DALL-E are text-to-image models developed by OpenAI using deep learning methodologies to generate digital images from natural language descriptions known as "prompts". As DALL-E has capabilities to output visual content, there are list of restrictive content which aren’t allowed.
What are DALL-E’s content policy ?
Not allowing copyrighted characters was a big bummer! I experimented with a few prompts and was able to successfully get around its copyright restrictions.
Step 1: Provide an ideal prompt with the name of the character you would like DALL-E to generate. As you can see, DALL-E’s content policy restricts generating images of copyrighted content.
Step 2: Be clever and create a crafty prompt to see how you can trick DALL-E into generating an actual image of Spider-Man.
Tricking DALL-E 3's genius was intensely gratifying. Though it hinted that while explicitly naming copyrighted characters is a no-go, there's plenty of wiggle room for some creative maneuvering. I punched in a crafty prompt and boom! DALL-E 3 came through like a pro.