This one is for my coaching members. I’ve helped several of you set up your first AI agent. Tomorrow, we’re going beyond the setup and looking at what you can actually delegate to it, plus how to teach it to do things that are useful for you. I’ll share some valuable skills you can give your agent to help with things like cleaning out your inbox, browsing the web, and shopping on Amazon. I’ll also show you a couple of tricks that make it easier to connect your agent to almost any tool. We’ll also look at Meta’s new AI agent, Muse, including what it’s good at, where it falls short, and whether I think it’s worth using. We’re meeting tomorrow at 2 PM PT/ 3 MT/ 4 CT/ 5 ET That’s one hour later than our normal time. If you’re not a coaching member but want to join us, message me for special pricing. See you there!

Most of us are still using AI like a smarter search bar. Open chat. Type what we need. Get an answer. Ask for the next thing. That is useful. I use chats constantly. A chat is great when you want to think through an idea, ask a quick question, improve something you wrote, or work through a problem one decision at a time. But even a really good chat is usually still waiting for you to lead. You start the work. You decide what should happen next. You keep the process moving. An agent is different because it has a job to handle, not just the request sitting in front of it. It can look at the work, decide what needs to happen next, and carry more of the process inside the boundaries you give it. For example, if I want help with a weekly email, a chat can absolutely help me. I can ask for topic ideas, pick one, ask for an outline, ask for a draft, ask for subject lines, and make revisions. The AI may do great work, but I am still managing every stage. An agent responsible for preparing that weekly email starts with more of the job already defined. It knows who the email is for, what I am selling, how I write, where to find approved information, and what still needs my review. That does not mean it gets to do whatever it wants. A useful agent needs: - A clear responsibility - Enough context - The right access - Boundaries for what it can decide and what needs approval You do not need an agent for everything. A chat is often the right tool. The simple test is this: - Who starts the work? - Who decides what happens next? - Who has to keep it moving? If the answer is mostly you, you are using a chat. If the AI can carry more of the process inside clear rules, you are moving toward an agent. That distinction matters before you start building anything, because a lot of people call every AI tool an agent when it is really just a chat with a nicer interface. Lesson 2 walks through the difference and when each one makes sense. READ THE FULL LESSON HERE

An AI skill is a set of instructions that teaches your AI how to do a specific job your way. Your process, your examples, your standards, your definition of done, all packaged up so the AI follows them every time. That's what gives you high-quality, consistent results. Without a skill, you get a different answer every time depending on how you happened to phrase the prompt. With a skill, the instructions are locked in, so the output comes out right again and again. And here's the part most people miss: skills are shareable. A skill captures how you do something well and packages it so someone else can do it too. Your team member. Your client. Another AI agent. Skills are how capabilities move between people and agents. That's what I've started building in the Classroom: skills you can install and use right away. If you've never used one, start with the how-to lesson: https://www.skool.com/ai-growth-lab-9975/classroom/1bd330d1?md=775b882059444981b5a68ec17fa07a58 Then grab the basics. Just pick the one that solves what annoys you most this week: Humanize - takes robotic AI writing and makes it sound like you https://www.skool.com/ai-growth-lab-9975/classroom/1bd330d1?md=73a187e05229491b81c214a3cbd9ae38 SlopMonster - cuts the AI slop out of your copy, the words that scream "AI wrote this" https://www.skool.com/ai-growth-lab-9975/classroom/1bd330d1?md=8489d1eb4e4c400fa008d183f9470027 Post Writer - writes your social posts in your voice https://www.skool.com/ai-growth-lab-9975/classroom/1bd330d1?md=d42535cb188f43aebda717d98f45da05 Viral Hooks - writes opening lines that stop the scroll

Alex Hormozi has built a reputation for being direct and specific about business, marketing, and sales. He's got a lot of documented thinking out there about what makes an offer work, how to price it, how to sell it, and what most business owners get wrong. ChatGPT has read all of it. So when you put HORMOZI: in your message, it filters its answer through that specific way of thinking about business. Instead of getting ChatGPT's usual general advice, you get something shaped by someone who has very specific opinions about offers, pricing, marketing, and growth. I used it when I was brainstorming my coaching offer. I didn't want a list of features or a generic pricing breakdown. I wanted pushback on the structure. I wanted it to tell me what was weak, what I was missing, and whether the offer actually made sense. And that's what it did. How to use it: Put HORMOZI: at the beginning of your message and describe your business idea, offer, or marketing strategy. Like this: HORMOZI: Here's my coaching offer. Tell me what's weak about it and what you'd change. or HORMOZI: I'm thinking about pricing this at $X per month. Is that the right structure for this kind of offer? or HORMOZI: Here's how I'm planning to sell this. What am I getting wrong? It works because there's so much publicly documented content from Hormozi that ChatGPT can draw on. It's not actually him. It won't be perfect. But it does shift the angle of the feedback in a way that's a lot more useful than a generic answer. And once you try it, you can swap in other names too. Any thinker with a lot of documented public thinking works the same way. Try it on your offer, your pricing, or your sales approach. What did it say that surprised you? 👇