TensorArt bragged how their FluxAI tool can generate comics... The results attached to the post speak for themselves. It's just 8 months old, so not an early model.
Transcript
AI-generated comic, mostly black and white with a few shades of grey.
Big panel at top: Fashion store with T-shirts and an undistinct item on display. Two very similar women walking into a wall next to its entrance, as well as a girl with impossible legs walking either left or right. Store sign: "Size leye light Is got neaıl boutique"
Panel 2: Woman in a shirt, cross-eyed: "Decitions, erctiʋns..."
Panel 3: Same girl, looking as if around a corner, with a disfigured disembodied hand holding a bag-like item: "We.let tame it caƨt harrng sishil.. ଚaraa I tine ջur ડeas tmall shopping ppree..
Panel 4: Same girl, looking at one or more large items held in her 4- and 11-fingered hands. Concerningly blank stare, circles around eyes, eyeballs lined with red: "ooh! the bres or ኗ5бo season.
Panel 5: Different girl with no arms and disfigured legs next to clothes racks. Confused look, indistinct symbols above her head: "What!' "thought I has cute shop, top.?"
Wide panel 6 at bottom: Both characters from above looking at a group of schoolgirls in front of badly drawn lockers; one has question marks above her head. The girl from panels 2-4: "Will a hebes. Red!" in a two-tailed speech bubble.
Indistinct signature and initial in the bottom right corner.
The prompt (which I now found and pasted in the body text) includes the entire script so any coherency in the story is not by the image generator. Yes, the script is so stupid it was probably by an LLM.