to architect the systems to work together. And if you're not in your business every day, playing around with it, doing it, I would have never discovered that change order app, gotten that idea to build it.

Brandon: And this is why I have over 111, we'll call them agents, but a lot of them are tasks, skills that are running on a cron schedule. But this is why I have 111 of these automations and agents going is because I get these little ideas every day and I'm slowly just automating one piece of the business at a time.

Seth: Yeah. Would you say these things that you've built are better at saving you time or making you money? Because when I look at most of the stuff that I've built, it seems like they're all huge time savers and it allows me to do a lot more stuff that I otherwise would have had to hire people for. And I don't have to do that anymore. Like it'll just handle a lot of this stuff for me. I think they do make money, but it's like, uh, it's not a direct correlation. Like I make more money because it enables me to do more stuff that eventually leads to that. But I feel like the bigger leverage point is just time savings. But like, do you have any things where it's like, yeah, when I trigger this agent to do something, I make money from it. Like, is it that simple or is that just not how it works?

Brandon: It's funny. So if you are set up the right way, the time savings should translate into making more money. The one specific agent that I can point to that's the closest to what you're talking about is I have a sales coach agent. And this is one of the first ones that people build when they come in our community, if they have a sales team. And what that agent does is it listens to my sales calls because I do a lot of sales calls. It's trained on our sales process. And of course, it's got a one bajillion IQ. And it gives me a report on how I did during each part of the sales process. And it'll actually coach me and say, Brandon, you said this but you should have said this. You said this, but you should have said this. And the first time I saw one of these reports and it was like, you should have said this. I go what? Let me just get my phone out. Let me text that guy back. And I texted him exactly what I should have said according to AI. And he goes, "Dude. I'm in." $25,000 sale. And that's happened multiple times since then.

Brandon: That's the closest one that I can point to and say like AI is literally making us money. Now, every business knows that it is running some kind of funnel. So it doesn't matter if you're looking at land parcels like us to do development and sell national builders or you're flipping land or you're flipping houses, whatever it is. Because you know that one of the biggest bottlenecks to get lead when you get leads in is being able to qualify those leads, because the more leads you can go through, the bigger you can make your funnel without it getting bogged down by the qualification stage and then making an offer, the more money you're going to make. And it is nailing that you can call like our land due diligence report. It takes us about four to six hours to underwrite a piece of land. Right. And if you're in the multifamily world, it's 20 to 30 hours to underwrite a multifamily deal before AI. Now, we can analyze so many more deals, and that's leading to us doing more deals and making more money. I'll give you an example.

Brandon: The lot that we put under contract, we're building a $4 million home. The reason we got that lot is it was speed to lead. No one's going to put something under contract until they've qualified it and they know it's a deal. We were able to run that analysis so much faster because we had an agent that was built and trained to do it. Now, it takes me about two hours to really underwrite a new build. And, doing it properly is a whole nother story. You kind of cut corners in that two-hour period. You can't go walk product and really look at all the competition. The luxury world is a whole different space. And you've got to be very, very careful not to miss anything because you miss one thing, and that's $200,000 off your sales price. We got that deal because of AI. We were this close to not being able to grab it. And so that's something that I can point to and say, AI is literally allowing us to do more deals, which make us money.

Seth: How do you know when the answer is not build it now, but the correct answer is wait six more months and it'll be a lot easier. When I say that, I'm thinking of countless examples of this where like things I was trying to do in December of 2025 were either very difficult and cumbersome to do or just not possible. Like I knew what I wanted. It just couldn't be done yet. But come June of 2026, oh yeah, you can totally do that now. Or maybe like the technology was like, eh, it was okay. But now it's like, ooh, it's really good now, six to nine months later. And I realized like, you can't really know this until time has gone by and you've seen it. But how often do you see that where it's like, man, I was working way too hard at this. I should have just waited a little bit and it would have been easier.

Brandon: Well, hindsight's always 20-20. You just don't know. You can only play the cards that you're dealt today. You don't know what cards you're going to have tomorrow. That's kind of why I'm so obsessed with this is like, I want to stay on the cutting edge of it. Otherwise, like that's what was keeping me up at night in March, April of this year was if I don't figure this out, my competition is, and like, I'm going to be out of business. That's what was going through my head. So I'm going to do everything I can to stay on the absolute forefront of this stuff. I'm going to pay every consultant. I'm going to buy everything out there. I'm going to be playing with this stuff every single day, staying up, working with it, because history shows that those who embrace the new technology get rewarded. That's become evidently clear. The first ones that really understood what the Internet was and the implications went on to build multimillion dollar businesses and some of them, billion dollar businesses.

Brandon: The first ones that really, understood, the implications of aviation and what that was going to do with the world, again, went on to make a lot of money. So you've got to you've got to be very aware of what's happening with technology because technology creates entire new industries that didn't exist before. So all my doom and gloomers that say, oh, the job loss and yada, yada, I don't think that's going to happen. I think AI is going to open up a whole other world that didn't exist. We don't know what it looks like, just like the internet did. The internet created tens of trillions, if not hundreds of trillions of dollars of new market value that did not even exist before. And that's what technology does. And so you want to be on the end that embraces it, that uses it, and those that do are going to be rewarded. And I could talk about computers. I could talk about the iPhone. Every time you embrace the technology and you get ahead of everybody else, you are going to be able to profit or make your life a lot better.

Seth: Let's talk a little bit about your community. Who is the ideal person for the system you're teaching?

Brandon: Yeah, that's a great question. It's so it's not for everybody. What I teach is advanced AI and And the big thing I like to lead with here is I'm not a tech guy. Okay. I was in medical device sales before I got into real estate 10 years ago, and I'm not super techie. So if you think you got to be super techie to figure this out, my whole setup is like AI managing and walking through this super technical stuff and doing it. I've learned to kind of speak the language and sort of read the matrix because when you're around it and exposed to it, you can't help, but it rub off on you a little bit. So I'm not a technical guy. I'm not. So you don't have to be technical to be in this. It's not good for regular day consumers. It's good for business owners, especially if you're a solopreneur. Or you've got like a team of people and you really want them to go fast and you're already using AI and you've got 12 different applications that you're using and you're kind of feeling overwhelmed by what's the new shiny tool and am I using this the right way?

Brandon: If you're getting frustrated with your current setup and you're thinking to yourself, dude, I hear all the amazing things that AI is doing for people, but I'm just not feeling it at the level other people are in my business, this is for you. It's the most exciting thing that I've ever come out with. And I've had everybody go through the program and everybody's had a great experience so far. I've gotten a lot of really good feedback. And I get text messages from people that are like, dude, you saved me nine hours today on this thing. This is wild. So it's definitely just for business owners. I don't think people that work in a corporate world have a W-2. We don't have any of those people in the community. But if you are actively. Doing business, especially in real estate, because a lot of what I built is real estate specific, it's a great fit. And I came out with something really cool new, and we can kind of go over that too if you want.

Seth: Yeah, sure. What do you want to do?

Brandon: So where I think things are going is everybody that builds their own second AI brain, I created a repo for my community. Which is basically just an API key. So everybody can plug their brain into mine and you can literally just download every single thing that I'm doing. You can have your brain go, hey, here's all of Brandon's systems and operations and all the bugs that he's fixed and all the agents that he's built and all of the systems and how he's managing things. I think we need to take a couple of these. These are really good ideas and they can just download my repo every single week because as I'm doing things and building things and training things on my server is I'm training my AI brain. My AI brain is training theirs. And so what we're doing in my mastermind community is everybody is now setting up their own repo where they're building in their brain and everybody's repo is connected to each other's and it's just this self-learning system. So the old way was, oh, Brandon built a cool agent.

Brandon: All right, let me go share that in the community and kind of copy and paste it in the Skool community and people can go download it and get it themselves and plug it in their computer. The person was the bottleneck. Now, we can just have our AI brains constantly learning from each other and sharing information, sharing what's working in real time, and it can just produce a report once a week and be like, hey, I scanned everybody's repo in the community. Here's all the cool things that they figured out that we need to implement. And you want to implement it. And if the person says go, yes, I do. Boom, boom, boom. It automatically does. I think this is the forefront. This is the next way of doing AI. So I think you'll start seeing a lot of the big AI people coming out with their own repo and saying, hey, you can download everything in my repo, get it for yourself. And so that I think like it's going to give a whole new term to the word mastermind because it's going from the individual people brains that are making up the mastermind to the like the AI brain.

Brandon: So it's something I'm really, really excited about that I think is going to be a game changer.

Seth: How much time should somebody realistically expect to spend learning this and implementing it? And it sounds like that repo thing kind of allows you to go at warp speed when it comes to implementing it. But still, like, you have to understand what's going on. So, like, when you look back at how much time you spent setting all this up, I know you probably spend a lot more time than the average person because you were doing it from scratch. But, like, if somebody were to join your community today and be like, I want what Brandon has, how long is it going to take him to get there? Like, what should the expectations be?

Brandon: I tell people, if you can go get rid of the wife and kids or the husband and kids for a weekend, give me two 14-hour days, just burn the bridges. I can get you probably the most productive weekend you've ever had. I feel that confident. I remember when I went through it. So that's what I tell people. For most people, they're going to spend a couple hours a day kind of going through it and getting it set up. And that's why we've got the coursework in there, because you do need to understand the back end and the infrastructure and how all this stuff works. Like you need to be able to speak the language a little bit. You don't have to know programming or writing code or anything crazy, but it does help to be able to know kind of what's going on. And that's why you've got to build it in a specific order, too. You've got to build the infrastructure first. You've got to get your server set up. You've got to make sure that your server is set up so it doesn't get hacked. You've got to make sure that it's set up or you've got a backup that's set up.

Brandon: You've got to make sure you've got the right gigs and CPU, RAM, and all that fun stuff. And then you've got to train it. You've got to build the infrastructure. It needs to know about your business first before you start building these agents so that the agents are high-performing agents when they come out of the box because they're trained on the context of your business. So there's an order that you've got to do. And let me be honest with you, it takes a few hours to get all of your apps connected to it. Like it is connected to absolutely everything. Go pull up all the software applications you're paying for. You're going to have to connect it to all those. But once you're there, once you're set up, I'm rolling. Like I don't spend a whole lot of time messing around with reconnecting things or anything like that. So I'd say 30 hours. Like you give me 30 hours over whatever period of time, 90 days, whatever. Like we will do some cool stuff.

Seth: Tell me about like the costs. And when I say the costs, I mean like, what are the different subscriptions or AI tools that you're paying for that you're assuming other people will be paying for when they start trying to do what you're doing? So for example, I look at my own setup, like I'm paying 200 bucks a month for Claude Code. I've got my server, 50 bucks a month for that. I got ChatGPT. I'm only doing the $20 a month plan for that because that's not really the main thing I use anymore. But like, what's the assumption of like, you're spending this much on these tools in order for any of this stuff to work?

Brandon: So I've got a $200 per month Claude Max subscription. I've got an OpenRouter subscription. OpenRouter is where my server can go and use all the 80 plus models around the world, Chinese models, Japanese models, all that kind of stuff. So that's why I don't use one model. So I got an OpenRouter subscription. I think my API tokens are like 60 bucks a month on OpenRouter, maybe on a high month, most of the time, 30 bucks. I've got a DigitalOcean server that runs me $55 per month for that. And I, I might have some API usage, maybe 80 bucks a month, somewhere around there. Just if I want to hit OpenRouter and use some different models to build things. And then I've got a RealVNC and Tailscale subscription. I think it's like, I don't know, maybe a hundred bucks a year. And that's mainly for my Mac mini. And the only reason I have a Mac mini is there's certain things that my server can't do. So you've got a headless browser and a regular browser. Headless is what the server is doing. You can't see it. And then a real browser is like you clicking on your computer.

Brandon: Certain things the server can't do, like manage my social media. I'll get banned if my server signs in. But my Mac mini, it can sign in fine. It can drive the actual browser. So I need connections there to do that. I paid $600 for a Mac mini. But again, most people don't have to have the Mac mini to do what I'm doing. All in all, probably spend $500 or less on my entire setup.

Seth: Yeah, the cost is real, but it also pales in comparison to paying even a single human to do this kind of stuff, let alone probably multiple people you could replace.

Brandon: I was paying $129 a month to have somebody manage each of my websites. I had three websites, $129 for each of them. AI is now managing everything. All of them saved me $390 a month, 129 times three, like little things like that. And I just, I just automated that and it's been running perfectly. So, the cost savings are absolutely insane. But again, don't go, I'm still using QuickBooks. I still use ActiveCampaign, GoHighLevel, like don't try to recreate those.

Seth: I've got a lot of different friends. I would consider each of these friends I'm referring to in my head to be like probably advanced amateurs, not like the creme de la creme of experts, but like, we're all pretty good at this stuff, but like, we're good in different ways. Like I'm further down the road in one direction of AI, whereas they may be further down the road in a different direction. So like, we're smarter in different areas. And for somebody like that, who already has some of their own infrastructure built out and just the way that they do things, like at what point is somebody kind of far enough down their own road that switching to your system doesn't make sense? Like, how could they self-evaluate and be like, yeah, that's great, Brandon, but it's not really for me. I've already got my own stuff figured out.

Brandon: So I'm assuming that they're either on Claude or ChatGPT, because those are really the only platforms that people are using to actually build off of. You could just have your server connect into your chat or Claude account by an API key. And download the entire setup into your server. Copy and paste it. Because my interface that I use in my server is the Claude interface. Like one of the tools that's connected in my server is Claude code. That is one tool that it uses to build and do things. It uses a lot of different tools. That's one of them. So the same interface that someone in Claude is using is the same interface that I use in my server to do things. So it's not super complicated. You just kind of got to get around some of the navigation when you first set up the server and kind of how it works. It can be intimidating, looking at a terminal and you're like, oh, my God, what is this? This command center. Am I going to launch a missile or type something that's going to screw my laptop up? What in the world? It's interesting.

Brandon: Again, it's like kind of going into the matrix, but it was not a whole lot. It wasn't difficult for me to transition over because there's not a huge, huge transition.

Seth: Yeah, I got you.

Brandon: But I know you, you had, you had like, I think you had stayed on like the Claude method because you really liked your Claude setup.

Seth: Yeah, well, for me, it's not like that I'm in love with Claude per se, but just looking at the Claude desktop app felt way more familiar and simple than looking at the same thing in VS Code. Just kind of a different look to it and seeing the stuff on the sidebar, I'm just like, why am I even looking at what is all this stuff? It just felt simpler on the desktop app version of it. But I actually realized you can tap into your server from the desktop app. So you don't necessarily have to open up VS Code to do it. But your whole OpenRouter thing where you can use any LLM that's out there, that's definitely something you can't really do from the Claude desktop app. So you'd be giving up that piece of it.

Brandon: Yeah, if you can wire in the Claude into your server, you can do that. So you keep your interface. But yeah, you lose some of the flexibility and the functionality.

Seth: Where do you think this is all going? In the next five years. I realize it's impossible to predict, but like it's just changing so fast or even three years, even one year. Like, what do you think is going to be the future of all this stuff?

Brandon: We're living through the singularity. This is it. If you look at what's happening, AI is just one technology. So for those that want to go deeper down the rabbit hole and all the stuff I'm about to explain, go check out Peter Diamandis' Moonshots podcast. That guy's got the, he's got his thumb on the pulse of everything that's going on. AI combined with blockchain technology. Let's talk about blockchain because, crypto, crypto is one way of looking at it, but the blockchain is the new financial system. These agents are. Are going to be transacting very soon. And they're not going to be using KYC or the SWIFT banking system. They're going to be using blockchain where they can settle instantaneously on something that doesn't get hacked. So you're going to a world where no more three-day ACH payments, no more 3% stripe fees. Wall Street getting their 3% cut is gone. That's going back into the pockets of the consumer and the business owner. We're going into an age where markets will be open 24 hours a day, seven days a week instead of the banking hours.

Brandon: Imagine how much more growth in business that's going to create alone just in that. Okay. So the next is the financial system getting upgraded with and combined with AI. Now let's look at the next frontier that AI hasn't conquered. We've got self-driving cars. My Tesla drove me to Atlanta. I didn't touch the steering wheel once. We're about to have self-driving cabs everywhere. They're already there. Waymo, Tesla's launching theirs. We're not going to be driving. We're going to be in a world where we'll one day, 10 years from now, look back and go, wow, can you believe people died from car accidents? Oh my gosh. And my kids won't ever have to learn how to drive, which is crazy enough to say. So that's coming. And then we've got the Optimus robots. When you take this body's brains that we've been talking about and you put it in a robot, what you just did was create a new species. And when the robots start building the robots, I don't know what that world looks It's like, we'll have a billion robots probably in the next 10 years.

Brandon: And then over the next five years after that, more robots than humans. What is that going to do? To the economy? If you think about all the change that we see, the change that we see is GDP, gross domestic product. And GDP has been growing at about 2% to 3% per year over the last 10 years. It was 4% or 5% before 2008. Then we had the debt crisis and yada, yada. So over a 10-year period, over the last 10 years, we probably have somewhere around 30% GDP growth. So imagine all the change that's happening in your life over the last 10 years. That's about 30% GDP growth, give or take. Because we have an infinite number of agents, we have an infinite number of potential workforce. And the formula for GDP is the number of people working times the hours worked, how productive they are. That's a simple formula to produce the change that you see, GDP. When you introduce an infinite amount of the workforce, these agents and these robots, what does that do to GDP? And that's not just the workforce.

Brandon: What happens when the hours work, 24 hours a day, seven days a week, and their efficiency 10x is because they're robots and not humans? I think that we're going to see in the next 10 years at some time. You will see 30% GDP growth in a year. So you have to imagine in a world where in one year, the equivalent change is what used to take 10 years to achieve. And every year, it's going to speed up and it's going to go faster and faster and faster. That is the world that we're going to. It is not an if that's going to happen. It is a when. And so when you combine that technology together and there's some cool stuff like breakthroughs and fusion and energy happening, all these technologies are merging. Again, it's not just AI. You got AI, you have energy, you have robotics, you have blockchain. They're all merging together that I think is going to create, the singularity. And I think that we're already living through it. So it's exciting. You're right. I can't see the future, but I am very, very excited about it.

Seth: Man. Like you said, predicting the future, it's like, why even try to do that? But as I hear you saying, it's going to go faster and faster and faster. I just started thinking of like basic physics. Like, where does this stop? Where does this plateau? Where does this blow up? Like, does it blow up? Like it can't keep going faster and faster forever. So like maybe it's five years, maybe it's 20 years, maybe it's a hundred. I don't know. It does make me wonder like, where is this headed? But at the end of the day, I can't control it. What's going to happen is what's going to happen. Anything I do is not going to like slow it down. So you might as well figure out how to live in harmony with it and make the most of it.

Brandon: Or figure out how to use it so that you're prepared for whatever is coming. Because Seth, I'm going to be honest, I really don't know if I'm going to have a business 10 years from now. I think that AI is going to create a world where we have to reevaluate and ask ourselves, what the heck does it mean to be human? And I think we're going to go back to our roots because I don't think people are going to have to do the same type of work that they did before. I think community is going to become very important. I bet sports teams probably do really well because that's a sense of community. Nature is going to be extremely important as it takes me several years to develop a piece of land and sell the houses and all that. We're going to compress that down into nothing. Nothing. And by the way, it's going to be the biggest deflationary bomb we've ever seen because what happens when robots are harvesting the materials and robots are manufacturing it and robots are building the homes? What's the cost of that?

Brandon: So it's a world of abundance where I think we'll continue to consume things at the same pace, but just like intelligence at our fingertips, magic intelligence that's doing things like technology. They can't fill this need. The need was there before, but the more they give us, the more we use. I think that everybody will probably end up living in a super nice luxury house because it'll cost nothing to build and do by the robots. Food costs hardly nothing because we're farming it and automating it. This is the world we're going into. And get ready because, again, it's going to be crazy.

Seth: Yeah. Well, we just spent a few minutes talking about the future as a final closing question. Looking at the past, can you think of anything that you have changed your mind about regarding AI over the past year? Like, is there something you thought in the end of 2025 that you're like, oh, that's totally different now. I think something different.

Brandon: Oh, man, that's a great question. I'm trying to think of what I thought about, AI in 2025 at the end of it. I don't think anybody saw it moving this fast. A lot of what I just talked about changed my mind. I did not know. I thought we were way far off on a lot of this stuff. We're a lot closer than I thought. It's just crazy because, yeah, I was using ChatGPT a whole lot and just asking it things. And I guess not. Nothing comes to mind that I can really think of that's changed significantly other than the speed at which things are happening and changing. How amazing that I knew the technology was going to be made and it was going to change the world, but I, I were, I didn't think it would be this close. Like we're, we're knocking on the door.

Seth: I think a similar thing. It's not that my mind has changed. I just wasn't aware as much. The big 2026 thing is just this agentic ability for AI to not just talk with you, but like, just do stuff for you. Like that is just such a huge deal. It really multiplies every human's effectiveness if they know what they're doing and how to take advantage of it. It's not that hard. It's just being aware of what's even possible. I was just talking to somebody yesterday. And I think people like you and me, Brandon, maybe even more you than me, you're so advanced and so far into it. It's easy to forget that most people are like 1% as deep as they can go. Like they think chatting with ChatGPT means that, yep, I know what's going on. Like not really, man. Like this goes so deep in so many different directions. Like there's just so much to learn.

Seth: I don't know that any single human can really get to the bottom of it, but it kind of goes back to this point of like, given how much it's capable of, you better start figuring out something because there's a lot that can be done right now and it'll only get more so.

Brandon: Yeah. Just start, just start learning it. And if you're somebody that's struggling, to learn it, I've got a, I'm doing another webinar here and I think it's like the first week of November or something like that. So if you go to joinaiaccelerator.com forward slash webinar. So it's joinaiaccelerator.com forward slash webinar. You can sign up for it. I'll show you all the cool stuff that I'm doing. And if you like the idea of being a part of a mastermind community of other business owners that are doing this stuff and sharing what's working and staying in touch with all the frontier technology, you can go to the same website, joinaiaccelerator.com, and there's a video. And if you want to book a call with me and we'll explore your business and what you can build, then happy to do that too.

Seth: I'll be sure to include links to that and Brandon's other stuff in the show notes for this episode. If you want to check it out, it's retipster.com forward slash 294, because this is episode 294. Brandon, as always, great to talk with you. Thanks for hanging out with me for a bit. All the listeners out there, we will talk to you next time.