A poster post drew 890 comments.

Its author argues repetitive AI designs reflect default prompts, not model limits.

He told ChatGPT to treat its first poster as a “what not to do,” producing a Bauhaus-style layout.

Naming the style lets him request it again.

How each outlet framed it
Hacker News
demonstrates ChatGPT reverts to identikit despite specific prompts; identifies repetition fatigue not quality as the core problem

Sources: Hacker News