Marcel Lagerwerf

A reorganisation cut my role. I was unemployed, and I hadn't chosen it. A job hides its risk. That day it stopped hiding. Freelancing doesn't remove risk. It makes it visible, so you can decide how you will carry it. I trade risk for freedom. That's a choice, not a promise of an easy road. AI is changing what IT freelancers sell. But one question comes first: what do you actually want your business to give you? That's what I help IT freelancers with. Find out what you want, grounded in psychology. Then shape your business around it. The ideal life differs per person. My DMs are open.

After the reorganisation I didn't look for another job. I went independent straight away. Not because it was safe. Because I wanted more freedom in my choices. So a question for you: why did you start freelancing? Is your work today still the answer to that?

AI didn't take your work. It took the part you loved. You used to build. Now you review what a model wrote. If the fun went out of your work when AI came in, you're not imagining it. Staying relevant matters. So does liking your work. Don't give up one without noticing.

Run more than one business and you notice something odd. The one getting less of your time can even grow faster. Attention is not the same as progress. Work in sprints per business, not on everything at once.

"Lifestyle business" is often meant as an insult. As if a business that pays well and leaves you time for lunch with the people you love is the lesser goal. Scale is a choice. So is freedom. I don't want to grow my businesses in a bigger team, I want them to grow my freedom in life to choose how to spend my time.

AI takes over more and more of the work. Fine. But which skills you keep doing yourself should be a choice. The ones you stop using don't send a notice. They just fade. Which ones are you handing off on purpose, and which ones by default?

Want to own your future as an IT freelancer? Start with your ideal life. Not your business. What matters to you? What should your week look like? What work do you want to do? Who do you want to work with? Then shape your business around the answers. Do the work exceptionally well, long enough to build a reputation. After that, the world is basically yours. I coach IT freelancers through this. DM me 'Flamingo'.

Worst news of my career: a reorganisation cut my role. I hadn't chosen it. Best thing that happened to my career: that same reorganisation. Freelance since 2021. Now I decide, without asking a manager, to make time for: my kids my wife travel sport A job hides its risk. I trade risk for freedom. I coach IT freelancers to shape their business around their own ideal life, grounded in psychology.

Extended per quarter. One client. One agency in between. That's not a business you steer. That's waiting for someone else's decision. I'm not saying go find new clients tomorrow. I'm asking one thing: who decides your next assignment? If the answer isn't you, DM me.

AI makes the work easier and faster. More and more companies are using it. Code gets built faster. With more risk. And faster rework. So the question for a tech freelancer isn't whether you can build it. It's whether you can see the risk before the rework does. Can you?

Agent-written tests are so cheap that many people skip reviewing them. But a test nobody reviewed mostly checks what the agent already believed. Who checks the checker? Cheap to write doesn't mean cheap to trust.

The security of a salaried job felt real. Until a reorganization cut my role and I was out of work. Not my choice. A job hides its risk. A business shows it. I trade risk for freedom because I'd rather see the risk I'm taking. Which risk in your work can't you see right now?

Al agents shorten that chain. A small business run by agents is an asset that buys back time sooner, if you test it like an asset and not like a toy.

The real test for Al-native: who checks the agent's output, and what happens when it's wrong? Most companies can't answer that. The ones worth copying can.

If Al writes the code, interviews should test judgment: which questions you ask, which risks you spot, how you verify. That's a QA mindset, and it's hard to fake.

A job feels safe because the risk is hidden. A business feels scary because the risk is visible. I'd rather see the risk I'm taking.

Reviewing Al code line by line doesn't scale. Reviewing by risk does: where can this change hurt us, and what evidence do I need before I trust it?

Fun is the hardest acceptance criterion there is. "It runs" is easy to check. "People want to play it again" still needs a human who knows what to ask.

Most Al agents ship at level 1: they tell you something broke. The real work is designing and testing them until they reach level 4: the cause, the options, and the one they'd pick.

A post is going around: put your VPS behind a private network, your site behind a tunnel, and close the firewall to everything else. Good advice. If you have a VPS. I looked at my own setup and found three different situations. Shared hosting. A tunnel isn't even possible there. The real risk is an upload or a plugin. A managed platform. The provider already runs the firewall. Copying the advice adds nothing. One machine at home. That's the only place the advice really applies. So I'm having it checked first, before I change anything. Same advice. Three different answers. Security doesn't start with a tool. It starts with a question: what am I actually exposing, and where does the risk sit?