The infinite loop

When AI stops inventing and starts connecting

We're used to thinking of generative artificial intelligence as a tool that creates images or videos from scratch. You enter a prompt, wait a few seconds, and the system produces something completely new.

The experiment described by DropFrame Instead, it proposes a different approach: using AI not to generate content, but to provide continuity to real, existing material.

The author begins with a series of black and white street photography photographs, taken personally. Not artificial images, but photos constructed through observation, composition, and searching for the right moment. These shots are then processed with the transition function of Higgsfield, without asking the algorithm to create new scenes.

The goal was to achieve a continuous visual loop: a sequence where each image blends into the next, eventually returning perfectly to the starting point. The interesting part is that the AI takes care of solving the spatial connections between very different photographs, simulating realistic camera movements and transitions.

The result does not showcase the AI's ability to “invent,” but rather highlights the photographer's work. The AI becomes a sort of invisible glue that unites images already rich in identity and artistic intention.

According to the author, this type of use could transform how photographers and creatives present their portfolios: no longer static collections of images, but dynamic and immersive visual experiences, where authorship remains completely human while technology amplifies the presentation of the work.

SOURCE (Davide Vasta): https://dropframe.substack.com/p/il-loop-infinito-quando-lai-non-crea?utm_source=post-email-title&publication_id=8386355&post_id=196757494&utm_campaign=email-post-title&isFreemail=true&r=8c7hng&triedRedirect=true&utm_medium=email