Comparing Qwen3.8 Max and Fable 5 on UI (blog.kilo.ai)

I mean, isn't that fairly subjective in the same way human designers have been arguing about for decades though?

Without the prompts or knowing how specific they are I think it's rather disastrous how similar they all are. But that's probably why many can catch an AI generated/designed page in an heartbeat. If the prompts weren't very specific and you get similar results from different models... then I think we are looking at a rather bleak and boring web.

[flagged]