How to Build an AI "Second Brain" for Your Amazon Business

How to Build an AI "Second Brain" for Your Amazon Business Speaker 1: What's going on, Badger Nation? Welcome to the world's first and longest running show all about how to make your Amazon advertising life a little bit easier and a little bit more profitable. Today on the show, we have Dan Head from AMZ Elite, and we're going to be talking about AI and specifically how to build things like second brains, I set up MCPs, what a tech stack looks like, a sample first activity. This episode is going to be really good in case you haven't gone any deeper into AI other than using ChatGPT, the actual interface. It talks about how to customize your AI experience in whatever tool you're using. So this is a good primer for people. I want to do more episodes in this thread soon. I know that we internally are getting ready to release our own MCP server. So this is sort of slotted in. I'll probably talk a lot more about it as we Continue this path both by box checker is going to be rolling out with an MCP server as well as add badgers core tool also have an MCP server so that you can begin to have conversations with AI. It's a little bit more customized than the generic. So this whole episode is about customizing your AI to get more out of it. I hope this inspires you. And then in future episodes, I really want to do more building and demonstrations about how to use things like AI and MCP servers to get a lot out of your Amazon marketing experience. Have a good one. Let's jump in. What's going on Badger Nation? Yes, right on. Today on the show, we have my dear friend, Daniel Head, my favorite Florida man from AMZ Elite, a full service consultancy, helping people on Amazon. 20 years of experience with Amazon sellers. You've been on the show before. We talked about great weekly routines on how to Better Manage Amazon PPC. I've learned so much from you over the years. Every time you come to Austin, we get some steak, don't we? Speaker 2: I'll be there in October. Speaker 1: We get some steak. It's my turn to buy next time. Speaker 2: Okay, fair enough. Speaker 1: We get some steak and you know a lot about steak, don't you? You always are making very astute recommendations and you have a lot to say about the steak that I've seen you eat. Speaker 2: Oh, yeah. Yeah. You know, if you're going to travel, you know, I want to try to find the best steak in town. That's just kind of like my rule of thumb. If you're going to go somewhere new, let's find out what's the best. And, you know, I got spoiled when I first started in this business because my father used to be the vice president of hard lines at Target. And he made this good friend, Raymond, who created a company called Gus Busters. And back in the early 2000s, when YouTube wasn't like so big, Raymond made millions of dollars by taking his Chinese made umbrellas and sticking them in a wind tunnel and they wouldn't blow back. So that's how he built Gus Busters. And the first time I met him, he's like, I know your dad. I'm taking you out to dinner. I'm like, OK. And so I meet him at Charlie's Steakhouse and he comes busting in with like 15 people. And he's like, do you have a reservation? It's like, no. Let me call Charlie right now. So he gets on the phone, calls Charlie, hands it to him. He's like, OK, we're taking you back to the private room. We get back there, and I'm like, I'm a very, very picky eater. And I'm like, I'm like, oh, I'm gonna get the flag. And Raymond looks at me and he's like, you're not getting the fillet. I'm like, what? He's like, no, no, no, no, no. I'm going to get you the Japanese Kobe. I'm like, Raymond, I'm a very plain eater. He's like, trust me, you're going to get to eat this steak. And if you don't like it, I will eat it and buy you the steak that you wanted. I'm like, okay. He's like, trust me, this is a $250 steak. Okay. Then I got that steak and it was the best steak I had in my life. Yeah, it was. It really was. The wood-fired grill makes a big difference. But the one place that I found that is probably even better than Charlie's is Garden Grace and you have one in Houston, not too far from here, and there's another one in Denver. When I go to Houston to visit my sister-in-law, I take my wife there every time. I can't go to Houston or Denver and not go to Garden Grace. It's just not possible. Speaker 1: It's beautiful. So I guess you have your best steak that you've ever had in mind. I was going to say it would be really disappointing if you were constantly chasing a steak that you couldn't find, but you found it. That's nice to hear. Speaker 2: Yeah. Now, it's not to say there isn't one out there that beats it. And that's why I keep going. Speaker 1: It's great. You know, normally when people come visit me in Austin, I take them to like, we'll go to my favorite food truck. I'm like a food truck guy. Okay. Speaker 2: Well, hey. Food trucks are great though. And I've seen some great food trucks in Austin. Speaker 1: The best Italian food in Austin. We actually took you there. We didn't get the Italian food, but it's Patrizzi's Italian food here in Austin, Texas. The best Italian food comes out of a food truck. How about that? Speaker 2: Wow. Speaker 1: And I got a lot to say about Italian food. Dan, I love the topic that we're going to talk about today. And let me just talk to all the dear listeners out there on Out there in Badger Nation. A lot of times on this show, I will try to leave the listeners with really practical things to go and do. But before we talk practically about how to do what we're going to do, I also think it's just worth conceptually understanding this, which is basically like you called it a second brain. It's basically how to use AI And we're going to talk about how to absorb a lot of information, how to recall that information, and how to do it at scale with specific information for your company. So basically, the trend that I've seen is that basically when ChatGPT first came out, 2022, 2023, when it first came out, it was amazing because it's like, wow, I can talk to something that knows a lot about A lot of different things. However, since then, a very common quote that I love to repeat is that AI knows a lot about stuff that I don't know about. Meaning, I don't know about how to sail, but AI knows a heck of a lot more than I do. However, The best sailors in the world will probably laugh at the advice that comes out of AI and ChatGPT by default about sailing. So the savviest people I know, yourself included, I've been doing a lot of this. Other people I know are basically building internal AIs for their company, which recalls information that you want it to recall, prioritized, rather than just generic insight out there in the world. So there's a couple different actions here. Number one is absorbing all the information. And then number two is building an interface to recall the information. Now this has deep implications for any business, anywhere. I kid you not, I just had an HVAC person in my house and He was like installing, I just got a brand new air conditioner. My house is so cold. And I saw him using Gemini to ask questions about installing this like new kind of HVAC unit. And in my head I'm just like, I asked him, oh, what is that? He's like, it's Gemini. In my head, I'm just like, this dude is missing out on a huge thing by not building a database of all of the official manuals for all of his HVAC units that he installs. Imagine he had all of that in a bake-in. Speaker 2: PDF. Oh, man. Speaker 1: Right. So imagine he was getting like very specific, good advice on all the very specific things that he's doing. And the same is true for any business anywhere in the world of AI. And we can talk a little bit about today about how to do it for Amazon. But it's really like conceptually, if anyone has not set up their own private databases or own private MCP servers or all of this stuff, hopefully this episode is a nice primer for you. And then later episodes, we can get even deeper on How to turn this into reality might not be so relevant for this show. I might do it on my other podcast, which is a little bit more generalized. But anyway. Welcome back to the show, Dan. I know I just bursted through there. You've been pretty excited about building a second brain for a while. I want to say, months ago, you told me about this News Digest, which I think probably spurred it off for you. So let's just, in the same way that when you walk into a steak restaurant, you can smell the wood-burning stove. Did I get that right? You smell it. You see the sights and sound of the steakhouse and it tantalizes people. So tantalize us with This news digest that you have. Speaker 2: Well, I mean, I think everybody nowadays has the same problem. We all have the same problem. We have way too much information thrown at us and you can't possibly digest it all. And then a lot of times you'll watch some of these YouTube videos and there's 45 minutes of yakking and then there's the 15 minutes you need to know. And all this information from all these different points and all these different people see different perspectives. So they have very valid points. So, you know, what I'm excited about is being able to digest massive amounts of learning easily. Setting something up like an agent running, checking both your email boxes, watching your, your YouTube videos. Web Clipper and Upsetting is really cool. It pulls all the transcripts out of videos. So my agent will go out and check all these spots, all these learning spots and bring all this information in. And then it's going to boil it down to like two or three pages so that, you know, it tells me like the most important things. And then it looks out for things like how can this material, materially help AMZ Elite. So it's breaking down the information in pieces where I can actually understand it. And when you're getting emails this angle and you've got to be on this webinar and you've got to go here and you've got to watch this, it's just all too much. I mean, and you go to like one of these conferences and you watch 15 people speak. How many of those people are you actually getting back from that conference and actually actioning, you know? That was a little weird. Speaker 1: Actioning. Yes, you're absolutely right. People don't need to hear us talk about stakes. It's true, right? So I do think there's, I mean, this was one of the coolest use cases of AI in a while. Like I put all of my YouTube videos into something, you know, like it's amazing. So like the first thing of an amazing project, and again, maybe later episodes, we can actually like do some modeling on this. But basically it's like, you can these days, it's basically the transition of moving from like generalized AI Because if you just go to AI and you type it in, like what's new in Amazon PPC, you'll get some generalized stuff, but it doesn't compete from like, hey, scrape the Amazon official announcements every day and let me know if something changes. Scrape all these YouTube videos. Let me know what's being talked about. Scrape this all over the place. So it's like digesting that and consolidating that is something that is so useful. I've done it myself. Sounds like you've done it too. And that is an amazing use case. Again, it's this whole practice of like customizing the AI that you're using. News digest, pulse on the industry. That's an amazing first project. Now, if you think about this, you mentioned something really unique, which is like, well, how can this benefit your agency? That's the other element of it too, which is of course, pumping in your own You know, hyper specifying this so that it's trained on things that you feel are important for your agency. I can't tell you. For over a decade, I struggled with team members, especially in a remote environment. How do you get everybody on the same page? You can get people doing bid optimization 10 different ways across 10 different people. It's like, how do you get everybody on the same page? It's really tricky. Previously, the wisdom was like, go create an SOP. But then it's like, how do you make people use the SOP? And then it's like, okay, well, you know, what about the SOP goes out of date? How do you update it? What about this person learned something new or there's been a change? Like how do you go and update the SOP? Like SOP management in a way that's really easy for people to use has been such a game changer for us. Like getting some uniformity into a remote team, like getting some standards that are going to be applied in every different situation. That's been such an amazing use case. When we write an update for something, it gets written like this. When you're doing your post-call recap, it gets written like this. AI is incredible at this. So it's basically going from generic advice, like, give me a follow-up email template, versus these are the ways that we like to write email follow-ups, and then making that accessible through AI for the rest of the team. It's amazing. Speaker 2: Yes. Getting everybody on the same page, like you said. Speaker 1: For you, you mentioned something interesting about a daily triage report when it comes to Amazon. Tell us about that. Speaker 2: The daily triage report, I built what's called a command center. And the command center is set up to Amazon with the MCP and it'll hit inventory. It hits all of my PPC. The command center is more designed around PPC than it is about inventory, but you have to include that inventory factor in there. Like you were talking about so much yesterday in core community. It's like inventory does so much to your conversion rate. It's everything, you know, to a point. So having that in there and then like this morning coming in and saying, okay, This is a good point for the brain, first of all. Originally, like a few days ago, I had an item come out of stock. Because of the SOPs in there, it was saying, okay, you need to turn off these campaigns. And I'm like, No, this is not how we do this. We missed an SOP that did not get in here. But when an item goes out of stock, we turned a portfolio budget to a dollar so that when it does go back in, we just kind of take that off and we let it rot. So I went and I fixed that. And then my triage report came out this morning and it said, OK, you're out. These items were available, but you sold out of all available inventory. You need to turn these portfolios to a dollar. Then it alerted me that one random campaign suddenly spent 4X yesterday out of the blue for no reason. And then I had some other alerts with some good campaigns that were running out of budget that are performing better since my team did the adjustments to them. So now we need to take the budget caps off of those campaigns and actually let them feed because we actually turn the performance around inside that campaign. Speaker 1: You know, I think you brought up a couple of different great points, which is great AI use cases. Number one, it's Much easier today to combine like PPC data and sales data and inventory data or rating data or shipping time data or competitor data, like whatever data point you want now, you can shove into AI and it will ideally find trends there that you wouldn't have been able to pick up on previously. So like that's another amazing use case. So it's this sort of like information digesting that it can do a great job of too. And I would say like, you don't need a super duper fancy setup in order to just get there, right? Like you could theoretically have your SOP about how you want to handle something in a Google Doc. And then you can just download these reports in a CSV and then just give it all to Your favorite AI and do something with that, like be able to converse with your ranking data and your search query data and your PPC data. That's just three exports. And then if you want to get it to do the thing that you want it to do, that's just an SOP that you can give it to. So it's another place to begin. I've been talking a lot about that on this show, which is just like these two reports used to live so far away and now you can have them live together and stack the data. So like the inventory report with sales data, like how much you're missing out if you're not well stocked or how much you're losing if you're overstocked. So it's like, that's a great interface to begin. And then eventually you start seeing like, oh, wait a second, like, There's a way to not have to grab the CSV manually. I can pull this via API or even like browser control. Like you can spin up your browser just to once a week log in and then download it directly. You can do things like that. This sort of second brain, what has it done? Like tell me about How your day is different today versus what it was a couple years ago? Speaker 2: Okay, a couple years ago, I would spend a couple hours going account to account to account, checking, logging numbers or getting a staff member to do it. You know, looking at the talk lists, looking at all this. Now this stuff is all automatic. It saves me an incredible amount of time. I think you were the one who brought up the question. How much has AI made me? It hasn't made me money, but it's saving me a lot of time, and that's just as good as money to me. Speaker 1: Tell me about that. What is the practical impact of it? I know someone who really got into AI, and he's like, I'm firing 80% of my team. This is a small business, so he's getting rid of five of the seven people that he has. Then I have other situations where it's like, Imagine I get everybody so much more efficient. Imagine how much more we'll be able to get done. So that's one thing that's on my mind, training everybody on cloud code, being sure that we're really good with all this stuff. What's been your approach? Speaker 2: Yeah, I've been preaching that too, beefing up the AI talent inside my employees. I don't plan on letting anyone go. The people that I've had are almost family to me. My VA, we call it the AMZ Elite baby, and then my cousin bought a house. I don't plan on letting my staff go anywhere, but what is changing is their role. Their role used to be in the trenches and now they're starting to come out from the trenches a bit. Speaker 1: You know what I find a lot though? I find a lot like the business owner cares a lot about efficiency. And like productivity and effectiveness. Because I experienced this, like people come to me, they're like, hey, I want to sign up with AdVagr to like get my team more efficient. Like they're like managing their keywords bid. They're managing their keyword bids manually. Like we need to put them on an algorithm or we need to put them on a rule set. And then like you like look at the account like two weeks later and like the team members haven't logged in. There's like stuff like that because it's like incentivizing the team properly. Like, hey, there's a benefit in becoming more efficient. There's a benefit in becoming more effective. Do you have any pointers for people out there who like want their teams to become more efficient? Like what's a good mindset for them? Speaker 2: You need an EOS for that. You gotta drive that. It's all about the carrot and getting them to chase the carrot and then giving them the carrot. So you set goals, you set rocks, and then you have them go and achieve those. And if you achieve those, you give them your word, you do this, I'm giving you a $5 an hour raise, you do it. I just did that. One of my guys just got a $5 hour raise that's going into effect on Monday. It's about driving people. It's still management. What you need to watch out for is the people that are so resistant to change that they don't want to be out of the trenches. Now, I'm lucky. I don't have that kind of a Corporate atmosphere, you know, everybody in my company, we know each other pretty well and we all interact pretty well. So they all want more. They, they, they want to do more. And that's, that's the kind of people I have on my staff right now. So, you know, getting them more motivated to actually go in, follow these processes, use the second brain, this is things they, they already want to do. So when you kind of have that kind of corporate atmosphere, it feeds itself. Speaker 1: You know, these are things that I've always loved. You know, it's true. Like it used to be like there might be a report preparer at an agency, someone who just logs in and grabs- I had one. Right. And grabs the report. And I think it's like now that person can do other things as well. Like, so it's kind of cool to watch, like, hey, like we have these people on the team. Like, how do we get even more output from everything? Like, because that's just the pace of things. So I'm curious on your tech stack, can you share anything like what what are your go to tools at the moment right now to sort of orchestrate all this? You mentioned Obsidian, which is a thread worth pulling on. Tell me, for those that don't know, Uh, what is Obsidian and how does it get incorporated into your process? Speaker 2: Uh, Obsidian is, um, you know, you can call it the brain. Some people call it the brain. Some people don't. Some people call it the vault. Uh, it's essentially a vault with all these markdown files that has all these rules that, you know, you always follow. So, so for instance, uh, launching a new listing, you know, you knew how tedious that used to be. I mean, like, Okay, you're gonna launch new listing. You've got to go look at the top five competitors. You got to go analyze what their weaknesses are, what customer review insight says, what this and that and this and that. Now it's a simple process. It's, it's okay. I check customer review insights first. I check the top five competitors. I copy all their A+, copy all their info graphics, feed all that in there. And then I break down every single level of their reviews. Five star, four star, three star, one star, all the way down. And all that information is always ready to go. And so by the time I'm 30 minutes in, I already have customer review insights, a full breakdown of all the competitors that are, what they're winning on, what they're not winning on. And then I can turn around and build a listing on another 30 minutes. Speaker 1: And for those that don't know, like also Obsidian, the history of it, it started off as like a note-taking app, like a competitor to like Evernote. We just had this simple thesis that what if we just had simple markdown files so that people could manage it on their computer, like they could own their notes. And then in the world of AI, AI really does love markdown files, like these .md files, because they're lightweight and easy to peruse. Then you can just take all that information And then just like give it to AI to refer back to periodically. And there's lots of different ways to do this. At our company, we put in Google Docs and then any document in a particular folder will get scanned periodically to check for updates. And then we like vectorized it into a vectorized database that way. And then of course you hook up your MCP so you can converse with all that data inside of it. So we then put that MCP inside Claude. We put it inside Claude Code Projects. We put it on a private URL. We put it in Google Chat. We put it in ClickUp. So you can tag the company knowledge anywhere you're at to get insight on like, how do we do this thing? Which has been a big stumbling block for me over the years of like, how do you get people to use SOPs internally? And it's just like, make it really easily accessible, make it queryable in plain English. So like, instead of having to search like product launch checklist, or like what if somebody called it a product listing checklist, they would never find it historically. And now like AI will query it a lot more effectively to give them the process. And then of course, the next evolution of that is, imagine you took that process And you gave it to AI as well, which is like, let's do some competitive analysis. So we have the manual SOP, that information living somewhere. But then imagine your AI has the tools and the keys to go and do the thing that they need to do. So like, eventually you can start like building things into your setup. So you can add like a web scraper that scrapes The Amazon blog, right? I love that. Speaker 2: Thanks. Speaker 1: Right. So you can add a web scraper, which will go out and try to like grab information from a product listing and putting it, put it into the SOP for you. So it's just like finding levels to say like, oh, I'm going over there and clicking on it. Can I have AI like click on it and go and get it for me? And then again, you're like getting deeper in this like realm of just like it processes information. It can give me specific information and it can go and grab the information I need as well. And there's various ways to do it, whether it's giving it a spreadsheet or giving it an API connection or just making it scan through your browser. Speaker 2: It's cool. Speaker 1: But I mean, directionally, I think there's a huge benefit in any company doing that, especially a digital marketing agency, especially somebody on Amazon, because you can configure a weekly task that goes and It scans what your competitors are doing, who launched the product. So imagine you just had that humming in the background for you, or like a competitor changes the price or runs a promo or does whatever. Imagine you just have something pinging you just to alert you. Speaker 2: So like that's, that's been something I've been going after too. Those darn external factors that are so hard to figure out. Like, you know, ah, I got a, suddenly I got a 10% drop in conversion rate this week. What's going on? Usually that's your competition, but It's hard to diagnose it with just a human brain, you know? Speaker 1: Right. So it's like, what did search query volume do? What did competitors do? What did this do? So it's like, you could now like query a lot of these things all in one swoop. It's a new world. It's a new world. Speaker 2: It really is. It really is. Speaker 1: So yes. So just this mindset. So the thing that I hope people walk away with from this episode is really just like all the systems that you ever would want to deploy at your company. Like you could build up a Database of all those things so that you and the rest of the team. You can always refer back to the latest and the greatest updated things. You can even set up reminders like, hey, this SOP hasn't been updated in six months. Go assign somebody to update this SOP, right? So like that was another thing I suffered with too of just like getting documents fresh and updated. AI has sort of solved that because it could just scan, see what's out of date and do it. Speaker 2: Well, that learning thing, that's the first reason why I developed that so that I could make the second brain smarter. So it's like, okay, so these two or three paragraphs You know, this is really game changing. Like for instance, when Amazon announced we're going to do a 75-125 on the titles. You know, I had to sit there and change everything right there on the fly, but the information was already coming in and it was already there. And so being able to adjust that on the fly, all your processes are updated now. Then you go teach your team members and now they're updated in the weekly meeting. Yeah. Speaker 1: Yeah, man. I just had a brand come to me and they were frustrated with their agency because the brand had to bring them to their agency. The character limit change on the titles. The brand was like, why are we finding out about this? Shouldn't the agency know about this? It's like monitoring the chatter, monitoring the news and then shifting your SOPs. We live in such a fast-paced world now. You have to have some element of what's new for sure. Speaker 2: It's been like this too for years, but now we can actually do it. Speaker 1: All right, CounterPoint. Do you know the app Blinkist? It's like a book summarization company. And I've had a subscription for years and years and years. So if I wanted to listen to an audio book or read a synopsis, it gives you the cliff notes on a lot of books. So it's so funny. I remember five years ago, I was listening to some business guru, and they're like, I read five books a day. I'm like, Bullcrap. You don't read five books a day. And then on another episode of his podcast, he said, well, I use Blinkist, which is like a Blinkist version of a book is like 20 minutes. So in my head, I'm just like, you can't call that reading a whole book. You can't tell people that you read five books a day. It's like you read five book summaries. The thing that I find interesting about that is it's true. When I listen to a Blinkist and I do it a lot, like there's a business book that I feel like I just want the synopsis of, I'll do it a lot. Blink, two or three books a week. It's cool. I've been a subscriber for years. What is interesting about it is like, Sometimes I will listen to a synopsis of a book and read the book itself too. In fact, what I often like to do is do both. I'll start with a summary. So in 20 minutes, I have the gist of what the book is about. And then I read the actual book and I'm like, wow, I can really see everything here. It's almost like you could see the full matrix. Because when you read a book, you're spending eight hours with an expert. I'm wondering how to balance that. There's a great book that I wanted to listen to, Software as a Science, and I was listening to the audio book when I'm running, but then I'm like, ah, it'd be cool to be able to query this book with AI. Now, if I just ask AI for a summary of the book, it'll give me one, but it wasn't good enough because it didn't really know about the book. So I transcribed the entire book in text, and then I gave that text To AI. And I was like, let's dig into chapter two. And I was able to like interact with chapter two, and then like it recommended dashboard, then I took the information from it, and I gave it to Claude Coe to build a dashboard that I was talking about. It was amazing. Speaker 2: Sounds neat. Speaker 1: It was so cool. So like transcribing audio books is like, a new thing that I'm doing a lot of because it's like I want all the knowledge of this book. And I want to be able to like go deep into a thing, not necessarily just skim the surface. Anyway, I'm sort of giving the counterpoint, which is like, are we going to miss stuff? By only getting like the summary of stuff, rather than the nuts and bolts and like, does that even matter? Like, should we be concerned about it? I'm not sure. Speaker 2: Yeah, I'm sure there's information that falls through the gaps. There's no way to get it all. There's just too much out there. But I think because of the second brain is it, it kind of understands me, it seems. Because I give it context when I'm teaching. So I'm literally teaching the second brain. Speaker 1: Don't make this mistake again. When I get this inventory report, I want it formatted like this. Always check for that. Absolutely. Speaker 2: So I think it kind of knows what I'm looking for in general, but if there's a gap in what I'm looking for and the information coming out, that's always a real problem. That was like one of my favorite questions to ask AI and like back a little bit was what is the gap between my initial question and your answer? And then how could I have worded this properly to get a better result? Things like that. Speaker 1: You know, I don't know if this happens with you, but sometimes I struggle with, I mean, I think anyone's AI, the AI is trying to be correct. And not only is it trying to be correct, but it's trying to convince you that it's correct. Bias. What's that? Speaker 2: That's bias. That's the bias. That's probably the single most helpful thing The helpful prompt I've ever learned in my entire life is drop the regency bias, drop the showmanship bias, and drop the confirmation bias. And when you drop a bias out of AI, it gives you real answers. It really, it goes deep. Because AI is, it wants to make you happy. And that's a lot of the reasons why it hallucinates because it wants to make you happy. So it wants to believe in something that isn't true because it thinks it'll make you happy. Speaker 1: You know, one of my favorite things is I have Claude Code analyze my computer. So like move files, like just like general storage stuff. Like, hey, do you see any files that I haven't opened in a year? Can I delete these? How's my RAM doing? Like what's a RAM hog? Like all this stuff. And my computer was feeling a little sluggish because I like push this thing to the limit. I got like a fully spec'd out. Mac Mini. I love this thing. But it was feeling somehow slower than my weaker laptop. And I couldn't understand why. And I'm asking Claude. And this Claude session knows all about my computer. It's got saved memory files. I've gone through a whole bunch of iterations of, this is my setup. This is my Wi-Fi. This is everything that goes on in it. And it was telling me to close Chrome. And I'm like, stop. It's not Chrome. There's something awry. And it's like, well, what about this? And I'm like, no, I don't think it's that either. And on the third time, I had to tell it to look for Processes like pull up the activity monitor, which like lists all the individual microprocesses on a computer, and then it found something. It's like, oh, you had this stray Google Drive script that was writing and rewriting like 200 gigabytes nonstop like. It would write them, realize the database was not where it needed to be, and then it would delete it. And then it would try to write it again. And it's been doing that for six months. And I'm just like, I had to nudge you like 10 times to eventually look a little harder. And then I had to tell you where to look. So I feel like But then again, at the end of it, it's like, remember this. Don't forget this. Don't make this mistake again. Whenever we struggle with speed, go back and correct itself. I actually noticed too many people not telling their second brain or whatever that exact thing. When it messes up, log this to memory. Don't make this mistake again. What do you have to do in order to remember this next time? And it should develop skills differently or develop its core Markdown files better to mitigate this from happening again. Speaker 2: Well, I think when I understood what instructions were, when they started adding instructions to projects, I'm like, oh, this is great. Unknown Speaker: Let's fix this. Speaker 1: Let's fix this. Speaker 2: And that's all the brain is. It's like all the instructions. They get updated and they get refreshed and it won't fall apart. Speaker 1: That's right. So like the way that I'm thinking of this now, like I have like a choked RAM, basically like a choked RAM SOP. Like when RAM is degrading, go and do all these things. And then I like push it, like go scrape the internet to look for all the best people who talk about RAM usage on a Mac. Like let's become an expert In this topic and take our experiential learning and bake it all in there. And like just trying to go from like AI knows a lot about stuff except specifics to now you know a lot about stuff and you've learned specific information. Speaker 2: Yep. Context. Yeah. Yeah. That helps a lot. Speaker 1: Alrighty, Dan. Well, I think we've covered a lot. Hopefully we've inspired people to go and play more with AI. You know, in our core mastermind, I remember I was doing like a Claude code session of like, We're auditing bulk files with Claude Code. And I was telling everyone, there were some people on that call who had never used Claude Code before, but you can do a lot yourself with that. And I would say the first thing to do would just be, you just show up and you start asking it. I would build the news digest. That's an amazing thing. Relatively well known. To just go out there, scrape the internet, look for stuff about Amazon PPC and give it back to me. That's a pretty great place to start. And then eventually you get more and more to helping you with actual optimization tasks, which is like, now I've got a CSV of like my rank data and my search query data and so on. Like, let's find some insight, dump it all in there. And then eventually you'll get to a point where you don't have to grab the spreadsheet and you can pull directly. It's cool. It's an exciting time. Speaker 2: Yeah, it's saving a lot of time and it's definitely making us much more efficient and we're learning. Speaker 1: Is it making more money? I'm finally starting to see it because my big turning point is that for a long time I was having AI invent a task for me, meaning like, oh, I'm not doing this, but AI could do it. So then let me go build and manage a task that I've never done before just because AI can do it. And I feel like that was a huge basis. I pulled it on a lot of... Speaker 2: That's a trap. Speaker 1: I gained a lot of experience doing a whole bunch of different random stuff, but now I'm much more strict on like, what's the business use case? Is somebody currently doing this and we can like amplify this or accelerate it? That's been a huge game changer for me. Don't use AI to go invent something. Use it to like take what you're doing currently and like put that in warp speed. Speaker 2: Better and easier, faster. Speaker 1: That's the move. Dan, thank you so much for coming back on the show. Thank you for having me. Florida man. Can't wait to- Steak eater. Steak eater. Can't wait to have you come back to Austin. It's always a pleasure to get together. I'll be out there in October. Heck yeah. Speaker 2: A week before Halloween. Speaker 1: Let's do it. Everyone else, thank you so much. I'll see you next week on the PPCNN Podcast. Unknown Speaker: I've got my bits, set placements too. Now that we say. I've made a few, I've had my share Hi. We are the PPC Den, my friends. You are the PPC Den. No time for medicals, cause we fixed the game, baby.