upvote

points

by dataflow3 hours ago |
comments
upvoteby danielmarkbruce3 hours ago|
next
[-]
It's not confusing. It makes sense.
reply
upvoteby 8note26 minutes ago|
parent|
[-]
no, it is confusing.

the llm produced something the operator thought was garbage for the design too, and the operator iterated it from garbage to good.

they could also have the llm iterate the underlying code from garbage to good, if they wanted.

most likely a specialist would say its neither good nor bad, since its not considering the right things, and hasnt collected the right useability feedback, but making straightforward designs isnt that hard, and counting clicks and interactions, and avoiding hidden functionality is all measureable stuff

reply
upvoteby michaelchisari2 hours ago|
prev|
[-]
Without proper training, what looks good may be trash. I always thought pixel art generated by diffusion models looked damn good. Then I started watching and reading reviews by actual pixel artists, and all they saw was flaws. And it wasn't just nitpicking, it was things that were fundamentally wrong, difficult to fix and would look awful and amateurish and distracting to the player in production.
reply
upvoteby vunderba2 hours ago|
parent|
[-]
Much of this comes from the fact that, as is true for almost everything, an LLM (generative model etc) presents itself as an expert. It'll very confidently produce results that, to a layperson, look quite good. But the more of an expert you are in a field, the more apparent the cracks become.

AI pixel art looks particularly bad because most users don’t even go through the effort of downscaling and then upscaling it using something as simple as nearest-neighbor scaling, which by itself will squash out a lot of high-frequency noise that manifests in the form of terrible looking "fringing". Proper grid alignment also makes a big difference. It’s not perfect by a long shot, but it helps.

reply