>>5225216With images and text descriptions, you may know about
artflow.ai beloved on /tg/ for portrait pictures, GPT-3 and all those antics but I saw that the image version DALL-E 2 recently got some more billions of parameters to their models and have made apparently significant improvements.
https://openai.com/dall-e-2/Already with DALL-E 2 today you can type in a paragraph of text, and machine generate any completely original art aesthetic from the description alone at your command, like a reality-devouring Infernal Desire Machine lol. Maybe someday you can just type in "Charles Dickens David Copperfield, Virginia Woolf The Voyage Out and Gustave Flaubert Salammbo but as a Battle Royale set in WH40k universe with Shakespearean monologues and the voice actors from Dark Souls, also extra bosoms and guns and make it anime please" and an AI will just machine generate this as a qst or first person shooter or isometric CRPG or whatever format you want instantaneously.
It is not available for general release yet, because of "safety limitations" (they are scared of 4chan conspiracy meme power kek, maybe also using it for Ukraine lol). But this technology exists and it will probably unmake reality. I mean photoshop and twitter alone has already unmade reality, this will completely unravel and end it beyond human comprehension.
