From Atari to AI Agents — 40 Years of Tech, a Reboot, and Why Now.

An Atari that could do nothing. A six-metre dish at Mitsubishi. Eleven years at TIMOCOM. SEO before SEO was a word. A dog named Coffee. And somewhere in between: the realisation that the machine I talked to as a child can finally, really answer.

The first computer. And the first disillusionment.

At twelve I got my first real home computer. An Atari 600XL. I remember switching it on, the flicker of the TV picture, the blinking cursor. And the thought a child naturally has in that moment: I’ll ask it a question.

I typed something in. Hit Enter. And got: SYNTAX ERROR.

That was the sobering realisation. The thing can do nothing. Not on its own. I had to teach it everything first. Back then I thought that was a flaw. Today I know — it was the beginning.

Tubes, satellite dishes and a journeyman

In the early 90s I became a radio and television technician. And I got lucky. While my colleagues out on call lugged televisions through Düsseldorf stairwells, I got to stay in the workshop 90% of the time. Set on the bench, back panel off, signal in, measure. Only for important clients, or when it came to satellite and decoder gear, did I get sent out.

The biggest thing I helped build: a four- to six-metre parabolic dish at Mitsubishi in Ratingen. Set up entirely on our own with one journeyman. Full band, full room, reception at its finest. I was as proud as could be.

TIMOCOM — the best time of my life

In 1998 I joined TIMOCOM as one of the first employees. First man in support. Then network engineering. Later development. In the end marketing. I did almost everything you could do at TIMOCOM. The core the website, the intranet and their ongoing development always stayed mine.

We were the startup everyone smiled at. Then we were the market leader.

“The best time of my life — and it was because of the colleagues.”

In 2009 I sold off a larger project and left. The corridors grew too long. The spirit was gone. Sometimes you don’t notice it in the numbers, but in the fact that you walk across the same car park differently in the morning.

SEO, before it was called SEO

During that time I taught Jon from MMOGA SEO — at a time when SEO wasn’t even a word yet. We pieced it together ourselves: understand crawlers, structure content, earn backlinks, don’t buy them.

MMOGA was later sold for a sum in the hundreds of millions and pushed aside its German rival Randyrun. Back then I had the choice: Randyrun or MMOGA. Or really: Germany or China. I took the German company. China would have been big, too big for the step at the time.

Canary Islands, Hilden, Coffee

In 2010 I moved to the Canary Islands. Back in 2012. I actually wanted to go further — Turkey, with a girlfriend. It didn’t work out.

Instead: a fresh start in Hilden. With my dog Coffee, who had been with me since I went self-employed. You don’t often hear about dogs in an IT biography. But anyone who works alone knows what a creature like that next to the desk means.

The quiet years

Until 2020, everything was easy. Hardly any outside contacts, but a great chat household with people who supported me and my work on the Farmerama fan site farmeramania.de for fifteen years.

Fifteen years. That’s easy to say.

2025 — the year three pillars gave way at once

In 2025 three good sources of income gave way. Almost in parallel. I was facing the end.

What followed was self-destruction. Nearly a whole year of 12 to 20 hours a day. Sometimes more. AI agents. I’ve known AI since 2019 — I had taken my own LLM all the way to a finished product. The first revenue came in.

And then came ChatGPT.

The decision

By the end of 2025 it was clear: in this new market I can’t get a foothold anymore without capital. For two, three years you might still earn something, until the revenue lands almost entirely with the big AI providers. Every system you build today gets devoured by the next wave.

So I wound down the company as of 31 December 2025.

“On the first day of 2026 I felt better. Out of the deep depression, into the energy.”

And then the setbacks

Honestly: a lot of setbacks came after that. Right now I’m not particularly good in a team. Communicating with people is hard for me. I can barely concentrate on a single thing. My whole body feels tense.

What works: I work on several projects in parallel and watch which systems really hold up. That’s not a problem — it’s actually productive.

A client job is doable, as long as the need for communication stays limited. Task in, result out. I always deliver 100%.

Why AI fascinates me

Back to the Atari. Back to the child who typed a question and got SYNTAX ERROR.

Forty years later I can ask the same machine a question, and it answers. Not perfectly. Not always true. But it answers. The twelve-year-old boy’s disillusionment has turned — late, but rightly — into wonder.

That’s why AI. Not because of hype, not because of money. Because a circle is closing that began forty years ago.

And because the machines are finally doing what I, as a child, thought they were built for:
Not to receive commands. But to listen.

Today

Sometimes I think: what if the Atari had already answered as an AI back then?

My life would have taken a different course.

Today I understand the systems. Tomorrow I’ll be talking with them.