Visions of Fungicene
The temporal-spatial crisis that we are going through today confront us more frequently with the signs of the melting of time, landslides that occur not only on a material level but also on an ontological one. What chance do we have of glimpse the functioning of other worlds and recognize the existence of pluriverses? Starting from this same estrangement from contemporary viscosities, opens up the possibility of imagining worlds designed by other non-human agencies, even after our time.
Visions of the Fungicene is a speculative approach for the articulation of new space-time frameworks on future ecosystems, which digitally/immersively shows a possible scenario of planetary terraforming (in a post-anthropocene era) co-designed from biological principles contained in the morphological patterns of various species of fungi that, guided by artificial intelligence algorithms, generate ecosystems, planetary landscapes and neo-topographies: synthetic territories that would reconstitute the current scenarios of devastation in another world.
In the approach to a design process of this imaginary, a bridge of symbiotic methodologies between the intelligence of fungi and artificial intelligence is enabled, where from databases that contain photographs of three basidiomycete genera of fungi: Agaricus, Amanita and Suillus, a machine learning model is trained based on StyleGAN2 (an artificial intelligence program) that learns the patterns of the textures and morphologies of mushrooms to generate new series of images based on these patterns. Based on these results, a futuristic fictional narrative is reconstructed about a world without us, where non-human entities are the protagonists.
The installation is made up of three simultaneous elements: an audiovisual narrative, a 360 video as a virtual journey through this new territory, and a 3D printed sculptural object from this world.
Fungicene: VR360 Immersive Space
Installation in the gallery space
Fungal Terraforming Metaverse on Mozilla Hubs
Virtual environment experimentation carried out in Mozilla Hubs using Rhinoceros + Grashopper for 3D topographic reconstruction of the images generated by the model trained with machine learning.