I rarely had time to explain myself. You meet people, you work, you move on — and there’s never room for the whole context of why you do what you do. This site is a bit of the opposite: my attempt to start honestly for once, to be brave and take my own thing seriously instead of looking for shortcuts.

I’m 29, thirty in half a year. During the day I deliver for Flaschenpost; the rest of the time I build on projects like this one. This is an attempt to tell a decade in short — the detour that’s slowly turning out to be a throughline.

A YouTube kid

I started making videos at twelve. My brother and I shot short films on a Samsung S1, I cut gameplays, and I understood early that the internet holds all the knowledge you need. Editing, shooting, putting stories together — I just pulled it all in myself. It turned into a kind of obsession: to learn every skill I could find. At some point came the realisation that what you can do, you can also turn into money. For me a little earlier than for most.

Far too early in the industry

I needed money, so I got in first as an extra and then landed on a daily soap, where I played a small role. At some point they noticed I could also shoot and edit and that I understood YouTube — and suddenly I was in the engine room: first in one of the early YouTuber networks, where I learned how marketing and sponsorship deals work, then in format development at a production company. That’s where I first understood this is real work — with concepts, deadlines and the job of selling something to a broadcaster. That company taught me an enormous amount.

Then my grandfather died, and that was a rupture. Suddenly the question was there of what I was actually doing with my life. I went to South Korea for two months to visit my Korean grandfather — and came back with the feeling that I had to do something else.

Directing, debt, an award

Years as a freelance editor followed. I co-edited an independent documentary — the biggest editing project, the one where I really learned to cut. I produced a brand’s first TV spots, really cheap and really good: for the first time researching an ad script, thinking about what symbols stand for, building everything around the brand’s colours. It worked, I got to make more spots — and even to work in Cape Town for them. That’s where I learned what it means to direct on a bigger set.

With a small team I then made my first genuinely good short film and won an award with it; people I knew from the industry helped me. At the same time I was in debt, because I’d never learned how to handle money. I started a VFX degree and dropped out after three months. After that I waited tables in a tapas bar for a while and just worked.

The detour through computer science

In my free time I started rebuilding small games in Unity — and for the first time actually wrote functions and classes instead of just fiddling with website builders. Through C# I slid into Python, ground through bootcamps, solved puzzles. It was just fun.

The key moment was a VFX trade fair in Stuttgart, right in the middle of the Marvel boom. What fascinated me most wasn’t the finished effects, but how cheaply and quickly you could suddenly try things out with pre-visualisations — and a talk about how a big sports brand uses Unity to design shoes. I asked someone which tool to learn. The answer: something like Unity.

At first I thought I’d study game programming. Then I happened to meet a professor who said: if you only want to build games, come to us — but if you want to understand how the language underneath works, study computer science. That was exactly what interested me. So I moved to Aachen. The first semester was hard, my maths background not exactly glowing. But I sat in the library, gave it a shot, and it worked.

By the way, computer science isn’t what most people think. It’s not about coding all day. It’s the study of how systems work and how you build, automate and prove things. That became my most important skill later on. For my bachelor’s thesis I built my own browser for VR in Unity and tested it in a study with 21 participants against the standard browser of the time: for finding information again across several tabs, it was noticeably faster.

Four days

While I was writing the thesis, Midjourney came out. Four days after I handed it in, the first version of ChatGPT arrived. That moment changed my life. I’d already worked with neural networks and reinforcement learning during my studies — I roughly knew where this was heading. But suddenly holding it in my hands was something else. I locked myself away and learned one last, hard subject in a week with ChatGPT, and passed. Back then it was all still buggy and far from today — but the progress since then alone shows where the journey is going.

Towards the end of my studies I opened a small coworking space in Aachen with friends — an empty shop unit, a coffee machine, tables, WiFi. That’s where I first experimented with local marketing and noticed how well you can pull people into a real shop with Instagram stories. Then a lot collapsed at once, including a relationship that had held through my whole degree. I’d saved a bit of money. I went to Thailand.

Thailand

I’d had a history with Thailand for a long time. I’ve done martial arts since I was three, got my black belt in taekwondo at 14 and never really stopped. At 19 I was there for the first time, a month in a Thai boxing camp outside Chiang Mai. This time I stayed longer, trained Muay Thai and fought my first fight — and won on points. Shortly after I had a heavy injury, the worst I’ve ever had. It humbled me in the truest sense.

Alongside that I worked on Mindset — it started as a small app for friends to share their real progress, later a brand around martial arts, the thing I love. I started posting daily reels for the fighting community, and at some point it went through the roof: first a hundred followers, then a thousand, then thousands overnight, at the peak 20 million clicks a day. I wrote e-books, built whole brand worlds with AI, ordered products from all over the world. A lot half-worked and frustrated me just as often — the AI jumped ahead, but the thing I’d built often just didn’t run. I burned a lot of money and countless tokens.

Along the way I learned a ton I’d never had on my radar: international payment rails, because a large part of my audience came from India and couldn’t pay with normal cards. LLCs, verifications, permits. And something that still occupies me today: that sometimes you’re stopped from buying something simply because a third party stands in between and says no. That’s where I started to really understand crypto.

But the most important lesson was a different one: not everyone understands what you see. And that’s okay. It’s not a short-term decision — it’s a project I’ll pursue my whole life.

What AI removes

At some point something became clear to me. The problem is rarely that people don’t understand your vision or your work. The problem is that doing it — programming, sorting, structuring, writing, editing — is tedious, slow, often boring work, and few are willing to put in that time. For a website it’s the code. For a film it’s the edit. For everything it’s that grindy part.

That’s exactly what falls away now. What remains are ideas, context and the question of what you actually want to achieve. AI builds nothing on its own. It needs input, a purpose, an urgency — just like you. It only mirrors what you put into it yourself.

For someone like me, whose head is constantly full of ideas and who everyone always told to focus on one thing and finish it, that’s quite a liberation. The grindy busywork I used to have to cobble together alone, I can now share. Intelligence is no longer the excuse. It’s becoming more and more a question of “with whom” rather than “how”.

Now

Today I deliver for Flaschenpost. It’s the exact opposite of two years in a dark room in front of the computer, and it does me good: eight hours of something physical every day, being among people again, collecting stories. I like the job. And on the side I’m building exactly these projects — because I want to help more people again, with the skills I have.

Honestly, a big part of this decade was dark too. A lot broke, my ego most of all. But I have the feeling that with these new tools a new decade is starting, one in which I can finally realise the dreams that broke along the way.

What I really want isn’t complicated at all: to be in the sun more, do nice things with my family, sit in the dark less. At some point it struck me that in the end everyone just wants to eat and sleep in peace, without worrying. As humans we’re still very much driven by fear. If I have a mission, it’s to take a little of that fear away from people — and to show them how to build their own thing today. For everyone.

On directing, and AI as a camera

Between 17 and 20 I applied to thirteen film schools as a director and got thirteen rejections. Almost all of them said: most people make their first film at 31 or 32, wait, live first. I was impatient and didn’t get it. Today I do. I paused acting and just lived for ten years — because you probably have to live before you can tell stories.

The love of directing stayed with me, and the love of acting is slowly coming back. It had suffered badly, because I thought I’d be replaced by AI and was worth nothing anymore; I was also surrounded by people who had an interest in amplifying that feeling. I had to work on that self-worth. By now I see it differently: AI is basically just another camera, another brush. For a while I’d used it as an excuse to do nothing. Now, for the first time, I feel I can tell what I want — without putting on an act, without telling anyone what they want to hear.

Christopher Nolan once said the task is to tell complex subjects simply. That’s exactly what I try to do — in films, in code, and really everywhere.

In closing

This whole project here is also a kind of self-sorting. Maybe it’s almost a diary entry. But they’re things I’d share with anyone who wants to know them. So now they’re here.

If you’ve read this far: thank you. And if you’ve got a cool project — write to me.