AI can now easily help generate the first draft of most communications, which means the challenge now lies in exercising judgment about the quality of that output. To collaborate effectively with AI, you have to articulate things that used to go unstated. Below is a four-step process to build this skill. It applies to any task where AI is involved and your judgment matters.
Establish an initial point of view. Before opening any AI tool, ask: What specific question are you answering? Who is the audience? What will they do with the output? And what would make the result useful versus merely competent? Develop your own hypothesis that you can later compare against the AI’s output.
Collaborate with AI across multiple modes. Most people interact with AI only to generate output. But you should also be using AI to critique (pressure-test what was produced), compare (surface the tradeoffs between alternatives), simulate (test the outputs against what you know about real stakeholders), and challenge (test what the output can’t know).
Analyze the differences between your initial view and the output. Not every difference will matter; sometimes AI will have a better take than you, sometimes the reverse. The point is to diagnose which differences reflect AI’s limitations and which reflect your own.
Deliver the output with an explanation of how you and AI arrived at it. This “reasoning trail” should capture two things. First, what AI initially produced, the starting point. Second, what you changed and why, to make your judgment visible.
