Disintegrator 3. Deep Learning as Parasite (w/ Jon McCormack)
Nov 22, 2023
43:04
Jon McCormack has been investigating the relationships between machine intelligences and creativity for decades. In Episode 3, Jon joins Marek and Roberto to speak about the social and cultural implications of AI -- beginning with the parasitism that deep learning methodologies practice upon human culture and the downstream effects on how we think, learn, and act.
We had the privilege of meeting Jon at an event thrown by SensiLab, an incredible research facility founded by Jon within Monash University, at their Creative AI summit in Prato this past summer. For anyone interested in the creative dimensions of AI, we highly recommend exploring the link above.
A few notes from the conversation:
This episode continues the through-line of skepticism toward the recent hype around major commercial investments in generative deep learning -- enumerating upon their bottlenecks, biases, and social and cultural effects. Jon’s critique here is strong and pithy, as are his gestures toward alternatives.
We had the privilege of meeting Jon at an event thrown by SensiLab, an incredible research facility founded by Jon within Monash University, at their Creative AI summit in Prato this past summer. For anyone interested in the creative dimensions of AI, we highly recommend exploring the link above.
A few notes from the conversation:
- At the beginning of the episode Jon references his absolutely beautiful work Holon, which is worth virtually exploring on his site. For anyone with a cursory or recent understanding of what it means to make art with AI, a deep dive into Jon's work is recommended -- it truly hits so many dimensions of this topic in stirring and striking ways and has been incredibly influential on many disciplines. It will definitely shake your conceptions of "AI + art" from that of idle generativity on commercial platforms.
- Jon references Anthony O’Hear’s “Art and Technology” as a foundational work establishing the parasitic tendencies of AI back in the 90s.
- There are some quite evocative images of the “alien-looking” structures developed through generative technologies as mobilized for spacecraft here. Jon references objects like this as examples of expanded notions of creativity.
- We strongly recommend Art in the Age of Machine Learning by Sofian Audry (a future guest on the podcast and a member of the Choreomata team), as it engages with Jon’s work.
This episode continues the through-line of skepticism toward the recent hype around major commercial investments in generative deep learning -- enumerating upon their bottlenecks, biases, and social and cultural effects. Jon’s critique here is strong and pithy, as are his gestures toward alternatives.