By now you have used AI to improve something you wrote. Today you use it for a different type of task: structuring or organising information you already have.
Today's task
Pick something you are currently trying to figure out, a decision, a research question, a planning problem. Give the AI your raw thoughts and ask it to structure them:
I am trying to decide [your decision or question]. Here are my raw thoughts so far:
[paste your notes, bullet points, or stream of consciousness. This can be messy.]
Please organise this into:
1. The key options or possibilities I am considering
2. The main trade-offs between them
3. What information I am missing that would help me decide
Do not make the decision for me. Just help me see the structure of what I already know.
This task works because it plays to the AI's strength, pattern recognition, and avoids its weakness, making judgements that require your specific context. The AI can see connections in your own notes that you might miss because you are too close to the problem.
AI excels at pattern recognition in material you provide. It is bad at making judgements that require context you did not give it. This distinction is the single most useful thing to understand about these tools. When a task plays to pattern recognition, finding connections in your notes, surfacing themes in raw data, AI is genuinely useful. When a task requires your specific judgement, AI is a draft generator at best. Most frustration with AI comes from using it for the wrong type of task.