Abstract-Codex | index | works | bio |> memory | network | contact

AI driven generarive artworks.

M0T3TU5 is a project placed at the intersection of generative art, artificial intelligence and data visualization. The title of the work refers to motetus the Latin name for motet: the traditional form of composition typical of Western polyphony; this title, finding its root in mot the French term for word, implies how in this installation text plays a central role.

This work is based on a machine learning algorithm that in recent years has received great interest for surprising results: it is the so-called char-RNN; a recursive neural network that once trained with a certain text (or any sequence of characters) is able to generate pseudo-texts based on the probability and prediction that a letter follows another one thus forming sequences of words, phrases and sentences to which the reader tries to attribute a direction of meaning.

MOT3TU5 is based on texts generated by a neural networks trained with some of the fundamental books of Western scientific thought (in this version Dialogo sopra i massimi sistemi by Galileo and Opticks: or a Treatise of Reflections, Refractions, Inflections and Colurs of Light by Newton) made readable through the visualization generated by a custom made software. The contemporary presence of texts in different languages and a spatialized soundscape refers to the poly-linguistic and polyphonic nature of the medieval and early renaissance motet. This work uses pre-existing texts as a mould to create and shape new materials, expanding and taking to the extremes this concept we could use literature as an art material, without the intention to create coherent narratives but on the contrary to warn us against the trust that we instinctively put into our interpretation. In a age where storytelling, propaganda and fake
are dangerously intertwined ascertain the facts should replace the XX century faith in hermeneutic.

>>> Download_documentation


MOT3TU5 video documentation ( mix of the 4 channel audio visual installation).

Alessandro Capozzo, "MOT3TU5", 2019. Multi channel audio visual installation, Variable dimensions.


MOT3TU5 video still frames.

Alessandro Capozzo , "M0T3TU5_CLUSTER_I", 2019 . Dyptich, lambda prints mounted on aliminium. 50 x 50 cm.

Alessandro Capozzo , "M0T3TU5_CLUSTER_II", 2019. Polyptich, 4 lambda prints mounted on aliminium. 30 x 30 cm.

^ up

  << back ^page-up
  > works > M0T3TU5