It’s great inspiration for surrealism, when you try getting before-and-after or side-by-side images, and it doesn’t understand separate frames are discrete. Like the sloppy accidental version of Gunnerkrigg Court’s Coyote prancing all over the page.
I asked ChatGPT “Draw a 3 panel comic that is a clever response to this comment” and it told DALL-E 3:
Photo comic strip with three panels: Panel 1: A cartoon character, a young woman with short brown hair and glasses, reading a comment on her computer screen that says, ‘These examples tell me that it doesn’t understand the narrative underlining a comic. I mean, why should it? AI gets a bit overrated sometimes.’. Panel 2: She ponders with a finger on her chin, looking thoughtful, and says, ‘Hmm… challenge accepted.’. Panel 3: She proudly showcases a well-drawn comic on her computer screen with a smirk, and says, ‘Never underestimate AI!’.
Yeah tbh DALL-E generates some nice comics… Most don’t have any sense but some have some kind of idea on it that you can kind of get…
Like I had this other one of a plane crash…
These examples tell me that it doesn’t understand the narrative underlining a comic. I mean, why should it? AI gets a bit overrated sometimes.
It’s great inspiration for surrealism, when you try getting before-and-after or side-by-side images, and it doesn’t understand separate frames are discrete. Like the sloppy accidental version of Gunnerkrigg Court’s Coyote prancing all over the page.
Ooh I get that reference!
I asked ChatGPT “Draw a 3 panel comic that is a clever response to this comment” and it told DALL-E 3:
Which then drew this:
Not quite there yet lol