Step 1 of 4Knowledge
A prompt is the instruction or question you give to an LLM. It is your input text -- everything you type into the chat window of ChatGPT, Claude, or any other LLM. Prompting is the art of phrasing these instructions so that you get the best possible response.
Why does this matter? Because the same question, phrased differently, can produce completely different results. A vague prompt leads to a vague answer. A precise prompt leads to a precise answer.
Imagine you are giving a task to a very smart but literal-minded intern. If you say "Do something with the data," they will do something -- maybe not what you wanted. If you say "Create a table of the sales figures from the last three months, sorted by region, with a summary at the end," you get exactly that.
Step 2 of 4Understand
There are three basic rules for good prompting:
- Be specific: Instead of "Explain AI to me," try "Explain what Artificial Intelligence is to a career starter in 5 sentences."
- Provide context: Tell the LLM who you are, who the text is for, and what situation you are in.
- Define the format: Tell the LLM whether you want a list, a paragraph, a table, or bullet points.
Here is a comparison:
Bad prompt: "Write something about dogs."
Good prompt: "Write a short paragraph (3-4 sentences) for a pet website that explains why dogs are good pets for families with children. Tone: friendly and informative."
The perfect prompt rarely succeeds on the first try. Treat prompting as an iterative process: send a prompt, evaluate the response, then refine your wording. With each round, you learn what works.
Step 3 of 4Apply
Take a task that you would normally give to an LLM. Before you send it, check: Have I formulated a specific task? Have I provided context? Have I defined the desired format?
Even with the best prompt, an LLM can provide incorrect or outdated information. Always verify important facts against a second source -- especially for medical, legal, or financial topics.