A Generative State of Mind

" frameborder="0" allow="autoplay" allowfullscreen>

1: Spacing in Time – 0916653+1005344+1007629
De-synchronization
Time Pollution
Cyberbells and Ghosts
Untitled Sample Diffraction in Time [Beat Box
(Diversion One) - Art of Noise - 1984


2: Clara
Alarm
Wave


3: Anonymous
Field Recording 13.10 – ClipGraveyard


4: 8
Messy Loops We Live in

Welcome to "A Generative State of Mind", a broadcast of the master experimental publishing, piet zwart institute and hosted by Louisa and Federico aka. Bowser and Funix.

What you are gonna watch is a series of AI-generated images through a GAN algorithm that analyses the sound of a certain audio track. Generative Adversarial Network: this Machine Learning technique uses two neural networks. One creates a new possible picture from Visual Noise, the other one attempts to discriminate the other neural network, in order to improve the output result. Does it create possible realities? Does it create probable realities? Does it generate microcosms? A few questions can come up. A multitude of answers could be given. We would like to question the objectiveness and truth in which AI is hyped. There is a chat: use it, for fucks sake.

In a system dominated by measures, demanding an ever increasing synchronisation between individuals and their environments, they propose "Spacing in Time" as a way to escape pre-determined variables. Through these audio experiments, they encounter a new moment of experiencing different time flows. The radio turns into a time's perception box where various tools for time measurement and orientation are being de-fragmented and re-booted. While creating audio environments with sound repetitions, reverbs, overlapping and echoes, the experiments lead to a sense of physicality which allows the full immersion into a broadcasted time travel. Shut down the clocks and de-synchronise yourself from the established recognition of time, while un-practicing your common perceptions. What follows will be experimental tracks by Martin Foucaut, Jacopo Lega, Nami Kim AND Kendal Beynon AND Clara Gradel AND Floor van Meeuwen. Camilo Garcia and Euna Lee.

The tracks have been composed using the open source software Sonic Pi. This radio stream is accompanied by visuals generated from sound using deepsing; thanks to Nikolaos Passalis and Stavros Doropoulos to let the repo be open source. This stream is hosted in the confederated project peertube, in Swiss Olympiad in Informatics's server. The title of this broadcast has been generated by https://www.name-generator.org.uk/rapper/