PUBLIC_KEY_PEM before distributing. Submissions currently save as unencrypted JSON.
The same request, worded differently, produces meaningfully different results. This module makes that visible — so you can start doing it deliberately.
By the end of this module: You will have run the same request three different ways across two AI tools and seen what changes. Then you will have written your own variations and tested those too.
The goal isn't a perfect result. It's seeing the pattern.
You will begin to identify patterns and differences across the two tools.
The tool responds to exactly what you give it. It doesn't know what you meant — only what you said. That means two people asking for the same thing, with different words, get different results. Not random results. Predictably different results.
Casual language gets casual output. Precise language gets precise output. Specialist language gets specialist output. The tool matches your register.
"If you use a $20 word, you'll get $20 words back. If you use amateur language, you get the amateur result." This isn't about being fancy — it's about being precise. The same request at three different levels of specificity produces three different levels of output.
Here's the same request written three ways. Notice what changes — and what the tool infers about the context from your word choices alone.
Don't just read the outputs — compare them. Look at length, tone, vocabulary, and whether the tool added things you didn't ask for. The differences tell you what information the tool extracted from your word choices.
Part 1 uses the three provided variations above. Part 2 is yours to write. Both parts run on both AI tools you signed up for in Module 1.
You're building a baseline for comparison across modules. Stick with the same two tools you used in Module 1 for now.
Always start a clean chat for each variation — do not use an existing conversation. The tool carries everything in the current conversation as context. If you send Variation B in the same chat as Variation A, the tool's response is shaped by both. Fresh chat, every time.
Don't use an existing chat — start clean each time.
Then start a NEW conversation. Send Variation B. Copy that response. New conversation again for Variation C.
Fresh conversations for each. Six responses total across both tools.
You will perform this exercise twice. Each time, you choose a request from your own work, write three variations of it, and run all three on both tools. Two requests, three variations each, two tools — twelve responses total.
Anything practical: summarise a document, plan a schedule, explain a concept, draft a proposal. Pick things you'd actually use an AI tool for.
Ideas if you're stuck: summarise notes from a meeting, plan a week's schedule, explain a concept to a colleague, draft a client proposal, write a job description.
Casual, professional, specialist — or three different levels of specificity. The point is to deliberately change the wording and see what happens.
Twelve more responses. Paste everything into the form below.
Part 1: outputs from the three provided variations. Part 2: your prompts and their outputs. Fill everything in, then save.
Paste the responses you received from each tool for each variation. Six fields total. Select your tools from the dropdowns — keep them consistent.
Choose a request from your own work. Write three variations, then paste the response from each tool for each variation. Six fields total.
Choose a different request from your own work. Write three variations, then paste the response from each tool for each variation. Six fields total.
Your work is saved locally. Download the file and keep this file for offline reference.