
This is a picture that a friend of mine took on a trip to England. This was probably made by OpenAI's latest model, because this is also one of many abominable Ghibli images that were probably part of some kind of meme. You can see that the text quality is infinitely better than before. It spells, displays and even puts it all into perspective correctly. However, it seems to only really be able to output a few different fonts, which you can even spot in the post that we're commenting on. The slop mutates ever closer to slipping past your defenses...

EDIT: Bonus picture, I can never see Ghibli anything anymore without this coming to mind