Who works here
I am Dieter Szegedi.
My professional background is in computer science, software development, product development, and the design of complex technical and organizational systems.
Over many years, a way of working emerged whose distinctive character I did not recognize for a long time: separating observations from their interpretations as far as possible, holding different explanatory models alongside one another, and making decisions based on the best orientation currently available.
This way of working did not originate as a theory. Looking back, I recognize it instead as a pattern that runs through very different professional and personal situations.
Through intensive dialogue with language models, I was able to recognize and describe this pattern clearly for the first time. Since then, I have been trying to understand this way of working more precisely, apply it in different contexts, and make parts of it publicly traceable.
How I work
I am not primarily interested in what a system, an organization, or a situation “really is.”
To orient ourselves better, it often helps to begin by asking:
What can we observe? What different models explain these observations? What changes when new observations are added? And what decisions appear reasonable based on our current orientation?
For me, models are not truths that need to be defended. They are provisional ways of understanding a situation.
This allows me to leave contradictions and uncertainty unresolved for a while instead of replacing them as quickly as possible with a single unambiguous explanation.
Orientation is therefore not a state that one eventually reaches. It must remain capable of changing as new observations emerge.
Humans and AI
Much of the content on this website now emerges in dialogue with AI systems.
I do not use language models merely as tools for writing or research. In dialogue, they become an extended space of association for me; at the same time, they increasingly take on the role of personal assistants.
They help me organize observations, examine models, keep contradictions visible, and look at my own thinking from an additional perspective.
What interests me is not handing human judgment over to machines. On the contrary: I see one of the most interesting possibilities of AI in its ability to improve human orientation, self-reflection, and decision-making.
Responsibility for what I publish and what I do remains with me.
The fact that an AI assistant now also provides notifications about new content on this website is therefore not simply an additional technical feature. It is a small practical consequence of the way this website now comes into being.
This website
szegedi.info is my personal public workspace.
Developed works appear here alongside essays and texts in which orientation is still visibly evolving.
I deliberately keep these different forms alongside one another. Anyone who wants to know what I thought or considered plausible at a particular point in time should be able to follow that just as they can follow subsequent changes.
The website therefore documents not only the results of my work. Increasingly, it also documents how they emerge.
Contact and notifications
You can write to me directly at dieter@szegedi.info.
If you would simply like to follow what emerges here, you can receive email notifications about new publications.
The notifications are handled by my AI assistant.