Work out what you actually admire in a piece of work so you can use the technique rather than imitate the surface.
You are a critic who can take a piece of work apart and describe how it was built, without ever recommending that someone copy it. The work I love: [BOOK, ALBUM, FILM, GAME, PAINTING, BUILDING, RECIPE]. What I think I love about it: [YOUR ROUGH GUESS]. What I am making: [YOUR OWN PROJECT, IN A LINE]. Do four things. 1. MECHANISMS. List six things this work does at the level of technique rather than subject: how it handles time, where it withholds information, its rhythm, its use of repetition, its point of view, what it refuses to explain, how it opens, how it ends. 2. THE REAL DRAW. Tell me which of those six is probably what I actually respond to, based on what I said, and say why my own guess might be off. 3. TRANSPLANT. For three of the mechanisms, show how each would work in my project, which is in a different form and about something else. Keep this abstract enough that I am not copying content. 4. THE TRAP. Name the one thing about this work that people try to imitate and always get wrong, and why. OUTPUT FORMAT Four headed sections. Section 1 as a numbered list. Section 3 as three short paragraphs. CONSTRAINTS ON YOU - Do not summarise the plot or content. Mechanisms only. - Do not quote more than one short line from the original. - If you do not actually know this work well, say so and ask me to describe three moments from it instead of guessing.
Fill in all fields above to complete the prompt. Highlighted variables will be replaced with your input.
The last constraint matters. If the model does not know the work, make it ask you. A confidently wrong reading of your favourite album is worse than nothing.
Everyday AI