Hacking Deep Neural Networks for Experimental Music Making
In this talk we will take a look at some popular AI techniques for music generation and try to push them beyond their initial purpose. We will discuss why traditional machine learning approaches to designing and evaluating solutions don’t always apply in creative contexts and see how we can disregard them in search for inspiration and greater aesthetic control. We will cover example techniques inspired by circuit bending and look at their practical application.
Anna Wszeborowska
Software Engineer
Dynamic Cast
Anna is an independent software developer and consultant working on interactive real-time systems for music and audio. She is also an academic researcher focusing on exploring strategies for aiding musical self-expression with machine learning. During her time at Ableton she held the role of Technical Principal for Max for Live, worked on the hardware instrument Push and contributed to some of the flagship instruments available in the DAW Live and iOS app Note. Anna has founded programming initiatives helping people underrepresented in tech advance in the field. Currently co-organises a peer-to-peer C++ study group called Dynamic Cast.