Your goal today is to get a genuinely useful result from AI, something you would actually use, not just a novelty. The trick is to pick a task that is small, familiar, and has a clear success condition.
Do not start with: "Write my business plan," "Analyse this spreadsheet," or "Explain quantum computing." Those are too large. You will get a generic answer and feel disappointed.
Instead, pick something like:
- Rewrite an email you already wrote to make it clearer
- Find a better title for a presentation you already finished
- Shorten a paragraph you know is too long
- Brainstorm five alternatives to a decision you are making
The key word is "already." You are not asking AI to do something from scratch. You are asking it to improve something you already have. This is important because you already know what "good" looks like for this task.
Your first prompt
I'm going to paste something I wrote. I want you to help me improve it.
Here is the context: [one sentence explaining what this is for, who is reading it, why it matters]
Here is what I wrote: [paste your text here]
Please do three things:
1. Tell me what is working well in this text
2. Suggest three specific changes that would make it clearer or more effective
3. Write a revised version incorporating the best changes
This prompt works because it gives the AI context, a concrete task, and a structured output format. The "what is working" step is deliberate, it forces the AI to engage with your actual content before rewriting it.
Read the revised version carefully. Do not accept it blindly. Does it sound like you? Did it change something you meant to say? Take the suggestions you agree with, ignore the rest. You are the editor. The AI is a draft generator.
You should finish today with one piece of real work that is genuinely better than it was before you started. If the AI output was unusable, you learned something: your input was too vague, the task was too large, or you didn't give enough context. Tomorrow's session will fix that.