Test-Driving AI Tools on a Real Task
Sveip for å vise menyen
A common mistake is writing a vague prompt and blaming the AI for a weak result. The fix is usually in the prompt, not the model.
Take a vague version of a prompt:
Rewrite this document in American English.
This produces a result, but often loses technical terms and breaks the original formatting.
Now compare it to a more specific version:
Rewrite this document in American English. Keep all technical identifiers unchanged. Do not translate function or variable names. Preserve the original markdown structure exactly.
The second version tells the AI exactly what to protect and what to change. The difference between the two results usually comes down to how much context and constraint you provide.
Task
Write a simple prompt for your own task, then rewrite it with added context and constraints. Run both in an AI chat of your choice and compare the two results.
Takk for tilbakemeldingene dine!
Spør AI
Spør AI
Spør om hva du vil, eller prøv ett av de foreslåtte spørsmålene for å starte chatten vår
Test-Driving AI Tools on a Real Task
A common mistake is writing a vague prompt and blaming the AI for a weak result. The fix is usually in the prompt, not the model.
Take a vague version of a prompt:
Rewrite this document in American English.
This produces a result, but often loses technical terms and breaks the original formatting.
Now compare it to a more specific version:
Rewrite this document in American English. Keep all technical identifiers unchanged. Do not translate function or variable names. Preserve the original markdown structure exactly.
The second version tells the AI exactly what to protect and what to change. The difference between the two results usually comes down to how much context and constraint you provide.
Task
Write a simple prompt for your own task, then rewrite it with added context and constraints. Run both in an AI chat of your choice and compare the two results.
Takk for tilbakemeldingene dine!
