Public debate of artificial intelligence today oscillates between the emergence of superintelligence in which mathematics solve all problems and the apocalyptic scenario where the machine runs amok, take control from humans, and in the worst case, destroy them. This dichotomy propagates itself into the relationship between machine learning and art, limiting the understanding of algorithmic and human entanglement. Libidinal Geometry examines an alternative mode of “being-with” with algorithms as material processes through the practice of subverting machine learning, hacking 3D printers, modifying the grammar of their language, and performing algorithmic/rule-based gestures. Going beyond machine learning as a mere statistical inference tool, the enquiry looks into a particular configuration, what Hayles call a cognitive assemblage, in which algorithms act as cognitive agents entangled with human cognisers.
Constantly shifting and translating the artwork between mathematical, physical, and mental worlds, the research foregrounds errors in search, training, and coding as an indispensable tool in cognition processes. Through those translations, the nature of the artwork is examined away from the aesthetic towards the found, technical, and 3D printed object in the reconfiguration from unity to multiplicity. It seeks to answer the question of what happens to the artwork if cognition and responsibility are distributed between human and algorithmic agents and what kind of art emerges from this entangled field? Methodologically, the enquiry uses a post-modern (Deleuze: 1990, Deleuze and Guattari: 1987, Lyotard: 2004) and new-materialist (Barad: 2007, Golding: 2010) approach that resides outside of dialectic division between subject and object, the self and the other and enables thinking in distributed environments, quantum entanglements, and creative forces and intensities.
In detail, the research brings together the concepts of non-integer (multi-)dimensionality (Mandelbrot: 1982, Penrose: 2004) and the libdinal (Lyotard 2004) to advance the framework of libidinal geometry and thus enabling a form of algorithmic expression and self-actualisation (Fazi: 2018, Heidegger: 1977, Simondon: 2017) that is not miming human expression, but is the result of emergent processes (Prigogine and Stengers: 2017, Stengers: 2010) and artificial cognition (Downing: 2015, Hayles: 2017, Penrose: 2005). In this framework the libidinal operates in the form of artificial desire that is embodied via modifying and hacking the code and the 3D printer in the translation and materialisation process from geometrical models into printed sculptures. Using this methodology, the thesis foregrounds artists like Marcel Duchamp, Allan McCollum, Sol LeWitt, Pierre Huyghe, Wayne McGregor, Felix Gonzalez-Torres, and Sougwen Chung, whose works map out the conceptual and algorithmic field through appropriation, rule-base and machine-like actions, as well as human and non-human collaborations as the grounding of libidinal geometry.
Keywords: found(ing) object, Duchamp, errors, distributed cognition, ana-materialism, 3D printing, digital sculpture, libidinal geometry, algorithmic art