All through December, a social media consumer often called Stelfie the Time Traveller has been crafting a time-hopping travelogue utilizing generative AI. Due to Stable Diffusion and fine-tuning, an nameless artist has created a fictional photorealistic character that he can insert into fake historic images set in numerous eras, resembling historical Egypt or the time of the dinosaurs.
Secure Diffusion is a deep studying picture synthesis mannequin that enables folks to create fictional scenes utilizing textual content descriptions known as prompts. With an extra approach known as Dreambooth, folks can insert their very own topic or character into scenes generated by Secure Diffusion. It will also be used to insert real people into fictional conditions.
Thus far, “Stelfie” has taken historic selfies throughout the ice age (being chased by a woolly mammoth), in historical Egypt (throughout the building of the pyramids), in historical Greece (with the Trojan Horse), hanging out with Leonardo da Vinci (whereas creating the Mona Lisa), within the old West, whereas operating from a tyrannosaurus rex, and whereas crusing with Christopher Columbus.
The artist behind Stelfie writes social media posts (on Twitter, Instagram, and Reddit) in character as playful dispatches from a 41-year-old time traveler as he visits completely different places, for instance:
Log #3 – Columbus ship, 1492 DC. “Laaaaand” Now, I do not wanna brag however… After months of crusing Columbus was misplaced within the Oceans and properly for example I’ll have kinda instructed the right route…
The nameless artist (a self-identified “humorous previous man”) detailed some of the process he makes use of to create the pictures in a number of Reddit feedback: a mixture of Secure Diffusion 1.5, a customized AI mannequin for the panorama, and a customized AI mannequin skilled on the Stelfie face, which is seemingly a fictional person created utilizing Character Creator. He makes use of “a number of inpainting,” which implies inserting AI-generated imagery into the pictures to repair errors and sculpt the scene, and every picture takes three hours to create.
Some could discover Stelfie’s attempts to sell the images as NFTs unpalatable (the tech is controversial), however simulated time journey is arguably a novel leisure use of know-how that may in any other case create deepfakes and probably deceive others. Utilizing Dreambooth and Stable Diffusion, anybody can now insert a personality (or themselves) into implausible and fictional conditions, so we’ll doubtless see much more fictional leisure eventualities within the close to future.
Itemizing picture by StelfieTT
