My first computer was a budget supermarket PC. A beige box that sounded on startup like it was about to take off. Twenty years later I build companies in which software agents handle tasks on their own. In between lies a story that says more about the present than about nostalgia.
I am not writing this down because everything was better in the old days. Quite the opposite. I am writing it down because I have rarely lived through a time when the leap was as big as it is now, and because I believe the experience from back then is becoming useful again.
The budget PC and the first modem
The appeal was never the machine itself, but what it made possible. With the modem the internet came into the house, with an agonizingly slow dial-up and the sound a whole generation still has in its ears. Suddenly you could talk to the world from a child's bedroom. That experience, that a tool on your desk multiplies your own possibilities, never let go of me again.
I started building things with it early. First small websites, then first businesses, then companies. The common thread was always this: new technology where no one takes it seriously yet, try it before everyone else does.
What has not changed in twenty years
For all the change, one thing has stayed the same: the winner is not whoever has the most expensive technology, but whoever understands fastest what it really changes. The budget PC did not win because it was good, it was not, but because it reached millions of households and got people to try things out. Access beats perfection. That lesson still holds today.
The other constant is the attitude. I grew up with a handshake mentality, where your word counts and you simply get things done. Why that stays important to me precisely in the tech business, I wrote down here.
The leap to autonomous agents
Today I sit again in front of a technology that feels like the first modem: AI agents that do not just answer, but take on tasks independently. They research, write, plan, execute. What used to take a team whole afternoons now runs partly overnight. How big this shift is shows in the fact that companies like Anthropic now treat agentic systems as a software category of their own, no longer a mere chat feature.
For me this closes a circle. The budget PC showed me how a tool makes the individual stronger. Agents do the same, only more radically. What this autonomous work looks like in concrete terms inside a company, I describe in The Autonomous Organization.
What this means for you
You need no computer science degree and no starting capital to be part of it. You need the same curiosity as back then in front of the beige box. Start small, automate one annoying task, watch what happens, and build on it. Anyone who waits until everything is polished misses exactly the moment when the lead is created. How to get in practically, I described step by step in this post.
Frequently Asked Questions
What are autonomous AI agents in simple terms?
They are programs that do not just answer questions but independently handle multi-step tasks: researching, writing text, planning and executing processes. So they work more like a digital employee than like a pure reference tool.
Do I need technical background knowledge to work with AI agents?
No. You can get started without programming skills. More important than prior knowledge is the willingness to start small, automate a concrete task and learn from it. That is exactly how I got started twenty years ago with my first computer, too.
Why do you compare AI agents to the old budget PC?
Because both show the same pattern: a technology becomes broadly accessible and empowers the individual enormously. It is not the most expensive technology that wins, but the one many people try early. Whoever seizes that moment gains a lead.
Warm regards,
Dennis Weidner
