A prompt generator feeds text into image-generation models, and a taste model filters the results before $BOTTO holders vote weekly on which fragments they like best.² The winning image becomes a finished artwork, historically minted and auctioned on SuperRare, with the proceeds redistributed to the community.^1,2 Both the prompt generator and the taste model began from random inputs, and have been shaped over time by the accumulated weekly votes rather than by any human-authored artistic direction. Botto has produced more than eight million images in total, the great majority never made public because they were not chosen by the taste model.²

Botto later gained a "sovereign memory" built from its own exhibition history, press coverage, academic research and governance record, letting it maintain continuity of authorship as the underlying AI models it draws on change. This memory was first put to use when Botto began writing its own generative code in p5.js. It was later extended into a parallel "intent" process, in which Botto retrieves an idea from its memory and iterates toward it, alongside its original open-ended exploration of the image models' latent space.²

Botto's genesis work, Asymmetrical Liberation, realized with Klingemann's involvement, sold at auction for about 79.421 ETH, roughly $325,000, when the project launched.¹ In July 2023 it took part in “Future Frequencies: Explorations in Generative Art and Fashion”, an auction staged by Christie's and Gucci.³ For Seeds of Inspiration, it collaborated with the photographer Ryan Koopmans and co-creator Alice Wexell, who set previously unminted Botto image fragments into photographs of abandoned architectural interiors in Georgia and Poland.⁴