The Problem with the Blank Box
When you sit down to use an AI like ChatGPT or Claude, the empty text box can feel like a test. Most users try to write one long, perfect paragraph to explain their goal. They might ask for a weekly meal plan or a professional email. Because the AI does not know who you are or what you like, it often gives a generic, boring response that requires more editing than it is worth.
This struggle happens because you are trying to act like a computer programmer instead of a manager. You feel pressured to include every possible detail in one go. If you forget to mention your budget or your specific tone of voice, the AI simply guesses. These guesses are usually wrong, leading to a cycle of frustration and manual corrections.
Let the AI Lead the Discovery
To fix this, you need to change your role. Instead of trying to provide all the information at the start, tell the AI that its first job is to interview you. This is called discovery. By asking the AI to lead the conversation, you ensure that it gathers the specific context it needs before it starts writing a single word of the final project.
You can start with a simple sentence: I want to write a project proposal for a new community garden, but I want you to interview me first. Ask me ten questions one by one about my goals, the location, and the budget so you have all the facts. This forces the AI to identify its own knowledge gaps. It shifts the mental heavy lifting from you to the software.
Converting Raw Thoughts into Quality Output
This technique is especially powerful when you have a lot of messy ideas but no structure. You might have a rough voice memo transcript or a few scattered notes from a meeting. You can paste those raw notes into the chat and tell the AI to use them as a starting point for the interview. The AI will then ask questions to bridge the gap between your rough thoughts and a polished final product.
This process acts as a filter for your expertise. If you have been doing your job for many years, you have a lot of hidden knowledge that you might forget to include in a standard request. The interview helps pull those valuable details out of your head. It ensures the final result sounds like you and reflects your specific standards rather than a generic template.
Building Your Personal Source of Truth
Once the AI has asked its questions and you have provided the answers, you have created a foundation for the work. This set of answers becomes the source of truth for the entire conversation. The AI now understands your intent. It can now act as your researcher, writer, and strategist because it finally has the necessary data to be useful.
When you finish the interview, tell the AI to produce the final draft based on everything you just discussed. You will find the results are far more accurate than if you had tried to write the perfect request from scratch. By treating the AI as an inquisitive assistant rather than a magic wand, you save time and get results that actually work for your daily needs.
Key takeaways
- Start your session by asking the AI to interview you for a specific task instead of giving it a long list of instructions.
- Specify how many questions the AI should ask to ensure it gathers enough context before it begins the work.
- Paste raw notes or transcripts into the chat and ask the AI to help you organize them through a Q&A session.
- Use the AI as a virtual staff member that handles research and drafting while you provide the expert knowledge.
- Only move to the final writing phase once you are satisfied that the AI has a complete source of truth for your project.