Article
Full AI agents or fully offline.
Four days in Catalonia. No computer, no AI, almost no social media. I bought this notebook so that I could write down what I would think about, and what I would come across and learn on the trip.
The notebook also records the driver who took us to where we were staying — a man who has lived in Barcelona for decades, although he was born on the opposite side of the Iberian peninsula, close to the Portuguese border. During the ride he spoke, among other things, about the tension between the Catalan question and Spanish centralism. He was clearly on the Spanish side.
We spoke about other topics as well, and he recommended places to visit.
The next day we went into the city. And alongside the pleasant views, we came away with some very unpleasant experiences.
The notebook records, for instance:
“7 May, in the afternoon. A side street running directly off Barceloneta, ten metres from us, in broad daylight: someone violently snatched a valuable straight out of another person’s hands and ran.”
Or:
“9 May. For the second time in three days, again about ten metres from us, in broad daylight, a valuable snatched by force out of another person’s hands, followed by a run. We saw it from the taxi.”
I had never seen a theft like this in my life, and now twice in three days. It would deserve its own article. So would other notes that I am not writing about here.
I have plenty of material. On everything I lived through, I formed my own view, and I wrote it all down. Only now, after returning, am I moving it into digital form. This process makes a great deal of sense to me.
TWO EXTREMES, NO MIDDLE GROUND
For as long as I can remember, I have been surrounded by digital technology. Email, chat applications, project tools — all open, all the time, whether I was working, resting or travelling, the phone usually in my hand. The only stretch of my adult life when that was not true was during my master’s degree, when I read philosophy books in the library, and when something existed only in digital form, I printed it and went to the park to read it there.
Outside that period, the digital world has been everywhere around me. Over time my sense of being overwhelmed by it grew, and my attention became persistently fragmented. The information smog in my inbox and in the other tools I use to run projects reached the point where I could no longer find my own priorities through it.
That is how I arrived at the way I now work: I move in two extremes — either I solve everything digitally, mostly through AI agents, or I work entirely without a computer. The things that are genuinely important to me, I write down on paper. On my desk there are about twenty sticky notes with the tasks and projects waiting for me, and I record my thoughts and findings in a paper notebook.
Join the Library
Full access to my thoughts, personal stories, findings, and what I learn from the people I meet.
Join the Library — €29.99 per yearSummary
Related articles
Will AI take my job? A certified Google trainer told me in June 2024 that my profession would cease to exist. Twenty-two months later, my job title has not changed — but ninety percent of what I do during the day is different. I have delegated more of my thinking to AI agents than I thought possible. I am not afraid. This is why, and what it means for anyone asking the same question.
It’s a strange feeling. I haven’t fully processed it yet.
The more I talk with friends and acquaintances about AI, the more I notice something alarming.
More articles
I am building an AI system to predict the S&P 500. It runs on my own machine, uses free public data — yfinance, FRED, the Shiller dataset — and grades every forecast against reality. This series documents the build itself: the decisions, the methodology, the mistakes. What I will eventually share from the running system is a separate question, and an honest one.
Yesterday I could not tear myself away from the computer. When I lifted my head, it was half past eight in the evening. I had been sitting alone upstairs for about three hours.
One hour. Fifty-five minutes. That is how long it took to build what a Czech software firm had quoted at over €50,000. I built it with Claude Code. Not a prototype. Not a proof of concept. A working tool — the one the company actually needed. By the evening of the same day, it was running on staging. This is not about Claude Code. It is about what Claude Code exposes.
I have conducted roughly one hundred and fifty practical interviews over the past four years. Fifty for data specialist roles. A hundred for advertising and performance marketing specialists. Almost every one of them involved sitting down with a candidate over a practical task — something close to a real problem we actually need to solve at the company. Not theory. Not trivia. Applied problem-solving. Over time, I started noticing a pattern.
Before you can teach AI to understand anything, you need to see what it is hiding from you.
The moment other people needed access to it, the problem changed completely. It was no longer about whether the agent could learn. It was about who gets to teach it.
I wanted to build an agent that doesn't just assist. One that acts.
This is what I learned about local vs cloud AI, and why I switched to Claude Code.
What happened — and how can it be reversed?
