BSC scientists collaborated in the creation of the digital artwork, which was on display at CCCB in the exhibition 'AI, Artificial Intelligence', co-produced by BSC
The S+T+ARTS prizes, awarded by the European Commission through the Ars Electronica platform, are considered the most important in the European digital arts sector.
Maria CHOIR is based on a model trained with the voice of singer Maria Arnal, which allows for a unique duet between visitor and AI that evolves with each interaction, becoming a chorus. Maria CHOIR is not only an artistic experiment, but a social experience about the creation of new musical tools with AI, about consent in collective datasets and the potential of synthetic voices to extend beyond the mere replication of human capabilities.
It also invites participants to contribute to a collective voice model, used for both artistic creation and research, highlighting the project's commitment to ethical engagement and transparency. During the five months that the exhibition has been active at the CCCB, more than 12,400 voices have been recorded, from among the 112,000 people who have visited the exhibition AI: Artificial Intelligence (CCCB 2024), co-produced by the BSC.
Maria CHOIR is the first of the artworks that the artist has developed in collaboration with scientists from the BSC. The researcher Joaquim Moré was responsible for training the model that allowed visitors to sing with Arnal's voice. The piece, developed and exhibited at the CCCB, was curated by Lluís Nacenta and directed by Arnal herself, with funding from the Fundación Española de Ciencia y Tecnología (FECYT) and the collaboration of BSC. The studio axolot.cat (Iván Paz and Lina Bautista) and the visual artist JP Bonino also participated.
Maria Arnal is currently collaborating with the BSC's Data Viz Group within the framework of the S+T+ARTS AIR grant, which proposes a reflection on creativity, ethics and musical creation with AI. The artist is working on the Impossible Larynx project, which integrates voice processing models and a 3D visualisation of the vocal tract to develop a new AI music tool to overcome the physical limits of the human voice, as well as medical and social applications. This new piece was unveiled at the latest edition of the Sónar+D festival.
The S+T+ARTS prizes, awarded by the European Commission through the Ars Electronica platform and considered the most important in the digital arts sector, will be presented during the new edition of the festival, which will take place from 4 to 8 September in Linz (Austria).