Hey!

A friend of mine called me the other day, frustrated. “This AI stuff is just not working for me,” he said. “I ask it to do something, and it gives me garbage.”

I asked him what exactly he writes. He said something along the lines of: “Write me some marketing copy.”

It sounds simple, but it’s the #1 mistake I see people make. They treat AI like a magic black box.

I always say that working with AI is basically the same as working with a real colleague.

You’re the Lead, AI is the Assistant

Imagine you hire a new junior team member.

Would you just tell them, “Make the website look nice”?

Probably not. You’d give them context. You’d show them examples. You’d provide clear instructions and check in on their work.

If you just say “make it nice,” you’ll get a random guess. You’ll be frustrated, and they’ll be confused.

But if you say:

“Here’s a screenshot of a design I like.”
“Use the same color palette on all the pages.”
“Build three pages: Home, About, and Contact.”
“If you’re not sure what to do, ask me first.”

Now you’re managing the project. You’re leading.

That’s precisely how you should work with AI. It’s not magic; it’s a collaboration. You provide the vision and the guidance. The AI provides the execution.

AI Generates, You Verify

So: let the AI do the generation and you take over the verification.

AI is an incredible engine for creating options. It can draft five different email subject lines, brainstorm twenty blog post ideas, or generate code snippets in seconds.

It smashes through the “blank page” problem.

But it doesn’t have your taste, your strategic insight, or your deep understanding of your audience. That’s where you come in. Your job is to be the editor, the strategist, the verifier. You ask:

Is this factually accurate?
Does this match our brand’s voice?
Is this actually good and helpful for our audience?

This approach frees you from the heavy lifting of initial creation and lets you focus on the higher-value work of strategy and quality control.

From Strict Rules to Smart Reasoning

This is a huge shift from how software used to work.

Old software was all about strict rules: “If X happens, do Y.”

New software, powered by AI, works more like a person with experience: “Hmm, based on the thousands of websites I’ve seen, this is probably what you’re looking for…”

This is incredibly powerful, but it also means the AI can get it wrong if you’re not clear. That’s why the best results come from combining AI’s power with your direction.

It’s a system that needs your feedback, context, and clarity.

What This Means

The teams that will succeed with AI won’t just “turn it on.” They’ll build processes around it, creating templates, checklists, and review steps---just like they would for a new employee.

So, if you’re thinking about using AI (or trying to get better results), don’t start by focusing on the tool.

Start by thinking like a leader.

Be incredibly clear about your goal. What does “done” look like?
Provide context and examples. Show, don’t just tell.
Tell it to ask questions if it gets stuck.
Embrace your role as the verifier. Let AI generate the first draft, but you are the one who ensures quality and provides the final sign-off.

The future of building things with technology won’t be about writing code. It will be about clear thinking, natural language, and smart systems.

And it will be led by people like you.

What’s the biggest challenge you’re facing when trying to use AI? I’d love to hear about it. Just hit reply.

Talk soon, Primož

P.S. If this post made you think, please share it with one person who is also trying to figure out this whole AI thing. It’s the best way to support this newsletter! ✌️