Nike Air AI: Fashion in the wrong time
Is named str4ngething, and his specialty is creating artistic pieces using artificial intelligence. According to this creator, he makes ‘Fashion and modern objects that appear at the wrong times’.
If we take a look at his Instagram profile, we can see that str4ngething has a collection of images that seem to be taken from a museum. However, in all of them we can see that the characters in these paintings wear Nike clothing. Sometimes those clothes don’t really exist, and have been created by parsing the text that str4ngething has entered into the prompt.
In the latest images, str4ngething has left the boxes aside and dedicated itself to generating images that could be futuristic products, but inspired by the Renaissance. The results of these products are really funny, since it is true that they are so well made that they cause a certain sensation of having seen them for sale before. Hence the famous ‘Mandela Effect’ that the author spoke of.
DreamBooth, is that you?
We know exactly what you are thinking. Can I make my own Balenciaga collection? Could I make my own concept of the iPhone 17 Pro Max? If yes… how could I do it?
Well, everything indicates that str4ngething has used the same tool as everyone else: stable diffusion. However, if you install it on your computer and start issuing commands, it is possible that you will never generate an image that comes even close to these results.
the secret is in DreamBooth. This is a program that was launched a few weeks ago and that allows train your own customized artificial intelligence. What DreamBooth does is use the base model of Stable Diffusion and do a new training with a series of images and keywords that we are going to give. After the learning process, the AI will be able to generate what we want. In recent days, Twitter has been filled with people who have used this application to have a Stable Diffusion model capable of generating images with their own face.
In this case, str4ngething has done much of the same thing. This artist may have selected a series of images of actual Nike products, along with some concepts distributed over the Internet. Then, by selecting the right words, the Stable Diffusion engine is able to create all these product images. Of course, we already guarantee that each image requires hours, because the world of AI is not as simple as writing a line of text and crossing your arms. Is required long time and hundreds of iterations to get a decent and quality result.
