Who Chooses the Artwork? Curatorial Agency and Distributed Intent in Botto

AI authorship is more complex than the algorithm alone: Botto's artistic output results from a recursive loop involving image generation, community ranking, voting, and minting decisions—meaning no single entity (neither the AI nor the community) can claim sole authorship. This challenges the common framing of AI systems as autonomous artists.

Power in AI systems is distributed but unequal: While community members influence Botto's artistic direction through voting and governance, the underlying AI models determine which images become visible and selectable in the first place. This means the community's choices are constrained by what the system makes available to choose from.

Explainability for AI art requires transparency beyond model behavior: Understanding how AI art systems work requires examining not just how the algorithm generates images, but also the "curatorial procedures"—the ranking, filtering, and selection processes—that determine which outputs are presented as finished artworks to the public.

Governance structures matter for AI creativity: Botto demonstrates that decentralized decision-making in AI systems doesn't automatically distribute power equally; the technical architecture and curatorial rules shape whose preferences actually influence the final output, raising questions about fairness in collaborative AI systems.