During lockdown i’ve been studying with the University of Karlsruhe, undertaking the ‘Introduction to AI and Neural Networks’ course. The module is part of the media art and philosophy department at HfG with the research group KIM: Critical Artificial Intelligence. Led by Professor Matteo Pasquinelli, the course was hosted online and welcomed students from across the globe studying media art, data science and many other humanities and science subjects.
The course was a theoretical and art historical introduction to artificial intelligence charting the the AI winters and summers of scientific discovery, investigation and academic and industrial funding. I was fascinated by the technical development timeline and trajectory of the AI from the Perceptron to the Deep learning revolution.
Practically we learned about the constituent parts of a neural network from training data, model and algorithm to try to establish a mental architecture of these statistical systems. In this crazy year of 2020 the explosion of facial recognition deployment in many sectors and increased dynamics of human quantification and the entanglement of big tech, corporate and academic sectors this course was politically pertinent for the current climate.
Crucially as a media art course, we studied the ethical implications of training datasets, the power dynamics, biases and issues of ownership and authorship inherent with these systems from how they are created, who by and the purposes they serve explicitly and privately.
My tutor Matteo Pasquinelli was an excellent and enigmatic teacher, the start of the course coincided with his releasing a new work ‘The Nooscope Manifested: AI as Instrument of Knowledge Extractivism’ a collaboration with Vladan Joler, an essay and cartographic map aimed to demystify the processes within AI from human and technical perspective. Viewable here
I recommend exploring the KIM research group website, so many interesting papers and research angles in this area.