A prompt is the message, question, instruction, or data you give to an AI system so it knows what to do. In tools like chatbots, image generators, and coding assistants, the prompt is the starting signal that guides the model toward an output. Learning to write clear prompts matters because AI is becoming part of schoolwork, science, programming, and data analysis.
A better prompt can save time, reduce confusion, and produce a more useful response.
Key Facts
- Prompt + model = output, where the prompt gives the task and context.
- A clearer prompt usually produces a more accurate and useful response.
- Good prompts often include task, context, constraints, format, and examples.
- AI models predict likely next words or symbols based on patterns learned from training data.
- In statistics terms, a model estimates P(output | prompt), meaning the probability of an output given the prompt.
- Prompt quality can be tested by comparing outputs for accuracy, clarity, completeness, and bias.
Vocabulary
- Prompt
- A prompt is the input text, question, instruction, or data given to an AI system to guide its response.
- AI Model
- An AI model is a computer program trained on data to recognize patterns and generate predictions or responses.
- Output
- An output is the answer, image, code, summary, or other result produced by an AI system.
- Context
- Context is background information in a prompt that helps the AI understand the situation, audience, or goal.
- Constraint
- A constraint is a rule in a prompt, such as length, tone, format, or required information.
Common Mistakes to Avoid
- Writing a prompt that is too vague: this is wrong because the AI has to guess the goal, audience, format, and level of detail.
- Forgetting to include constraints: this is wrong because the output may be too long, too short, too advanced, or in the wrong format.
- Assuming the first answer is always correct: this is wrong because AI models can make errors, so facts and calculations should be checked.
- Asking for hidden or unsupported certainty: this is wrong because AI responses are based on patterns and probabilities, not guaranteed truth.
Practice Questions
- 1 A student tests 20 prompts for a science summary tool. 14 prompts produce accurate summaries. What percent of the prompts were accurate?
- 2 An AI response is limited to 150 words. It produces 6 paragraphs with 30 words each. Does it meet the constraint, and by how many words is it over or under?
- 3 Compare these two prompts: 'Explain gravity' and 'Explain gravity in 5 sentences for a 9th grader using one everyday example.' Which prompt is likely to give a better student answer, and why?