Tell HN: Un-Cherrypicked Stable Diffusion Outputs Are Horribly Wrong

Despite what the hype says, I never bought SD and DALL-E2's claims. We're not closer to generating art, unless art means creating images made up of shapes that don't make sense at all (women with 3 hands, pregnant woman with weirdly inflated belly, etc.). See this for some examples (and possibly nightmares): https://lexica.art/?q=women