The art of taking good selfies seems to be about to reach a new level previously unknown thanks to the power of the Artificial Intelligence (AI) platform. stable diffusion.
And it is that in the most recent example of use of this system, someone had the brilliant and crazy idea of using the power of AI to take the most impressive series of self-portraits in the entire history of humanity, literally.
Since “the subject” took advantage of the potential of the system to create a series of selfies of himself in different periods of human history, both in the past, in the present and even in the future.
So now, thanks to this technology, we have given rise to the first time traveler powered by Artificial Intelligence.
The idea sounds like a bad joke perhaps, but in reality the results of this self-portrait session are admirable, as a clear demonstration of the power of this new technology.
Based on selfies, this subject traveled in time with Stable Diffusion
During the last days on Twitter, a spectacular Twitter account has gone viral, under the name of @StelfieTT, an account that is dedicated to showing the results of Artificial Intelligence applied to recreate selfies of the same subject at different points in the history of humanity.
The idea of just imagining it may sound silly or even ridiculous, but once we see the results of each photograph, the only thing left to do is stop and admire the potential of this platform. Here are some clear examples of this:
📳Log#1
I’m Stelfie, a 41 year old mf.
Thanks to modern sd technology I travel through time and take stelfies.
It’s messed up but I’ll make it back.
A parody of life and history through a crypto time machine adventure.
Next is my remote control…#timetravel #stablediffusion pic.twitter.com/4hJQ4DiqRA— Stelfie the Time Traveler (@StelfieTT) December 6, 2022
The AI Stable Diffusion in its version 1.5, as reported by colleagues from Tech E Blogwould have trained in the development of these self-portraits with the fictitious character of ‘Stelfie’, where the protagonist that we see in each photo is not real.
In itself, the image generated by the platform would have been one and then it would have been necessary to apply some enhancement touch-ups to increase its realism and credibility, so that each portrait would have taken about 3 hours to complete.
But what’s interesting here is that the Stable Diffusion model has the ability to create new images from scratch by using a simple text request, describing the elements to include or omit.
Best of all is that access to this AI it is relatively free, although it is very demanding. So it is recommended to run Stable Diffusion on a system with at least 10 GB of VRAM.