What is a Prompt and Why Most People Write Them Wrong

Most people who start using AI tools treat them like a smarter version of a search engine. They type a few words, get something back, and either use it as-is or give up because it wasn't what they needed.

The problem isn't the AI. It's the assumption that a vague input can produce a specific output.

This chapter is about what a prompt actually is — and why the way most people write them guarantees mediocre results.

What Is A Prompt?

This distinction matters because most people treat a prompt as a question. In reality, a prompt is closer to a briefing document — the more complete and precise it is, the more useful the output will be.

A prompt can contain:

• A task or instruction ("write," "summarize," "compare," "explain");
• Background context the AI needs to understand the situation;
• Information about the intended audience or purpose;
• A required format or structure for the output;
• Constraints on tone, length, or content.

None of these elements are optional by nature — but each one you include makes the output more targeted and useful.

Why Most Prompts Fail

The most common prompting failure is not a technical mistake. It's an assumption problem: the user knows exactly what they need, but doesn't communicate it — expecting the AI to infer context it was never given.

The five most common mistakes:

• Too vague — "help me with this" gives the AI nothing to work with;
• Missing audience — the AI doesn't know if it's writing for a client, a manager, or a general reader;
• No format specified — the AI chooses a structure that may not match how the output will actually be used;
• No constraints — the AI writes 500 words when you needed 80;
• Single-attempt mindset — treating the first response as the final answer instead of a starting point.

The Shift In Mindset

The single most important reframe in this course:

A prompt is not a question. It is a specification.

When you ask a colleague to do something, you don't just say "write something for the client." You give them context — who the client is, what happened last, what outcome you're looking for, what tone to use.

AI works the same way. The time you invest in writing a more complete prompt is always paid back in a more useful response. And once you've written a good prompt for a recurring task, you can reuse it — which is exactly what Section 3 of this course is built around.