Skip to main content

What social challenges do prompt engineers face when using text-to-image generation systems?

Sarah Bradley avatar
Written by Sarah Bradley
Updated over 2 weeks ago

Prompt engineers face the challenge of predicting how others might describe and react to images posted on the Web.

Since text-to-image systems are trained on images and text scraped from the web, users need to imagine the image not just as a description, but as if it already exists on the Web.

Simply describing an image in detail often isn’t enough; one must consider how others would describe and interpret it.

Did this answer your question?