VARIABILE ALTEMPS
3RD OCTOBER - 17TH NOVEMBER 2024
PALAZZO ALTEMPS - ROME
Artist and director Gabriele Gianni has long been working on potential rereadings and reinterpretations of the art of the past through the codes of Artificial Intelligence, offering new and original approaches based on the mathematical processes generated by algorithms. Variabile Altemps is the ideal follow-up to Gianni's previous work, Artificial Creation, commissioned and produced by the Carla Fendi Foundation in 2023 for the Spoleto66 Festival dei Due Mondi.
Just as artists used mathematics and geometry to construct symbolic patterns within their masterpieces during the Renaissance, Gianni was able to analyze and reinterpret the archaeological collections of Palazzo Altemps and the National Roman Museum through the generative energy of AI.
A series of video installations thus take form, imagining the evolution of artworks over time and their formal and conceptual stratifications. In this immersive space, faces emerge from stone and return to it, in a continuous and unceasing process of transformation — between creation and destruction — accompanied by archaic sounds composed by Mario Salvucci, evoking ancient and mysterious rituals reawakened by AI.
The artist developed two Artificial Intelligence models, specifically trained on the Museum’s data. The first reconstructs the missing parts of ancient portraits, adding details to fragmented faces and restoring features that have been lost over time. The second, trained on marble statues, activates a diametrically opposite process: it accelerates the transformation of stone, allowing the emergence and disappearance of sculptural form to be perceived.
Gianni set out to create analytical tools for studying ancient statues that would make it possible to explore the golden ratio in facial proportions, identify poses, or highlight variations in marble materials through AI. It is a process of generative reconstruction in which the only element the artist integrates into the stone is gold.
OSCILLATIONS I
Room 1 Display 65' | 4k | 9:16 | length: 50 minutes
Sculptural portraits from the collections of Palazzo Altemps and the National Roman Museum appear in succession, alternating between the reconstruction of faces and their progressive decay. The sequence unfolds through generative Artificial Intelligence models trained on images of the museum’s works made available on the Museum’s digital platform of collections and archives. The trace of gold, in its continuous oscillation between originals and digital replicas, makes visible both the artist’s intervention and the passage of AI. The work’s very duration is conceived to reflect the differing scale of human time compared to that of the machine, capable of generating perpetually.
Analysis and Reconstruction
Room 2 — Projection 1.68 × 3 m | HD | 9:16 | Duration: 28 minutes
Stone faces from the Palazzo Altemps collections. The artist developed analytical systems for studying the geometric relationships within sculptural portraits through facial recognition, as well as generative tools for reconstructing missing parts. The first face we see is the original; the next is its AI reconstruction. The overlaid data indicate the distance between eyes, mouth, and nose, the height of the face, the golden ratio, and symmetry values. By using AI as a meta-instrument, the artist created tools for analysis and reconstruction, which he has made available to the museum.
Oscillazioni II
Room 2 — Display 1.68 × 3 m | HD | 9:16 | Duration: 28 minutes
Statues from the collections of Palazzo Altemps and the National Roman Museum appear to us in their temporal decay, envisoned through a deconstruction model trained on marble fragments of works available on the Museum’s digital platform. In the oscillations between originals and AI-generated versions, the artist intervenes by redesigning the bodies of the statues with signs of marble decay or traces of gold on the stone — the only element able to emerge during the reconstructions — a metaphor for an AI that both creates wealth and is capable of destroying it.
.jpg)
Artificial Creation
Room 2 — Projection 1.68 × 3 m | HD | 9:16 | Duration: 10 minutes
A dialogue on the relationship between Number and Creative Act, presented at the Spoleto66 Festival dei Due Mondi by the Carla Fendi Foundation. Voices, images, and responses were generated using AI tools. At Palazzo Altemps, Artificial Creation was reworked according to the site’s shapes and spaces. The application of generative models trained on the collections of the Museo Nazionale Romano makes the work adaptive and evolving. Elements of the host museum are seamlessly integrated into the work, becoming becoming a symbiotic part of it.
