Give AI permission to disagree with you. One of the biggest mistakes people make with AI is only using it to execute their existing ideas. You tell it your plan. It agrees. Then it helps you build it. But what if your original approach wasn’t the best one? Try adding this to your prompts: “Don’t automatically agree with my approach. If there’s a better way to achieve the goal, challenge me and explain why.” Now you’re not just asking AI to complete a task. You’re asking it to question your assumptions, identify blind spots and suggest approaches you might not have considered. You don’t have to take its advice. The value is having another perspective before you make the decision. Use AI as a thinking partner, not a yes man.

One of the easiest ways to get better results from AI is to stop asking it to execute immediately. Before building, writing or changing anything, ask it to create a detailed plan. Then question that plan. What could go wrong? What is it missing? Is there a better approach? What does the finished result actually need to achieve? Refine the plan until you’re happy with the direction. Then let AI execute. You’ll spend slightly longer before starting, but you’ll waste far less time fixing poor outputs afterwards. The better the plan, the better the result.

If AI keeps giving you generic rubbish, it’s because it’s guessing. Paste in three or four of your own emails first, then say “this is how I write, now try again.” Same tool, completely different result.

If you’ve just joined, here’s the two minute rundown. There are 21 modules in the classroom courses, guides, prompts and step-by-steps. All free, every one of them is something you can set up in your own business the same day you read it. No theory, just actionable steps. How to get value out of this instead of just collecting tabs: 1. Don’t start at module 1. Start with whatever is currently annoying you most. Quotes taking too long? Chasing payments? Drowning in emails? Go to that module first. Momentum beats order. 2. Work through it, don’t read through it. Have the guide open on one screen and your actual business on the other. Use your real quote, your real email, your real customer. Reading it start to finish and “coming back to do it properly later” is how it stays theory. 3. One module, one process, one week. You don’t need to automate your whole business by Friday. Fix one thing properly and it keeps paying you every week after. 4. Ask when you get stuck. That’s what the community is for. There’s no such thing as a stupid question here someone else has likley had the same one and hasn’t asked it yet.

Right, it's done. The classroom went from one course to nine, and all of it is free. Here's what's in there. Start Here is three short lessons and takes about ten minutes. Read those first. They help you work out which job in your business is actually worth automating, and it's usually not the one that annoys you most. It's the one that quietly costs you the most. Then there are 23 lessons across five courses. They go from what AI actually is, through building something that works, to keeping it working on day ninety when you have stopped watching it. The Systems Library is 15 builds you can copy. Missed call text back, same day quoting, invoice chasing, review requests, onboarding, and ten more. Each one shows you the cost model so you can decide if it's worth doing, the exact prompts, the ways it breaks, and what to buy instead if you'd rather not build it yourself. The Prompt Library is 1,000 prompts, 50 in each of 20 categories. They're sorted by the job you're doing rather than by topic, and every one came out of something that got built for real. Two things worth saying. It's free and it stays free. Nothing is locked and there's no upsell waiting at the end. And the numbers in there are honest, including the embarrassing ones. There's a cold email campaign in the course that sent 50,738 emails and got 23 replies back. That's 0.05 percent. It's in there because the stuff that failed teaches you more than the stuff that worked. If you do one thing today, open Start Here and read lesson one. The live sessions start on Thursday 3 September, 12:30 UK time. Bring the job that eats the most of your week and we'll build the fix in front of everyone. Until then, there's plenty in the classroom to be getting on with.