What this course is about
How to make AI check its own work: prompts that reduce made-up facts, check questions, cross-checking, catching errors in figures, references and rules, and spotting a confident tone where there is no real certainty.
AI writes smoothly and confidently, even when it is wrong. It can mix up a number, invent a document reference or describe a rule that changed long ago. You cannot get rid of this completely, but you can cut the number of errors sharply and learn to find them fast. This course shows you how to build checking into every task {AI_WITH}, from the prompt to the finished document.
What you will learn
- A prompt that reduces made-up facts
- Before and after: a meeting summary
- Self-checking an answer
- What self-checking catches
- Check questions
- Let the AI ask the questions
- Cross-checking
- Cross-checking: an example
- Errors in figures
- Figures: a formula instead of an answer
- Made-up references and rules
- References: how to check them
What's inside
- Four PDF versions of the course: for ChatGPT, Gemini, Claude and Microsoft Copilot — each explains where to find the features you need in that assistant.
- 13 ready-to-use prompts and a prompt library you can copy from.
- Step-by-step tasks, before-and-after examples and a one-page cheat sheet.
- Free updates when the AI assistants change.
Who it's for
Office workers, managers, teachers, freelancers and small-business owners who want practical results from AI without technical knowledge.
KHIZHNYAKOV