How to Use AI to Make Money, Grow Your Business, and Be More Productive | Geoff Woods

There are three skills that somebody fundamentally has to learn. I don't care how good your prompt is. If you aim it at something that doesn't matter, it doesn't matter. You use this to solve your biggest problems or to help you achieve your most ambitious goals, that matters. Today I'm bringing on Jeff Woods, who's an expert at helping normal people use AI to change their lives. We're going to steal his homework and a four-step framework he uses with AI to completely change the way you work in almost every aspect of your life. What about people who say that AI kills creativity? The difference is people who are creatives have been creative with design, creative with writing. You're no longer the player of one instrument. You're now the conductor. What are some obvious things that you see that you go, that is AI and not in the good one? There are certain things that it opens a paragraph like. It seems like people have a lot of misconceptions about what AI can do, how it can actually help our lives and make them better. Can you give me like your most jaw dropping example of how AI has helped somebody in the last six months? Because it seems like it's changed so much. Yeah. The greatest use case I have found is not using AI as an assistant, but using it as a thought partner. Because your ability to think strategically is the difference between growing your business or going out of business. We had one company, venture backed, racing toward an exit, hopefully over the next 24 months. CEO calls me and says, we have a big problem. And I said, give it to me. And he goes, you know, despite we hit our goals every quarter, every year we exceed budget, no matter how good we do, we have a hostile board. Every quarter is an absolute bloodbath. I'd say 50% of my executive team's time is wasted every single month with distractions, from the board. The chairman just calls me and says, our relationship is so dysfunctional. He's going to give me six months to turn around, but it's by the second board meeting, I haven't done it. He's going to fire me and my whole executive team. And I asked him, what are you going to do? And he goes, honestly, Jeff, I don't know. I swear it's not us. It's them. It's like these guys have all this childhood trauma. They've never healed. They think we're their punching bag. I was hoping you could tell me, hey, I could help. Doesn't seem like the obvious use case. I think one of the greatest opportunities for people is to realize, don't go looking for an AI use case. Realize whatever problem you're solving right now is the AI use case if you know how to approach it correctly. And so we pulled his exec team on Zoom right then, and I wrote a context role interview task. Context, I literally said, I'm the CEO of and I named the company. We've been racing toward an exit over the next 24 months. Literally, Cody, everything I just told you word for word was the context section. role. We said, you are an HR professional with deep expertise in creating personality profiles. Interview us, ask one question at a time of the five questions to gain deeper context about one board member. And then your task is to create a personality profile for them we could use to train a project to create an AI board. AI conducts an interview, five questions on one director named Susan spits out a profile. I immediately look at the CEO and say, don't trust it. This is another one of the mistakes people make with AI. They're getting an answer and they're choosing to believe it as the final result. They're judging it. Viewed as a first draft. I told him, tell it what you like about it. What's good about it? What do you not like about it? What are the top changes you need to make to it? He gives that feedback. The updated profile was amazing. We did that for every single board member, pulled that into a custom project and then gave it instructions on how we wanted it to act as an AI board. To test it, we took the deck from last quarter. We pull that in and said, give us your feedback on what you think the board will say. How long do you think it took it to read over 60 slides as every single director and tell everything that they thought it would say. 90 seconds. Five. Yeah. Five seconds. One of the insights it said was on page eight, Susan's going to get distracted by all the details on the slide. This is going to lead to a 30 minute detour. It's going to derail your entire meeting. Instead of all the details, just say these three things. It's what she cares about most. I look at the CEO and he's like, that's exactly what happened in the real meeting. So we know it works. To make it better, we get permission from the board to use an AI notetaker in the upcoming meeting. We get an actual transcript of what they say. We pull that back to the AI board and told it to compare what it had simulated to reality and tweak the personality profiles so it could have simulated reality. Day of the second board meeting comes, my phone rings. And I asked the CEO, what's the verdict? And he said, well, the chairman just called me. He said, this is the best meeting we've ever had and the best deck we've ever seen. It completely transformed their relationship with the board. So instead of just hammering away at how do we get this board to get off our ass and be hostile, that we were able to just with a really simple prompt, get AI to interview us about the personalities of every board member so it could simulate them so we could anticipate what they might say and receive coaching on how to interact with them to transform that relationship. Could we rapid fire? What are all the ways that people are not using AI right now that they absolutely should? I'll give you one way that if you just focus there, you'll get 80% of the value. What is the biggest problem that you are facing right now? That if you could just solve it, it would unlock a huge amount of value for you. This can be personal or professional. And then instead of going to AI and asking it questions like it's Google, go to AI and engage it as a strategic thought partner to interview you about how you're currently thinking about solving that problem and then show you the non-obvious approaches you could implement that would lead to faster results. And a very simple framework for that is crit. Give it lots of context, assign it a role where you tell it the kind of expert you want it to be, but then make it interview you by asking one question at a time between three and five questions to gain deeper context so it can then accomplish a task, which is whatever you want it to do. So you have a formula where you can get AI to become your, you know, smarter, limitless, blue-pilled friend. And it's called CRIT. Can you break each one down quickly so that somebody listening can use AI way better than 99% of people out there? So CRIT is just a really effective way to talk to AI. It stands for context, role, interview, task. Context. While AI has been trained on the information in the world, it doesn't know you, Cody, your business, the way that you do. You can't give it a short answer. You got to give it lots of context about the situation. The more, the better. Role is where you describe the kind of experts you want AI to be because you literally have access to expertise in the world. And it's not like saying you're a marketing expert. Be really vivid. This is a creativity. You're an expert CMO with deep expertise in the CPG space, but you're really good at understanding the nuances of the whole food, healthy, holistic, like you get really specific about it. Cause I mean, it's tapping into 500 million books worth of data, access to the expertise of the world. If you're really clear about who you're looking for it to be, it'll be that type of person. So you give it context, you assign it a role. The most important part is the interview. I just don't ask AI questions anymore. I make it turn the tables and ask me questions to pull deeper context out of my head than I would have ever thought to have shared, but also to elevate how I think and to shift how I lead. That's a game changer. And with all that additional context, task is whatever you want AI to do. You aim that at any problem, at any goal, it'll blow you away at what this thing can do. It's such good advice because I remember when you first were talking to me about this, I have a real problem with AI slop, which I think is actually what most people are doing now. Most people who use AI are not using it agentically, which would be probably the most forward way you could use it now. but they're just asking a question, getting, and that question is usually a sentence, two sentences, getting a really big readout that has so many missing parts in it. That's hyper generalized. That is immediately obvious that it's AI. You know, we had a thing the other day I was interviewing a candidate and it was for a pretty big job. And I'm like, well, you know, give me the breakdown of how you would go about this. Like, I would like to see what a brief looks like from you on this guy must've gone to chat GPT, you know, listed out a few questions and wrote me the most obvious AI chat GPT brief of all time. And because of that, didn't get the job, didn't get a second interview because that showed me like the level of detail that he was thinking on. So I feel like a lot of times people with AI, you can go to the lowest common denominator, which is like a slop cannon, or you can go to actually a little bit of a superhuman with AI. What do most people do wrong with AI besides asking the questions? Because I found that was so important when you taught that to me. You familiar with the 80-20 rule? Yeah. So 80% of your results come from 20% of your efforts. Most people are using this for the 80% tasks that only drive 20% of the value. They're not asking what are the 20% applications that could drive 80% of the value in my business or in my personal life. They're defaulting to using it like Google or to write better emails. It's not entirely their fault. Most people spend most of their days in the 80%. That's just like this huge challenge across the board in work today. It brings some value, but it's not transformative. And by the way, I used to be that guy. When I first started using chat, I was a C-level exec for a public company based out of India. and I was literally using it to write better emails. And I remembered thinking, my greatest superpower is asking leaders the right questions to drive growth. Here I am asking AI questions. And that's when I wondered, Cody, could I turn the tables and make AI ask me the right questions to make me think more strategically? Because I know how valuable that is in a business or a personal context. And that's when I discovered this idea that you are the thought leader. AI is your thought partner. And if you choose to view it that way, it can truly be transformative. And I'm talking any generative AI tool, ChatGPT, Claude, Copilot, Grot, Perplexity, it does not matter which tool. We can deliver magic for a company. I saved one company $300 million in 10 minutes with ChatGPT because I knew how to get it to ask me the right questions. So I was doing a workshop with a small group of CEOs. And I asked them the question I ask most people, what is the biggest problem that you are facing in your business that if we could just solve it, it would unlock a huge level of growth for you? And the question itself is really strategic because for a guy who wrote a book about AI, I don't care about AI. It's not our goal. We all want to become a better leader who builds a better business and creates better lives. So how do we use AI to do that? One guy looks at me and he goes, I have a very big problem. I run a manufacturing company here in the US. I leased all this equipment from this company in Japan. But since I did that, the market has fundamentally shifted. That debt structure is killing us. We're going to go bankrupt if we can't get it restructured. And I go, okay, what have you done? He goes, man, I think I've done everything. And he lists five very specific strategies he's deployed. Cody, none of them have worked because the board of the company in Japan that holds the debt is refusing to allow the restructure because they think they'll lose face in Japanese society. He looks at me and goes, I have no next steps. We are going out of business. Do you think AI can help? maybe i literally opened up in this case chat gpt wrote context role interview task into the context window and in the context section i literally wrote i run a manufacturing company here in the u.s i leased all this capital equipment from a company in japan things have shifted in the market the debt structure is killing us we're going to go out of business if we can't get it restructured. I feel like I've tried everything I said to you verbatim was the context section. Role was you're an, you are an investment banker with deep expertise in restructuring debt. Interview me, ask me one question at a time, up to three questions to gain deeper context. Your task is to give me five non-obvious strategies I could deploy to get the board to restructure the debt. Here's what's interesting about this. If you think about how many books you've read in your life, whether it's 100, 500, 1000, thousands, what percent of all that knowledge can you recall and apply right now? Almost nothing. And look at how far we've gotten in our lives. These models have been trained on about 500 million books worth of data. It can recall 100% of it. In under a second, it was able to read through 500 million books to simulate an investment banker with deep expertise in restructuring debt who also knew 2000 years of Japanese culture and history. And with that knowledge, it asked a question. Do you have any relationships with any other executives in Japan that the board would respect? I look at him and he goes, wow what a great question i would have never asked that oh my gosh i do it asked two more questions like that before it turns the tables a final time and it said here's your five non-obvious strategies number one it called the saving face consortium said you have enough relationships with all the right people in japan why don't you just approach them to acquire your debt give them really favorable terms get debt get restructured the board saves face i look at him he's actually holding back tears. And he admits to everybody in the room, I haven't slept in 90 days. I've been making peace with the fact that we're going out of business, but in less than 10 minutes, I got hope. Two months later, my phone rings. It's a text from him. And he says, ball is moving. This is actually going to get done. It's amazing. And so for most people is what happens that we don't ask, we don't give it enough. So if you want to use AI right now, what you're probably doing wrong is you don't give enough context. You don't assign it a role and you don't let it ask you questions to get to the final results. And if you do those three things, then you can actually make your partner smarter and, and, and then focus on the output as opposed to, we just say, give me the output up front. I would say that falls into skill number two. Yeah. There are three skills that somebody fundamentally has to learn to really be able to harness this in a way that enhances them instead of just being a replacement. Step one, can you identify a 20% use case? I don't care how good your prompt is. If you aim it at something that doesn't matter, it doesn't matter. You can use this to write a better email that is not going to change your results. You use this to solve your biggest problems or to help you achieve your most ambitious goals. That matters. So pick something that matters. Two is, can you communicate effectively? Just like a marriage, this is a relationship with technology Crit is just a really simple framework to communicate really effectively with AI And then number three is can you remain the thought leader When it gives you an answer can you be skeptical Do you not trust it and submit the AI slop like that person did with you, but instead really be thoughtful about, here's what I like about this. Here's what I don't like about this. Here's the top changes I want made to this. When you interact with it that way, it then becomes a true thought partner going back and forth, not over one prompt, over a conversation. This will enhance the way you think and collapse the time it takes you to get things done. Quick aside, here's the thing about everything Jeff walked us through. AI can analyze where the growth is, find places to move the needle, even map out a plan, but a plan still takes execution. You know what kills the execution? You hesitate. I see it constantly. An owner knows the move. You hire, then you do the second location, the inventory you need to stock up on, and you sit on it for months because getting funding feels scary or buying stuff feels scary, and might give you the clarity. Capital is what lets you act on it. So let me tell you about a new business lending product from SoFi. There's something called SoFi Small Business Loans, and they're built for growth. With a quick digital application, you can receive a decision quickly, and if approved, get funding as soon as 24 hours later. But here's the part I actually really like. The pricing is really clear and upfront. So you see your terms before you accept. No surprises, no penalties for paying off your loan early and repayment terms tailored to fit your business, not the other way around. It's built for solopreneurs and small businesses. So if you're looking for capital to buy, build, grow, head to SoFi.com slash Cody or click the link in the show notes. Your business moves fast. I think your financing should too. SoFi small business loans are originated by SoFi Bank N.A. subject to eligibility and approval. Terms and conditions apply. This is a paid SoFi partnership. What are some other hacks to use AI better than anybody else? Speech to text is a big one. Yeah. I think about a quote that I, a piece of advice I was given my senior of college when I asked, what job should I get after school? And the CEO tells me, you're asking the wrong question. He said, focus on mastering the skills that will be so valuable, they'll serve you no matter where you go. Phenomenal advice. AI is changing what skills matter. speech to text is a skill that matters. I don't know this to be certain, but I think typing is going to go away. Yeah, I do. I think they'll re-engineer the computer entirely. Like it won't even look like this. Totally. And with AI, context is king. So when you type, you naturally edit yourself. You care about punctuation, sentence and paragraph structure. None of that matters with AI. And people feel sometimes uncomfortable talking to it because their thoughts aren't organized. That's okay. You can just start pressing record and you can pause and like you can be a hot mess. AI can handle all of this and the more you give it, the better. So that's one. Another pro tip, when you are writing crit, when you think you've given it enough detail, assume you have not and ask what else? And when you think you've given it enough, assume you have not and ask what else? So like when I'm, I mean, I did this with you. I took you three levels deep on the context section because it's just, it's not enough. So I think this habit of give AI more than you think it needs and the quality of the results will go up drastically. Another pro tip is that I play AI against itself. So once I get a result and I tell it what I like, what I don't like, the top changes, and I go back and forth multiple times and I like whatever I'm looking at, I recognize that's just the best that I can do. Not the best that can be done. so i will often have ai flip personas flip roles and say now be the challenger your world-class it's stress testing the insufficiency of everything i've come up with as that person i want you to tell me what you like you don't like the top changes it finds all the cracks in the foundation it highlights every bias every assumption that i've overlooked and once i work it there i'll then say, great, now be my ideal customer where I build actual personas of our customers. So we can simulate the CEO of this company. And as that CEO, it can give me feedback on our recommendation before I even give it. So I really get AI to play against itself to elevate the quality of the results. It's funny you say that because AI can be too kind to you often, I've found. And so, you know, it'll just say, yeah, that's totally reasonable that you could hit this goal when you're typing in unreasonable rules. It'll say, yes, of course, you're probably right in this endeavor. And one of the ways around that is by saying, no, now you have to attack it. So how do you get AI to not just be rose-colored glasses? Yeah. So there's two ways. One is you can do it in that very specific prompt. You can say, don't just agree with me. I love to have it be the aggressive growth-minded board member as part of the role. Don't just buy what I'm saying and tell me I'm great. I want you to push me for the next level. So this is how I talk to it. I have that's in the single prompt. The next level is once you figure out how you really want AI to talk with you is you put it into the custom instructions of the model. So whatever tool you're using, whether it's chat, co-pilot, Claude, any of them, you can give it instructions that will carry across all of them where mine knows I'm a high D. I have zero patience pants. Give me the 20% that's going to drive 80% of the results and don't fluff me up. You need to tell me, be the challenger, red team everything, fact check everything so that what you send me is solid. That's in the custom instructions. I can do any prompt and it'll give us an answer. And then it's going to say, here's my red team automatically. You have a use for AI and sticky notes that I thought was fascinating. Talk to me about that. So we get to go into so many companies and really change the way that people view AI. But how many keynotes or workshops have you sat and were like, that's amazing, and you did absolutely nothing? I didn't want that to be the case. It impacts a core value for me. I really want people to change. What's missing is a trigger, a trigger in their environment that will cause them to change. And I have found, I did this myself. I literally put two sticky notes on my desk. One sticky note said, how can AI help me do this? because I already said this. Don't go looking for an AI use case. Realize what you're doing right now is an AI use case. So that first sticky note of how can AI help me do this is that trigger. And then the second sticky note says context, role, interview, task. Literally sits on my desk. I was reviewing my financial statements one month and I'm literally going over it. And I'm going, this makes no sense. Like, what do I need to know here? And I looked down and I saw the first sticky note. I'm like, huh, how can AI help me with my financial statements? What would that prompt even look like? And then I see credit. I'm like, come on, Jeff. And I wrote context. Here's our financials for the last month. And I drag and dropped it in. Our chat GPT is locked down. So it has the security. Your role is to act as a strategic CFO who's world-class at telling a CEO the top five things they don't know about their business based on their financials that they should know about their business? Interview me. Ask me one question at a time, up to five questions to gain deeper context. Task, tell me what those five things are. Cody, that was so mind-blowing for me that I have now gotten to the point where I have built an actual agentic AI CFO that literally is in my email every single day. If I get anything from my finance team, it reads it, it opens up the Excel files. It has got probably 10 pages of instructions on how to run analysis on our financial statements. And based on that, it sends a Slack message to our finance channel with what are the things that we need to know about the business based on the financials. I mean, this is world-class level analysis. It took me 30 minutes to set up when I was on a plane because I got bored. And it happens automatically now. Yeah, the part that's crazy for me is that once you have your first agent set up, which I think can be really intimidating, actually. And I'm relatively tech savvy. And the first time I set it up, it's like, what the fuck's terminal? And what are we doing here? And it's gotten easier every two weeks, I feel like, the entire AI landscape changes. But I think it's worth playing around with if somebody hasn't done agentic AI. Because what you'll find is every small little thing you hated to do, it doesn't mind doing it and remembers the first time how you asked to do it. And then all you might have to do is reset it every so often and say, like, why haven't you done that? You're supposed to. So here's where I have a different approach to most people with AI. People hear about building agents like I got to build an agent for most people. That's a distraction. I really think of AI adoption like dominoes. if you stand them up and line them up correctly, how do you knock them all down? You just tap the first one. Agentic AI is like the 18th domino. Lead domino is the three skills I mentioned earlier. Can you identify a 20% use case? Can you communicate really effectively with crit? And can you stay in the driver's seat as the thought leader? Not trusting what it says. Tell it what you like. Tell it what you don't like. Tell it the top changes. If you do that once a day, every day for the next 30 days, you'll be playing in a different lead. You won't even recognize yourself. So just start there. If we want to talk about agents, you have to realize you don't need an agent. That's a solution looking for a problem. What's the problem you have where if you could use AI to take it off your plate, it would bring value. And here's where you have to think about your job. your job is just skills you apply and processes you follow so like you look at your job here at contrarian thinking there are certain skills that you apply and certain processes you follow i can segment those into two groups the 20 percenters that drive 80 of the value and everything else what i'm interested in is how do we build an agent that either allows you to do the 20 better or to take something annoying that's 80 off your plate and in that case you're looking at what skills need to be documented, what processes need to be documented. Crit is actually phenomenal here because you could have it, you could give it context about what that skill or process is. The role is to be the perfect agentic AI builder, have it interview you to gain deeper context. And the task is to create the markdown file. Markdown is the format that AI loves, the markdown file that it could save so that an agent could replicate that without you being there. And it's just about recognizing that any knowledge you have in your head for how you do your work that is not documented is an opportunity to document. Use AI to interview you to create that documentation in Markdown format. You save it in a file folder. Then you give an agentic AI platform like Claude Cowork the ability to tap that. And all of a sudden it can replicate you. It's so good. Also, when you realize all you have to do is just talk to your computer. Just talk to it. all right, I would like to build this thing. I don't know how, what would be the next steps? Can you walk me through it? I don't know what a terminal is. What's a terminal? And it'll literally take you through every single step. That's literally how I built my AI CF. I was like, here's my vision for what I want this thing to do. Assume I know nothing about how to build this thing. You're like my eighth grade teacher. You need to walk me through this. Like I'm an eighth grader, just step by step, help me do this. Yeah. It's so good. And that way I think it doesn't have to feel so overwhelming for so many people. Because sometimes if you go on the internet these days, everybody's an AI expert all of a sudden and everybody has 400 agents doing all their work and they've made $10,000 in 20 minutes. And by the way, that's all BS. That's all BS. If you really, yeah, if you pull all those things apart and ask, number one, did you build things that actually matter? Most of them, no. The number of people I talk to who go, oh, I save six hours a day with AI. And I ask, so what do you do with the other six hours? oh, I study new AI tools. They've actually gained no business value from what they are doing. So the real opportunities, are you using this in the things that matter? Is it showing up in revenue? Is it showing up in profit? Is it showing up in better retention or better CSAT scores? How are you using this in a way that matters? So let's get tactical. What would you do as a framework to determine what your 80% is that you're doing as opposed to your 20% that you're doing. So how do I know I'm doing the highest value tasks to AI? Let's start with the 80 and the 20. You can literally get a piece of paper and a pen. I'd go analog here. I would not go AI. I would write down everything you have to do this week. Everything. Meetings, email, phone calls, Slack, proposals, board report, like I'm literally thinking through all the things on my plate this week. And once it's all on there, I would start with the first one. And I would ask, does this single item represent a 20% priority that's going to drive 80% of the results in my business? That is an incredibly high bar. If the answer is no, just skip to the next one. When you get to a yes, just put a star next to it. Very fast. You go from all the things you could do down to the focused things you should do. And then I ask the question, if I can only do one of them this week, what is the one that is actually going to deliver the most business value? And I put a number one next to that. That is literally how I prioritize my life. That's how you quickly identify the 20% from the 80. AI throws a bit of a curveball. Because what we do as humans is going to change. If you study technology shifts throughout history, it didn't take jobs. Remember, a job is just skills you apply and processes you follow. What it did is it made the value of certain skills and processes skyrocket in value, and the value of certain skills and processes plummet in value. If one of the 20% things you're doing is something that AI can do better, you should not be doing it anymore. In fact, if AI is 50% as good as you right now, you should start working to have AI replace it because within 12 months it will be as good as you and then it's going to run away with it. the things that I think are going to matter from a human skill standpoint moving forward, the ability to think strategically, to solve problems, to communicate, to collaborate with people and machines, and to create, to be creative. Those are the areas where you will see this become a massive enhancement, not a replacement. But most people, frankly, their job is so focused on outputs, do this report, send this message, generate this proposal. That's machine-based work. In the industrial revolution, it required humans to do that kind of work, not in an AI-driven world. What do you think are the first jobs that go away with AI? Customer service, frontline customer service, paralegals. And maybe you'll still have paralegals, but document review, document generation, that type of stuff that goes away. I think a lot of entry level work goes away. And this is actually a bigger problem because you have to get the entry level job to develop the context to move up the ladder. And this is going to be one of the bigger gaps in the workforce is how do we bridge that gap? When kids coming out of school, we're not going to hire somebody to do this work because an agent can do it for basically free. How do you leapfrog somebody to this level of performance when they haven't had the reps? Let's say you have kids that are like of the age of going into the workforce. Yeah. What jobs or skills would you absolutely never let them do because of AI I am very skeptical I not saying you don do this I very skeptical of the number of people going into law, mostly because most of them go to law school because they lack direction, not because it's what they actually want to do. I think that industry will absolutely remain. the skills are going to be fundamentally different. I don't have a strong opinion on that. Where my opinion is, is if you are going into applying for a job where they still operate in a traditional hierarchy where you report to a boss, you show up to work, take direction from a boss and do something repetitively, I would not work for that company. And here's why. That's how we had to work in the industrial revolution. In an AI-driven world, AI-driven organizations are far more like networks and swarms rather than this hierarchy of top-down. It's far more, can leadership cast the vision for where we're going and the strategy and clearly communicate it? And this is about a company that's going to say, Cody, what are your God-given superpowers? What is it that makes Cody Sanchez, Cody Sanchez? And this is not a long list. This is probably three things. that truly make you, you? And how do we actually get clear on what is it that makes you, you? And how do we harness those strengths and then focus you on here's the 20% things of this role that these two things got to match up with. Here's how it lines up with our mission. And now I'm going to give you agency to harness AI, to build agents so that you can surround yourself with technology as well as human teams to go deliver more value than you would have ever thought possible. Yeah. Yeah. We call them daycare employees versus department chair employees. And I think historically you could have a lot of daycare employees. They need activity time. They need to be told what to do. They're there from a certain hour. It's not really about what the outcome is. It's about time in seat. And that existed for a long time. But now I think you're going to move to everybody becomes a department chair. I don't think you have to be an entrepreneur or a founder. And most shouldn't. Yeah. But you do have to have your own center of gravity. And if you don't, if you're waiting on somebody else to tell you what to do. You're literally acting like a machine. I know. It's very problematic because we humans forget. We don't put it on our calendar and we think it's okay to be reminded multiple times. Machines don't think like that. So I used to be partners with Gary Keller. Yeah. We started Keller Williams. I forgot about that. Yeah. The very first thing he ever said to me was on November 1st, 2015. He said, welcome to Austin. Good luck. You're going to need it. And I said, hi, thanks. Why? And he asked me, Jeff, do you know what the best part about your job is? And I said, no. He goes, that is your job. If you try to give me pieces of your job, you will no longer have one. That was the very first thing he ever said to me. I like it. Yeah. Now he and Jay, they hired me to turn their book, The One Thing, into this training and consulting company. And he said, look, you're going to want to get me on podcast, me on stage, me on video. But if you need me, then I don't need you. So go succeed without me. Cody, this was extremely valuable for me because how often do we look for our boss to tell us what to do? How often do we bring our work to our boss for their approval before we can submit it? We are allowing them to own pieces of our job. The best part about your job is that it's your job. You need to own 100% of your job, which means thinking leverage. If your boss has to follow up with you to ask if you got something done, you are not owning 100% of your job. It's very true. It also worries me for a lot of employees today. And I've been there. I'm sure I was at some point in time when I was younger and starting. But right now, if you are waiting for somebody else to tell you what to do, I mean, my little funny line is I always like to say when somebody brings me something like, well, what do I do here? I say, I'm really excited for you to figure that out. I believe in you. This is why you have the role. This is why the AI-driven leader is about leadership more than it is AI. There are things you can do as a leader to start to create this culture. So, you know, it is part of our culture. It is your job to identify what you believe the 20% priorities are that you need to focus on this week to be on track for delivering what you need to deliver by end of month. So you're delivering what you need to deliver by end of quarter. So by end of year in alignment with business plan, I'm not going to tell you what to do this week. You're going to show up to our one-on-one and you're going to tell me, you're going to show me your plan for what you think the 20% is. that's thinking leverage. Then it's my job to help you see what you don't see so that we can bridge those gaps. So week after week, after week, we get to the point that you can truly prioritize your work independent of me. How do you change somebody's mindset who thinks that that is like toxic culture? Like I think a lot of people today would say, well, your boss should help you and tell you what to do. You shouldn't be expected to do that. You're not running the company. That's like a common thread in society today. How do you flip that switch? Well, one, I didn't say that. My job as the leader is to be clear on what our goal is. My job is to be clear with you on what your 20% is and the results I'm expecting from you by end of year. But it's your job to show me how you're going to get there. It's your job to look at all the things that are on your plate and to ask yourself, what is the 20% that's going to drive 80% of that? And I don't expect you do it in a vacuum. That's why you get the opportunity to bring it to me. Show me your thinking. Show me you've done the work to think this through. You're not alone. We're locking arms. We're on the same team, staring at the same goal. I'm going to help you see what you don't see. Today, my executive assistant goes, all right, our agenda for today is one, two, three, four, five. And I immediately stopped her and said, are those in order of priority? and she goes no they're not did she roll your eyes yeah no she goes she goes they're not it really should be four three five two one i said good run the meeting that way because we may not get past the first item let's make sure we're investing in the things that matter most she's in her first 90 days that's me training her every single week but she's getting to the point where she used to show up and wait for me like first time she's we meet she like looked for me to drive I said, this is your meeting. I'm going to give you five minutes to think about what do you think the most important thing is. I'm going to go get a new coffee. When I come back, you're driving the meeting. That was our first conversation. And by the way, that's me empowering you. That's me saying, I believe in you. You're talented. You are capable. And I believe that you can think this through and you're not doing it alone. I'm going to be there to coach you and to guide you. I could tell so many stories of how people have grown inside my orbit as a result of this. It's crazy. That's really the only way to grow is the unfortunate thing is that I still don't think society thinks that this is the way that you get better. I think. I've also found myself asking, who do I want to serve? I've realized that with our company, AI leadership, if we have to convince you that AI is important, you're not our customer. We're for the people that know that they need to become AI driven and they're running into the wall. They're taking such massive action. They're running into the wall. My job is to take away the wall. What about people who say that AI is bad and it's kills creativity and it's bad for the environment and why are you guys even talking about this um creativity is a big one that i hear often i would ask them might it be possible would you change your stance if you found that this could help you become more creative they can't say no to that and then we do something together i mean this is such a creative enhancement i was with this last week one of the biggest brands in the world. I'm in the boardroom with the C-suite. They gave us the biggest problem that they are facing. I mean, you're talking billions of dollars at stake. We wrote a crit on how to solve it, totally came up with viable solutions. That alone was worth billions of dollars for them. That was a 20-minute use case, but it requires massive change, massive mindset shift. And I said, I had a, I flip it and I said, great. Now I want you to be Steve Jobs. Imagine you got to come back and deliver not a message, but a manifesto that I could deliver as the CEO of this company to our workforce that would ally all fear, all anxiety, and would galvanize people to run through walls with excitement. And this thing would be so damn good, it would put the release of the iPhone to shape. And it lays out this manifesto that could be delivered in a minute. The CMO of the company goes, can we put that on letterhead? I mean, that is, the creativity is not your ability to create the draft. Your creativity is the ability to cast the vision for where you want to go, to be descriptive with AI on the types of things you're looking for, to be creative in terms of the roles that you pull in or the type of tasks that you wanted to do. The difference is people who are creatives have been creative with design, creative with this specific type of art, creative with writing. Like you were the player of the instrument in the orchestra. You're no longer the player of one instrument. You're now the conductor front and center who's waving the baton and creatively engaging AI as a business analyst, then transforming it into a world-class communication expert with deep expertise on your IACP, and then turning it into an aggressive growth-minded board member who identifies all the non-obvious growth levers that you could be pulling that you're not pulling. Like even listening to me talk, like my brain is firing on such a higher level right now, even thinking about how I could use this thing. let's say that somebody's listening right now and they want ai to help them become uh more prolific in their writing you know they want to write way more but in their voice the way that they do it just a very tactical use case like what would you do to go in what would you upload what would you include yeah in order for somebody to turn their ai into their little mini version of that? First, I would feed it 20 to 50 emails. I would feed it past writing samples of what you would say is your work. And I would include all of that as the context. The role would be you are an expert prompt engineer, but also communications expert whose true superpower is actually creating a writing persona. I want to have a style guide of my voice that I can embed into AI and your world class at figuring out how to do that. Interview me, ask me one question at a time, up to five questions to unlock deeper context that you need to do this thing. And then your task is to create the file in Markdown format that would be my voice, that if AI ever were to write anything on my behalf, it would nail this. I have that, by the way. I have that. so how many m dashes does it put back in there that you don't want it oh see i have it in my custom my global custom instructions to never use hyphens are okay when they're appropriate yeah and dashes are not yeah and so it still doesn't mine still sneak them in i still gotta i still gotta keep my eyes on the prize there no i feel like there's still a 10 to 20 failure rate even with the best prompt that is the best thing ever and you still go in there and go that's an m dash that's a it's not x it's y like what i mean i'd be curious your take what are some obvious things that you see that you go that is ai and not in the good way m dash is number one yeah um there are certain things that opens a paragraph like in the ai era in this transformation like it's almost it's it's a little verbose hypey inauthentic yeah um i can also just tell when i'm reading and there's a lot of words that aren't delivering a message yeah it's just too much it's a lot of fluff yeah it's a lot of fluff and all that says to me is by the way i want people to use ai yeah i want you to use it to get to a draft faster but all that tells me is that they allowed AI to become the thought leader. Yeah, that really comes back to your point. Because for me, I'll lose my mind on myself and my writing or the team if there's the ever-present M-dash, that it's not X, it's Y. What nobody else will tell you, says that line a lot, in the era, uses the word era a lot. There are a few things that must be just so prolific in human communication that it assumes we do that constantly. And I think sometimes I think employees are scared a little bit to sound robotic in that way. But the cure is really just remember that you have to keep leading from the front. Remember whatever draft, whatever it gave you is just the first draft. And before I even look at it, I just tell myself, this is the bad answer. I don't care how good it sounds. This is the bad answer. I'm about to make it better. And then I review it and I literally tell it, here's what I like about what you gave me. Here's what I don't like about what you gave me. Here's the top changes we got to make. And then it creates an updated draft. And then I do it again. And I keep doing it back and forth, back and forth, back and forth until I like where I'm at. Yeah. What about, you have an AI board. You know, your AI board has a bunch of members. I'm curious. Do you think everyone should have an AI board? And if they should, who's on it? You should have an AI board if you are in a position where it's lonely at the top. and you would like to have advice on demand from people that will, or personas that will have your best interest at heart that are custom designed to augment your weaknesses. So I am the founder and CEO of a company called AI Leadership. I self-funded this thing. I have no investors. I have no board. At the end of the day, the buck stops with me and it's lonely at the top. and I see the value in boards because I'm inside a lot of public companies and I see the role that that thing plays. So one day I wrote a crit where I fed it my tenure vision. I fed it our business plan, culture documents. I fed it my personality profiles. I told it in the context section everything that I thought were my strengths and weaknesses but also called out. Those are just the ones I'm aware of. And the first crit, It had to interview me to identify what my other strengths and weaknesses were that I was unconscious of, that I needed to be conscious of. And based on that, to identify the skills that I was missing, that if I were to assemble an actual board, I would want to attract those skills to the board to get me to where I want to go over the next 10 years. And it gave me this list of skills. I didn't trust it. I applied judgment and I called it down, but like one was vision, one was product design, one was storytelling, risk mitigation, finance. And I said, great. Now research famous people, world-class business people or leaders who exemplify these skills. Well, three of those skills were taken up by one guy. Vision product design and storytelling who better than Steve Jobs So I literally built a personality profile for Steve Jobs but custom augmented to me So he has rules on what he can do and he has very specific rules on what he can't do. He is not allowed to give me advice on being a husband, a father or a leader. I'm dead serious. I'm not interested in his advice in those three categories. Warren Buffett is on my board for long-term planning and risk mitigation. Jeff Bezos is on my board for scale and operational efficiency. See, my most valuable board member is my future self. I created a persona of the man I want to be 30 years in the future so that that version of me can advise me every single day. This is actually wild. We've taught 1,500 business owners how to increase their revenue inside of one quarter by anywhere from 26% to 100%. How? Well, we stole the playbook from private equity. We use something called the 12 Ps, which is the framework that takes you from whatever your revenue level is now to a higher level plus more profits. You see, the thing about private equity is they don't guess how to grow a business. They know every single time. It's just never been applied to normal everyday businesses. We stole the homework, integrated into the 12 Ps, and deliver something called the owner score. Your owner score in our boardroom is what helps you move your business forward. So if you right now are wondering, how do I make more profit in my business? How do I grow my revenue of my business? Why is this one problem continuing to happen? Maybe it's hiring the wrong people. Maybe it's not having enough sales. Maybe it's not having enough leads. I can almost guarantee you it is one of the 12 Ps holding you back, except no one's ever taught you this. If this is you, I want to meet you. And if you're a fit for our Contrarian Thinking boardroom, I may invite you out to our headquarters in Austin, Texas, to learn more about how we help small business owners double, triple their revenue with just one of these 12 Ps. You can click the link below to see if you are a fit for the contrarian boardroom. So you say that AI could lead to cognitive decline or decay. What's happening there? How do we know if that's happening to us? I think this is the greatest risk with AI that is not being talked about is unfortunately, a lot of people are lazy and they're not realizing it, but your ability to think is a muscle. If you stop using it, it's going to get weaker and weaker over time. Most people are using AI in a way that it is replacing their ability to think. Anytime you ask AI a question, get an answer, and you just accept it. You made AI do the thinking leverage, not you. You can absolutely flip that, though, in a way that this requires more thinking on your end. And this is where Crit has been really good because when you describe the context, you have to think like, what's all the stuff that I need to tell? And you're recalling all these different things. Your brain is working more. When you describe the role, you're getting really creative about the kind of experts that your brain is working more. When the AI turns the table and starts to interview you, it asks questions that you would never ask yourself. Your cognitive demand goes through the roof. And with all of that, it then accomplishes the task, which you then still apply judgment and discernment. And then you're thinking more about what do I like? What do I not like? What are the top changes? So if you actually look at when you write crit, how much more you use your brain, it goes through the roof. In that case, it is enhancing me. Unfortunately, most people are allowing it to replace them. So how do you know if you are allowing it? to replace you. If all you do is you take the answers, you don't allow it to ask you questions. That is the biggest indicator. I think, well, one, those people aren't even aware that they're doing it. So there's a lack of awareness there to begin with. For me, the thing that I pay attention to is when I try to solve problems, can I still do it without AI? I think about this a lot. And And it's a dance, Cody, because there are times where I'm so clear on how inferior my human brain is to solve things without being augmented by AI. I know what it feels like to have superhuman powers. And when you suddenly can't fly anymore, you know what it feels like to not be able to fly. It's a really good point. And it's hard to have patience with it. You know, I was with our mutual friend, Charlie, the other day. And he's like, well, we can just upload this into Claude and Claude can give us it. And I was like, no, we have these sticky notes on the window. I said, I want us to think through this critically first. Then we can take pictures of it. We can feed the whole thing through AI. But I actually want us to just make sure that we like the output before we take an output. And I can see how we actually would get, it's annoying because it could be so much faster if you allow the computer to do the thinking for you. And so it can feel dumb. actually, to not do it. Well, I mean, in my office at home, I have a nice leather chair that is my thinking time chair. This is very manly of you. I like this. I only sit in that chair when I'm doing strategic thinking. Really? And I'm talking pen and paper strategic thinking. So the guy who wrote the book about AI uses pen and paper still, because I sometimes need to sit down and really process my thoughts. and it's not a one-shot thing, sometimes I need to sit there for an extended period of time to really think through, what am I missing? What am I not thinking about? And I get all that context on the page, then you better believe I'm bringing in my computer and going, here's the question I just asked myself. Here's all my notes from the result of my thinking time. And I'm talking to this thing, dumping all that context into AI and then having it play a very specific type of role to ask me deeper questions to pull even more out of my head to then do whatever tasks that is, whether it's creating files that I'm going to build into an AI agent, whether it's using it to harness that against our strategic plan, whether it's communication that's going to go to the company. That's what it looks like. Top three AI tools to use in 2026. Claude. That's fair for everything. Sad. No, we need some competition. We can't have another Google that owns everything. But yes. Okay. I agree with you. That's the one I go to. Although, you know what? Actually, I don't agree with it all the way because you know what I still like ChatGPT for? ChatGPT does a really good job of very short, prompted image creation. I still find even with Claude Design, Claude Design has a design that looks like Claude and you have to beat that design out of it still. I'm sure that'll be gone at some point, but Chat actually does not have a specific chat design, which is interesting. And that's where I can give you the high level answer. And then there's what's the problems that you're facing? Yeah. You know, my videographer will have a different answer in terms of the AI tools that he's using for that versus, you know, if you're in graphic design space, like you're going to be using something very different. If you're focused on research, you're going to be focused on something very different. But I think just broad based across the board, Claude. By the way, even the way you described that use case, was your brain working more or less? Oh, yeah. More. I have to. Yeah, exactly. Like I watched your eyes. You started looking up into the right and you're visualizing this thing. Your speech started talking faster. Your brain was firing. How many AIs do you use at once? Do you ever have like Claude over here working on an issue, chat creating an image for you, Higgs field rendering a video simultaneously? Sometimes, but that's not the 20%. Really? And this is again where I really focus on the 20% that's going to drive the 80. For most people, can we just identify a 20% priority that matters, write crit really well, and then apply our thought leadership to the situation? That's going to take you so much further than what tool and how many can I use? That is so what most people are focusing on. That's not going to transform the EBITDA of your company. That's not going to change the competitive positioning of your company. Now, when you do this really well, and then you start to build agents, do I have Claude running different things at different times? Sure. But that's not the goal. It's can I focus on one thing that matters right now and do it really well? And if that's a repetitive task, can I turn it into an agent so it starts to run on itself? And can I focus on making that thing really good so that it actually does bring me 100% leverage? Otherwise, it's not that valuable. And then I can move on to building the next one. What do you think about having centralized AI brains in companies? This is a little bit 202, but having a centralized brain in your company or having decentralized brains per division and also having decentralized agents in each division. How do you see companies doing that today? Let's call it companies that are doing anywhere from one to a hundred million bucks. They aren't. They aren't. Less than 5% are doing that. Less than 0.01 are doing that. And in most, the vast majority of companies, people are still using this to write better emails or to be their Google. They say they're building agents. They're not. They're building a project or a GPT. true agent that can do things independent of you. Now, the smaller companies that are startups, like I do believe that they are starting to go down this. I would really challenge how much of that is the 20% versus the 80. I think a lot of people are doing a lot of things that aren't actually delivering results. So I'm not even worried. Like most, most companies shouldn't even be worried about that yet. If you are, that is a high level of sophistication. And my guidance to them is step one, I want as much data centralized as possible because it's a massive, massive asset. The question is what access to what data, which is that's when you can look at divisional stuff like that. And then the question is how is the data structured? Is it actually in the right file format, like a markdown where these models can actually talk to it? And more importantly, can you give your people agency to start to use AI to document what they do So it can do that for them. So it can liberate them to deliver more value inside the company. Yeah. I mean, I think the one thing we found with the centralized, overly centralized AI database bases are they really need sub agents for the most part. And those sub agents, you're right, you have to have a data lake and then that data lake needs to have protections for all the individual, you know, subsets that you need to have. But I think inside of a business, what we're seeing, at least with our businesses that we talk to, is there's usually one, this is probably your 20%, there's one part of your business, usually sales, because that's where the money makes. And inside of sales is like the first place where agents make sense or, and I don't mean agents that are going to go do your sales even. I mean, is there an agent that can comb through all of your phone calls that your salespeople have to determine what are the most common objections, feed you the best solutions for the clients that were actually closed, and then help you train your salespeople without having 30 minutes. I got two cool use cases here. Voice of customer is super important to us. Yeah. So we want to understand what's on the hearts and minds of our people. Every conversation we have with a customer, there's an AI note taker there. And we ended up creating something where with the note taker, if a customer says something that we think is voice of customer worthy, we click a highlight button. AI rewinds to the beginning of that customer talking for that segment, records until we say stop, built it. So that snippet gets sent to a voice of customer channel on Slack once a week. an agent goes in that slack channel finds everything our customers have said over the last week sends that to another application where it's turned into a podcast when i wake up wednesday morning by 5 a.m when i'm at the gym i'm listening to an actual podcast of everything my customers have said over the last past week that's how i prepare for our product market fit meetings well that's cool that's a good idea so that's just voice of customer yeah we hired a new salesperson and i asked the guy who i was promoting him she was backfilling him i said how are you gonna train He's like, well, you know, I got my scripts. We're going to role play. I'm going to listen to some calls. I go, nope. And he goes, okay, you clearly see something I don't see. What do you see? And I go, hasn't every sales call we've ever done in the company been recorded by an AI note taker? He goes, yeah. Don't we have conversion data on what's happened with every single prospect? Yes. Why don't you feed AI the last 100 sales calls and the conversion data, as well as what you think your script is, write a crit and have it come up with what the ideal sales script is based on conversion data on that volume. Turn that into custom instructions for an AI sales coach so we can build an agent so that when this person gets off a sales call, that note, the note taker automatically sends the transcript to the sales coach who compares that against the best that could be done and slacks them on what they did well, what they need to change within the next five minutes before their next sales call. Yeah. That's what we do too. We take all the data from Avoma and actually feed it into something that in real time can give you feedback on these objections. Yeah. And it's sort of voice triggers based on what you say on the phone call. It's been a game changer, especially for, we had an underperforming sales rep that I love. He's a great jet and traditional training just wasn't working. But this iterative real-time training is so much more sophisticated that now he's one of our top closers. And he was the one that he was, you know, speaking of being an employee, he was one of the guys that came up with the V1 of this as a non-technical person because he was struggling. He was going to get fired. And then he showed it to the team and now we're like, okay, let's roll this out across a bunch of it. It wasn't even top down. That's awesome. So it's really cool what you're doing. So for somebody listening today, if you had to give them one thing to do with AI right now, what is it? Once a day, every day for the next 30 days, write crit in a strategic way. That's it. Pick something that you're doing once a day that matters, write crit, context role, interview task. And when you get an answer, don't trust it. Say, here's what I like about it. Here's what I don't like about it. Here's the changes I want made. That will blow your mind. If you do that every day for the next 30 days or the next 30 business days, you won't even recognize yourself. And here's what's going to happen. Cody, it's really tough to describe color if you've only seen black and white. Most people are living in black and white. I mean, even the use cases we were just sharing, people are like, I could never come up with that. Yes, you could, but you haven't seen color yet. And when you start writing crit on the 20% every single day, you start to see color. Then you start to look at a process and you're like, why the heck are we doing it that way? And you will be so much more creative about how you can approach something, not by bolting AI onto the way things used to be, but by truly imagining from first principles, how might we do this if I was an AI-driven leader or this was an AI-driven organization? Jeff, you're the man. I love how you take AI, make it really simple, but profound. And I always push you. I'm like, what about this hack? What about this hack? You're like, no, no, no. It's this. It's nothing else. Do this one simple thing. And if you do it, AI does not have to be so scary and in fact can really be life changing. All right. Thank you for coming. Jeff Woods everywhere. And I love AI-driven leadership, AI-driven leader, the book. And I liked it so much. We bought it for all the employees. If you had told me that last time you were here, thanks for being here, Jeff. Thank you. you