Something can seem obvious and still be wrong here. These four layers separate noticing, explaining, deciding and making.
“Something is here.”
Noticing what matters or does not fit, often before you can explain it. A figure looks wrong. A phrase lands badly. A brief no longer fits its purpose.
LimitBias, tiredness or overconfidence can mislead you. A possible explanation is not a conclusion.
“What might it be?”
Describe the detail. Consider possible explanations. Find evidence. Name whose view is missing.
LimitAI can suggest explanations and help test them. It should not change the question without telling you.
“What should we do?”
Weigh the evidence, what is at stake and what matters. Decide with authority. Take responsibility for the people affected.
LimitA named person or responsible body owns the decision. A model or unchecked hunch cannot own it.
“Make it so.”
Draw, write, code or render the work. People increasingly share these tasks with AI tools.
LimitManual skill matters. It is not the whole of human value.
Keep in mind
What we hold