In this session you will run the same climate question through nine different setups and see nine different answers come back. Same model, same question. The only thing that changes is the context behind each one: the persona it is given, the data it can see, and whether it reasons step by step before it answers. The gaps between those nine answers are the whole point. Once you can see them, you can control them.

The reason most AI output feels generic is not the tool. It is that the model is guessing at everything you did not tell it: who you are, who the answer is for, what counts as a good source. A climate analyst, a policy advisor, and an investigative journalist will frame the same 1.5°C question in completely different ways, and the model will too, depending on how you brief it. One of the nine boxes is fed fabricated climate data on purpose, so you can watch how fluently and confidently a wrong answer gets delivered when the source is bad. That box tends to be the one people remember.

This is the Context Chaos exercise from the Terra Studio curriculum, run live. You will work through it yourself, then Nilesh will reveal the hidden context behind each box and pull out the framework underneath: the few inputs that decide whether an AI answer is useful or noise.

What you'll do:

Take one climate question and send it through nine configured boxes during the session. Each box hides a different combination of persona, data quality, and reasoning mode. You compare the answers side by side, predict what changed, then see the actual setup revealed. By the end you can read an AI answer and infer what it was and was not given.

What you walk away with:

A repeatable way to set up any AI tool for climate work, plus the habit of checking context before you trust an output. Specifically:

• The context framework: the inputs that move an answer from generic to grounded
• A prompt structure you can reuse for analysis, briefings, and policy work
• A short checklist for catching confidently wrong answers before they reach your team
• Exclusive discount and access to Terra Studio

The format:

Ninety minutes, hands-on. You run the exercise alongside Nilesh rather than watching from the back. He walks through each reveal as it lands, with time to test your own prompts and questions throughout.

Go deeper:

This exercise comes straight from the Terra Studio curriculum, where context engineering is treated as the skill underneath everything else you do with AI. The related free course on becoming an AI-fluent professional is available at studio.terra.do.