DALL-E, OpenAI's system for generating images from prompts, is now open to whoever wants to sign up.
Here is the announcement.
I did a few experiments to come up with something appropriate for this group. My favorite is
this one
; the prompt was “a beautiful comic book superheroine with a determined expression flying over manhattan, digital art”One problem with it is that it seems to mangle faces. This might be a consequence of them trying to prevent deepfakes that look like a particular famous person.
Here's one more.
