AI is radically transforming the production and perception of images, raising new questions about authorship, style, originality, and responsibility. Starting from an analysis of generative text-to-image and text-to-video models, the lecture unfolds along three conceptual axes:

1. The role of semantic attractors: words or concepts that guide visual generation by acting as gravitational poles of meaning;
2. The ethical and aesthetic biases implicit in datasets and algorithmic filters, inevitable insofar as they mirror our own prejudices;
3. The notion of distributed creativity, which deconstructs the idea of the author as an isolated individual and instead recognises the network of actors (both human and non-human) involved in the construction of the image.

The lecture weaves together philosophical, semiotic, and technical perspectives, offering a critical and practical reflection on how to use these tools consciously, without succumbing either to technological hype or to a priori rejection. Designed for scholars, artists, and cultural practitioners, it invites a shift in perspective: not only to ask what AI does, but what it tells us about how we construct meaning, aesthetics, and collective memory today.

The lecture will be held in Italian.

Digital preservation faces a fundamental paradox: as our capacity to record grows exponentially, meaningful engagement with vast collections diminishes. Cultural artefacts in institutions accumulate in databases in digital format, disconnected from human physical experience—mirroring our personal struggle with ever-growing phone galleries. This talk presents “compositional interfaces”—frameworks developed during Hiroaki Yamane residency on the Island of San Giorgio Maggiore that transform data abundance into creative opportunity.

Rather than static repositories, these interfaces treat digital archives as dynamic materials for artistic expression. Moving beyond categorical organisation, they enable remix through computational techniques. The presentation demonstrates prototypes and research including: semantic extraction across 20,000+ images from the Fototeca of Fondazione Giorgio Cini, creating navigable landscapes beyond tabular constraints; and reconstruction of Giacomo Casanova’s legacy in collaboration with artist Sara Francesca Tirelli, through Gaussian Splatting, depth estimation, and LLMs to transform archival fragments into new narratives.

Unlike traditional digital humanities approaches prioritising structured databases, compositional interfaces embrace ambiguity and unexpected associations through computational threads. This method bridges technical digitisation processes with creative practice, offering new pathways for cultural heritage engagement in the digital age.

The lecture will be held in English.

The essence of complexity lies in connectivity. Whether we examine the vast expanse of the World Wide Web, the intricacies of a cell, or the marvel of the human brain, their complexity invariably stems from intricate interconnected networks. Connectivity is foundational to our biological makeup, our social fabric, and our communicative prowess. In this talk, Albert-László Barabási will chart the 25-year odyssey of BarabásiLab in crafting the visual lexicon of networks and connectivity. Finally, he will focus on our recent monumental project for the 2025 Venice Architecture Biennale, which charts 50 years of architectural history.

The lecture will be a journey from the “interactome” — the subcellular networks that underscore our health — to the “connectome”—the neural network that delineates our consciousness. Albert-László Barabási will chart the transition from the nebulous ‘hairballs’ to the clear connections of 3D printed physical networks, from data vis to the dataism, and from research papers to art museums.

The lecture will be held in English.