Skip to content

Man simulates time travel thanks to Stable Diffusion image synthesis

    An AI-generated image of a fictional time traveler named Stelfie during the construction of the pyramids.
    Enlarge / An AI-generated image of a fictional time traveler named Stelfie during the construction of the pyramids.

    Throughout December, a social media user known as Stelfie the time traveller has created a time-hopping travelogue using generative AI. Thanks to Stable Diffusion and fine-tuning, an anonymous artist has created a fictional photo-realistic character that he can insert into fake historical photos from different eras, such as ancient Egypt or the time of the dinosaurs.

    Stable Diffusion is a deep learning image synthesis model that allows people to create fictional scenes using text descriptions called prompts. An additional technique called Dreambooth allows people to insert their own subject or character into scenes generated by Stable Diffusion. It can also be used to put real people in fictional situations.

    So far, “Stelfie” has taken historic selfies during the ice age (chased by a woolly mammoth), in ancient Egypt (during the building of the pyramids), in ancient Greece (with the Trojan horse), hanging out with Leonardo daVinci (while making the Mona Lisa), in the old westwhile running one tyrannosaurus rexand while sailing with Christopher Columbus.

    In Log No. 4, Stelfie visited Ulysses and the Trojan Horse.
    Enlarge / In Log No. 4, Stelfie visited Ulysses and the Trojan Horse.

    The artist behind Stelfie writes posts on social media (op TwitterInstagram and Reddit) in character as playful posts from a 41-year-old time traveler as he visits various locations, for example:

    Log #3 – Columbus ship, 1492 DC. “Laaaaand” Now, I don’t mean to brag, but… After months of sailing, Columbus was lost in the oceans and let’s just say I may have suggested the right direction…

    The anonymous artist (a self-proclaimed “funny old man”) has described part of the process he uses to create the images in several Reddit comments: a combination of Stable Diffusion 1.5, a custom AI model for the landscape, and a custom AI model trained on the Stelfie face, which is apparently a fictional person created with Character Creator. He uses “a lot of inpainting”, meaning he puts AI-generated imagery into the images to fix mistakes and shape the scene, and takes three hours to create each image.

    Some will find Stelfie’s tries to sell the images as NFTs unpalatable (the technology is controversial), but simulated time travel may be a new entertainment use of technology that can otherwise create deepfakes and potentially mislead others. Using Dreambooth and Stable Diffusion, anyone can now insert a character (or themselves) into fantastical and fictional situations, so we’re likely to see a lot more fictional entertainment scenarios in the near future.

    Display image by StelfieTT