Talking to AI

In this part, you begin with prompt engineering in its narrow sense and gradually expand it. You learn that a prompt is not merely a command. It is a designed environment containing context, intellectual operations, dialogue structures, prior history, and quality controls.

Central transition : From better instructions to deliberately designed conversations

Use the same baseline whenever possible. By keeping the task stable, you make the effect of each intervention visible. The point is not that every reformulated prompt is universally superior. The point is to show that each design choice produces a different response because it defines a different task.

Baseline prompt

Write a paper on the economic value of beauty in tourism.

Part I · Chapter 1

1. Context I - Situating the task

Major question

How does giving the AI a specific context change what it considers relevant?

Description

The word " tourism" can refer to destinations, hotels, landscapes, architecture, attractions, marketing, mobility, or visitor experience. Without situational context, the model chooses the domain boundaries itself. A short contextual frame makes those boundaries explicit and changes the examples, mechanisms, and vocabulary that become salient.

You are working in the context of Alpine tourism destinations facing increasing international competition. Destinations invest substantially in landscape preservation, architecture, public spaces, hotel design, and visual destination branding. Write a paper on the economic value of beauty in tourism.

Expected impact

By situating the task, you change the response regime: the model stops producing generic, nowhere-anchored text and works within a field of localised constraints and stakes.

The instruction tells AI what to do. Situational context tells AI where the task belongs.