Panel session
Can Musical AI Be (A)Live?
DATE: July 6, 2021
TIME: 17.00 - 17.40 CEST
HOST: Ryan Groves
PANEL: Ash Koosha, CJ Carr (Dadabots), M. Stella Tavella
REGISTER: This session is part of the 4-day Music & Innovation Summit by Wallifornia MusicTech. You can register for the whole event here, and attend all sessions for free.
ABSTRACT: For many, it can seem like Artificial Intelligence is on the verge of achieving general intelligence (i.e., to learn and understand intellectual tasks like a human). In reality, AI is usually trained to do very specific tasks in controlled environments. In music, AI has been used for almost every step in the creative process. However, the challenge to extend these technologies into a space where the AI can interact or react to musical situations -- such as livestreaming or live concerts -- is daunting. Beyond that, the cultural relevance of musical AI and virtual beings is just becoming a possibility (outside of Japan, of course). So even if these technological hurdles can be overcome, there is yet another hurdle to make the music matter to the general public.
This panel features a group of experts who will discuss these parallel challenges in the technological and cultural domain: Ash Koosha, the founder of Auxuman, where they created the first AI musical supergroup (Yona, etc.), and recently released a music video created entirely with AI techniques (“Incomplete”); Stella Tavella, machine learning engineer and generative music hacker who works by day as a Data Scientist for Musicxmatch; Dadabots, a house name in the AI music space through their 24/7 livestreams of generated music; and Ryan Groves, founder of Infinite Album, where they’re using AI to replace game soundtracks and facilitate engagement through music in livestreaming.
Read more about the host and panelists below.
Host
Ryan Groves is an expert in AI-driven music composition. He built the Best Music App of 2015, Ditty: a musical messenger that sang your texts automatically. He then founded Melodrive, the only real-time AI composer. He got his M.A. in Music Tech from McGill, and his work on computational music theory won Best Paper at ISMIR in 2016. Now, he is using AI to generate adaptive soundtracks for Twitch streamers with his new startup, Infinite Album.
Panelists
To say Ash Koosha is an Iranian-born, London-based electronic musician is barely to scratch the surface. A synaesthete who sees colours as sound, a virtual reality pioneer, a software humanist, Artificial Intelligence (AI) innovator and a former rock musician, he has been pushing the intersection of creativity, AI and interactive media.
Ash founded Auxuman, where he built the first virtual group of AI-driven musical artists, including YONA. He has since also founded Oorbit, which looks to build immersive experiences accessible to anyone via the web, and still quenches his creative appetite by releasing music and AI-driven visual pieces such as "Incomplete" (a collaboration with Dalena Tran).
CJ Carr met Zack at Berklee College, back when they used to play instruments. After falling down the rabbit hole of python, arXiv, and github - they formed Dadabots, a mythical AI death metal band & neural audio synthesis research lab. As Prometheus brought fire to man, Dadabots bridged the ivory tower into music culture, carrying Theano models with buggy dependencies, GPU blazing, ushering the earliest neural audio synthesis experiments into the hands of musicians, crossing the deadly crevasse from PhD research into avant-garde metal, breakcore, beatbox champions, and grammy-nominated hip hop producers. Their musicianship has since deteriorated, and they are embarrassingly out of practice.
M. Stella Tavella is a Machine Learning Engineer with expertise in Music Information Retrieval. She graduated in Sound and Music Engineering & Computer Science at the Polytechnic university of Milan. She started her career in the belgian startup Musimap, working on music recommendation systems, she is now an AI Engineer at Musixmatch. Stella's passion for experimental electronic music production soon joined the Music AI world as a tool to enhance the sound design and composition process.