Common Prompt Mistakes
Most weak answers come from an unclear prompt, not a weak AI. In this lesson you'll diagnose the mistake yourself, then build a stronger prompt and watch its quality rise.
This is a learn-by-doing demo. The ideas here work with any AI, no code or vendor-specific tricks.
Small changes, big difference
When a prompt is vague, the AI has to guess what you want, and guessing usually goes wrong. A few small fixes, like naming your audience or asking for a format, make the answer far more useful. Below you'll practise spotting the fix and building it yourself.
What's wrong with this prompt?
Read the weak prompt and pick its main problem. Get it right to reveal a stronger version and why it works.
Turn on the missing details
Pick a weak prompt, then switch on what it's missing. The prompt rebuilds live and the quality meter climbs as it gets clearer.
Good prompt engineering is mostly about being clear, specific, and giving the AI enough information to do the task.
Name the audience · add context · ask for the format you want · set limits like length and tone · and split a big request into smaller prompts.