There are five levels to making money with AI, and most people quit at the first one. I walk through each level from my own experience: selling my time as a $5k a month analyst, scaling to a 30 person team, over $6 million in digital products, 1.3 million followers, and the level I'm focused on now.

Find it here: https://www.skool.com/milesdeutscher/classroom/7f34f45f?md=34b3a56d05d242e38657367ddf2bf7f4

Show and tell: the architect/builder method. I use it in real life at my day job, and now every time I build with AI. At work I architect solutions for developers. The loop goes like this. They go off and come back with a plan, we talk it through, I add the things they hadn’t thought of, and we land on a better plan. Then they build and I review. With AI it’s the same loop, just with two separate chats. The architect chat plans with me. It asks the questions I’d forget to ask, never writes any code, and the plan goes into a file. Then a fresh builder chat reads only that file, tells me how it would build it, and stops so I can look before anything gets made. After that it builds one piece, tests it, and reports back. Why I like it, at work and with AI: it tends to find problems before they’re built. The plan gets talked through before anyone writes a line, so you catch the gaps while they’re still cheap to fix. One thing has to come first, though, and that’s getting organized. The plan and the decisions live in files, so you can close a long chat without losing anything and the next one picks up where the last one stopped. That matters because AI gets worse the longer a chat runs. If you’re in here you’ve got a head start, since the knowledge base setup Miles teaches is the same idea. It’s what I did first too, before clients or business ideas, and I’ve had clients come to me just wanting that part. I walk through the whole thing here: https://youtu.be/rRRsZmKoAho Anyone else split planning and building like this? Or do you plan in the same chat you build in?

Thanks to everyone who shared feedback. We've made some updates to make the community easier to use and safer for everyone. What's new - New categories. Posts now go into General, Q&A, Show & Tell and YouTube Videos, so it's easier to find what you need and help each other. - Community rules. We've added 10 simple rules to keep the feed focused on AI, useful and free of spam. https://www.skool.com/milesdeutscher/-/rules - Scam protection. There's a clear guide on how to spot fake accounts copying my name and what to do if one messages you. - Pinned guides. Q&A and Show & Tell each have a pinned post explaining what goes there and how to get the most out of it. - Admins on the community are: @Vashti K @Ryan D @Victor A 👉 Start with the Welcome post Everything is in one place: where to find the free resources, where to post, how to stay safe from scams, and the community rules. Please take 2 minutes to read it: https://www.skool.com/milesdeutscher/welcome?p=cdb2184a More to come As the community grows, we'll keep improving things. If you have ideas, drop them in General and tag one of the admins. Let's keep building. 🚀 M.

I'm wondering if anybody can help me here. I keep on getting this bottleneck when I'm working on a project in Codex. It keeps asking me for approvals, and even though I have given the agent full access, it's still asking me for approvals. It's frustrating because I'm not always in front of my computer, so when there's no approval given, the work stops. Does anybody know how to get around this issue? Is there another way of dealing with this so I'm not constantly having to be at my computer approving certain aspects of the coding?