How to use AI without handing over the decision
A practice for organizing options, criteria, risks and missing information without transferring responsibility for the choice to the model.
ReadTag archive
#responsibility
Publications tagged with responsibility.
Are we afraid of AI, or of ourselves?
A research-informed essay on how AI can amplify goals, become a moral alibi and obscure institutional responsibility - while real limits of human control still matter.
ReadHow to analyze a message without diagnosing a person
A practice for organizing a difficult message with AI without diagnosing the sender, mind-reading intentions or handing over responsibility for the reply.
ReadHow to separate facts from interpretations
A practice for separating visible facts, interpretations, assumptions, judgments and missing questions before a conclusion or decision forms.
ReadHow to use AI as a second reader
A practice for using AI as a second reader: checking clarity, structure, tone and reception risks without handing over authorship.
ReadA good summary is not the same as a good decision
AI can organize information well, but decisions require criteria, responsibility and an understanding of consequences.
ReadDo not diagnose people from emails
AI can help organize difficult communication, but it should not be used to diagnose people, personalities or intentions from correspondence.
Read