Why it’s interesting
Infinite Are Better Than One treats the Mona Lisa as the center of an expanding online galaxy. Nicolás Ruarte, working as NoMa, trained twelve LoRA models based on Stable Diffusion to bring Argentine painting, Indigenous patterns, comics, street ornament, and other visual languages into conversation with the familiar portrait. The project is more than one novelty image: individual works can be opened to inspect the prompt, random seed, and combination of model weights behind them, making the collection useful to browse and compare.
How to try it
Open the official website and choose All Giocondas. Pick ten portraits at random, then open their detail views and compare the disclosed prompts, seeds, and model combinations. Look for what remains recognizably Mona Lisa and what changes when a different Argentine artistic tradition takes the lead. The gallery is free to browse in a web browser and does not require a local image-generation setup. The first page is visually dense, so sampling a small group works better than attempting to scan the full collection in one visit.
Requirements and sources
The project is credited to Nicolás Ruarte / NoMa. Its official site documents the gallery and the twelve Stable Diffusion LoRA models used to recombine the portrait, while the related 2025 academic page supports the project year. Because the available sources establish 2025 but not a precise first publication day, the catalog stores the beginning of that year as a sorting placeholder rather than presenting it as a known launch date. The site remains browsable. These images are contemporary generative reinterpretations and should not be confused with Leonardo da Vinci’s original painting or with historical Argentine artworks.