OpenAI is making a huge mistake by deprecating DALL-E-3

There’s a certain dalle-e-3 style that I can’t name but think 2.0 will struggle with still. I don’t know what to label way it rendered a billion different entities at once all tiny but still mostly a coherent image? You guys know what I’m thinking of? I wouldn’t know where to look for those images, maybe someone has a few laying around… I would like to try