The AI-Output Reflection Span

To what extent do you shape the outputs of your interactions with AI? Whether asking for advice, programming, writing, ideating, planning or decision-making, people increasingly outsource cognitive processes to machines. As a result, human input is often merely an intervention through which an AI system starts processing. This raises the question which underlying assumptions, ideas and beliefs are embedded in the outputs of AI systems and who reproduced them. Did they emerge from the worldview of the AI, or from ours, triggered through the prompt? From my perspective, reflection is the capability to distance oneself from and observe one's own thinking, in order to more consciously shape it, to notice underlying assumptions and to intervene in their reproduction if considered necessary. When we outsource cognitive processes to machines, reflection requires new dimensions that focus not just on the self, but on distributed cognitive systems. It becomes necessary to consider personal influences, the influences from the AI system and the interactions between the two dimensions. This simple methodology aims to do exactly that. It enables the identification of underlying assumptions in an output and the assessment who reproduced them and to what extent. It can be applied in workshop settings or individually as a structured process within everyday AI usage. It is also thinkable to integrate it as a functionality into user interfaces.

Aug 21, 2026