Description of prompt/desired imagery:
- “Playing frisbee with a cat in the park”
- I hope this creates a complete scene of a person and a cat playing frisbee in the park, as you might expect to see a person doing with a dog in a stereotypical scenario
Sketch:
Generation program:
- ChatGPT enhanced with DALL-E image generator
VERSION 1:
- Prompt: “Can you generate an image of a person playing frisbee with a cat in a park?”
- Result:
VERSION 2:
- Prompt: “This is almost exactly what I had in mind, maybe similar to that last one, but with the cat jumping for the frisbee”
- Result:
VERSION 3:
- Prompt: “You didn’t use the same style as the one I said I liked, this is more cartoon-like”
- Result:
RECAP:
- In conclusion, honestly AI is a good brainstorming tool rather than an actually accurate way to create something you desire. I would probably suggest using something like this to generate positional changes or different ‘views’ of standard objects or other things.
- Pros:
- It ‘can’ do the job
- really realistic in general
- great variations
- Cons:
- despite doing the job, it is not consistent
- cannot produce the same thing even when asked
- still very much can tell it is generated when next to real art