Someone saw that comment about making pixel art by just downscaling an AI image and took it literally, without any additional postprocessing, or even carefully picking a proper model/lora. They picked the most generic vanilla model
So few AI devs actually try to do a good job, but the few that do, from my experience seem to be exceptional, and or course produce greater quantity of content compared to traditional devs.Someone saw that comment about making pixel art by just downscaling an AI image and took it literally, without any additional postprocessing, or even carefully picking a proper model/lora. They picked the most generic vanilla model
What's honestly worse, is that they looked at the shit results they got, and went "eh, good enough, let's release it". In brand with most AI devs I guess....
That's exactly what they're doing, the all the dev had to do was make the images less blurry, simply changing how the pixelation process affected his images would've fix the problems.
