Introduction
The prompt decides the output
The story: You hire a brilliant freelance designer. Whether you get exactly what you need, or something beautiful but useless, depends almost entirely on the brief you give them. The better the designer, the more the brief matters.
In AI/AWS terms: As foundation models get more capable, the prompt (the brief) decides how good and accurate their output is. This course covers how to write, refine, and optimize prompts, the zero-shot, few-shot, and chain-of-thought techniques, and the risks of prompt misuse.
For the exam: The prompt is the main lever for output quality.
Running scenario
The story: You work at a finance app company and ask the designer, "Make a market report for a new product." They come back with a glossy report about smart doorbells, in a format you can't use. They weren't wrong; you just never said you're a finance company, who the report is for, or what it should look like.
In AI/AWS terms: At AnyCompany, a fintech company, you need market analysis reports. The first prompt, "Generate a market analysis report for a new product category", produces a report about smart home security: the wrong industry and the wrong format. Adding context is the first and biggest improvement.
For the exam: Output in the wrong domain or format usually means the prompt lacks context.