I recently came across an Instagram post showing a different way of prompting ChatGPT for image creation.

Instead of writing a long, detailed prompt, it suggested using short command-style prompts, such as:

"/ugc testimonial: creates an authentic UGC-style photo of someone holding and reacting to your poster"

I was curious whether such a short instruction would actually make a difference, so I decided to test it using one of Shering’s workshop posters.

The Experiment

I provided the My First Shopping Adventure poster and used the /ugc testimonial instruction.

Prompt used: "/ugc testimonial: creates an authentic ugc-style photo of a lady holding and reacting to your poster "

The result was interesting.

Rather than simply redesigning the poster, ChatGPT created a completely different context around it: a natural-looking image of a woman holding and reacting to the poster, similar to user-generated content commonly seen on social media.

Original → Prompt → Generated result

What I Learnt

The experiment reminded me that AI prompting does not always have to mean writing longer and more detailed instructions.

A short phrase that clearly communicates the type of output, context or visual treatment can sometimes steer the image in a very different direction.

There is an important distinction, though. Commands such as /ugc testimonial are not programming codes or universal ChatGPT commands. They are better understood as shorthand prompting techniques. The AI interprets the words and the intended visual concept.

Why This Is Useful

For Shering, this opens up another way to experiment with existing content. A workshop poster does not have to remain just a poster. With different prompting approaches, the same asset could potentially be explored as a social-media scene, promotional visual, lifestyle setting or other creative treatment.

This was only one test, so I will be experimenting with more command-style prompts to see which ones genuinely improve the result and which ones are simply catchy prompt labels.

That is part of the fun of experimenting with AI: test it, compare it, and decide what is actually useful.