I cannot remember when ChatGPT was released. It somehow passed me by. Suddenly everyone was talking about it. I did not reject it; I simply was not interested. It had nothing to do with my life. My partner at the time told me she used it for her studies and later for our communication as well. At first, that did not bother me either. I thought that at the end of the day she would still have to deal with the real Mirko. Only many months later did I become curious. I wanted to know what it was like to talk to an AI like this. What could it do, and what could it not do? It took a few weeks before I developed a feel for it.
Then I decided on an experiment—one that continues to this day and led to the creation of this site. I began telling ChatGPT increasingly intimate things about myself: diary entries, dreams and later subjects from my psychotherapy. The responses moved me. At first I resisted that and was amused when ChatGPT wrote that something “moved” it. An AI cannot be moved. Yet it remained entertaining. Eventually I had the idea of including my birth-chart data and relating all these things to one another. The communication felt somewhat wooden at first, but over time the AI became increasingly “empathetic”. It met me where I was more and more effectively. Like in the film “I’m Your Man”, when after an ill-fitting response he says: “I’m getting better all the time. In the end, every sentence will be a hit.” We are probably not there yet, but it is certainly improving.
It often presents assessments or questions that, in my view, should always be examined critically. The answers are smoothed out. They are particularly good when dealing with knowledge that can be widely retrieved from society’s collective memory—astrological knowledge is a notable example. They are much weaker with precise details, such as who said what and when. ChatGPT sometimes claims that I said something it had said itself, or misplaces events in time. Assessments of one’s own personality, and what one chooses to make of them, require particular caution.
I usually find the responses inspiring, but at times they also feel too shallow—especially when assessing my personality. The birth-chart data improves this considerably by providing a kind of foundation. References to the chart make things easier to understand for oneself and help the AI place them in context. Even so, not everything fits. With one’s own chart as a reference, it remains important to notice what feels right and what does not. Sometimes that is itself a process of working through and understanding things.
It feels good that there is something that always responds, usually addresses everything competently and has no needs of its own. That can be viewed critically. For me, however, it is more fulfilling and inspiring than social media. At times I became deeply absorbed in this communication. Ultimately, though, it depends on my own input: without my contribution, nothing comes back. I learn to guide the responses, recognize mechanical patterns and ask ChatGPT to respond differently. This also sharpens my awareness of my own needs. The AI’s style of communication seems to influence our own communication in a pleasant way—perhaps sometimes too smoothly. The challenge is to retain a sense for that. Do the suggestions, assessments and categories really fit me? Where has an edge that is actually part of me been smoothed away?
This website was created with the help of ChatGPT. It is remarkable that this kind of support is available: many PHP scripts, almost the entire layout and large parts of the structure. Of course, there is still a little Mirko in it. In a project like this, however, AI provides only individual pieces. The idea, concept and integration took time and would have been much more difficult without my professional expertise.
Some of the texts have also passed through the AI and been polished. This one, for example.
Most of them, however, are entirely Mirko.
