Slide the images <– –> left / right, or let it auto-play. There’s over 70 of them!
My friend, Edgar Stephen Curo, turned me on to a new app called Wombo Dream. You enter some text, and it feeds it through an artificial intelligence engine to generate abstract expressionistic art, in a variety of flavors – etchings, steampunk, baroque, fantasy art, and so forth. Whereas the app’s suggestions are on the order of “rainbow dolphin,” my natural inclination as a former QA engineer and present-day troublemaker is to feed it things like “The Incredibly Strange Creatures Who Stopped Living and Became Mixed-Up Zombies.”
You can see the results – if you dare! – and the prompts I used for each one at the bottom.
It’s from a Canadian firm, and I don’t know what they used to train their AI, but I am impressed that, even with wildly differing styles, the results do have a whiff of the spirit of their source material. Indeed, some of them, such as “Wrestling Women vs. The Aztec Mummy” suggest a romanticized, or perhaps just bigger budget, version. There’s three versions of “Fleshpot on 42nd St.” and I don’t know which one is most disturbing. After a while, I started to branch out into other art films and Giallo titles, famed for both their luridness and length.
The app has a “Generate again” button, and I presume it learns a little more every time you push it. I tended to like its first drafts best, when it made the most mistakes. For artists, mistakes are often hidden intentions. Teach that to the AI.