interdependence
patreon
Interdependence 2: AI and Music with François Pachet & Benoit Carré
🕑 Added 2020-05-12 15:59:09 +0000 UTCWe are joined by research scientist and Director of Spotify's CTRL lab François Pachet and his collaborator, the composer Benoit Carré to discuss how the latest in AI research offers a new paradigm for both composition and the concept of authorship and ownership.
This will be one of many times we pursue conversations with researchers at the cutting edge of Artificial Intelligence and music, not only because it offers a glimpse into the future of music, but also the greater economy. Music is often the first to feel the tremors for greater economic earthquakes, so it feels like an important conversation to jump into!
We can't play their music during the podcast, however we recommend that you check out these works in combination with listening to the conversation!
Daddy's Car: https://www.youtube.com/watch?v=LSHZ_b05W7o
The Ballad of Mr.Shadow: https://www.youtube.com/watch?v=lcGYEXJqun8
American Folk Songs EP: https://skyggewithai.bandcamp.com/
"Black is the Color” by SKYGGE featuring Pete Seeger: https://www.youtube.com/watch?time_continue=20&v=sv56kfvsQAM
On listener Jeremy Yuille's suggestion, here are a couple of tools referenced in this episode that you can use to get started with ML and creating art right now (no code required):
Comments
Reuben Nathaniel
Thank you for this interview guys! And now a fan of his music too! Wow
Mat
great tip, will edit the notes now. That is an incredible analogy, I need to digest Flatland now - it is true in many cases we feel that there are higher order processes happening across many fields that are really awkward fits in 20th century conceptual terms, not least when thinking about areas like political topographies or the creative act. One of our goals in running these conversations is trying to make sense of these new dimensions, and offer the opportunity for others to metabolize them. Exactly for the reasons you stated, it would be a shame to shut down new ideas because they don't port effortlessly to old models AND how can we learn from mistakes or naiveties from old models when considering how to get our heads around the new? Really appreciate your listening, hope we can keep it up!
Loved this episode thanks! You mentioned a few platforms throughout that would be great to have links in the show notes —ie runwayML and iva(?) etc Wanted to share a takeaway I picked up through the ep around how these technologically enabled abstractions introduce new—higher?—dimensions for expression which criticism and policing rush to address. Listening to your conversation, I’m reminded of Abbot’s novella Flatland: creative operations are happening in higher dimensions—François’s “bowls” for instance—that we can only really critically respond to when we render them back down to our perceptual bandwidth—audio files in this case. Critical appraisal follows up into this higher place, now needing conceptual frames and language of ML etc to keep up. Ways of framing access and the policing that often ensues then quickly follow. I’m looking forward to your discussion of these topics across the series. Thanks for the brain food!
