And maybe prompt engineering isn’t the invention of the AI age at all. It may be a return to something Confucius already intuited: that the form of a question is not overlooked but carries within it the power to direct thought, constrain possibility, and prescribe a certain horizon of answers.

Confucius believed that language does not merely reflect society but manifests it. Words commit us to courses of action; misnaming distorts not only what we say but also what we do. This is why people in his time, and often in China still, are careful with speech in a way that contrasts with Western free expression.

All this raises a provocative question: when we spend hours refining prompts, correcting outputs, and coaxing ChatGPT to deliver our ideal responses, does that make us the teachers of value? Specifically our moral values?

about Confucius: he often adjusted his answers depending on who was asking