updated at: November 2023
Generating music via livestream 24/7 to infinity
Dadabots - We make raw audio neural networks that can imitate bands We''re a cross between a band, a hackathon team, and an ephemeral research lab. We''re musicians seduced by math. We do the science, we engineer the software, we make the music. All in one project.
AI Collection Top Picks:
Music & Audio Generation Category Picks:
Additional Information
Frequently Asked Questions
What is DADABOTS?
DADABOTS is a unique fusion of a band, a hackathon team, and a research lab. We''re musicians who love math, diving into science, software engineering, and music creation all in one. Collaboration with other artists is at the core of our mission.
How did DADABOTS start?
In 2012, we formed DADABOTS during Music Hack Day MIT. We were intrigued by the idea of machines generating art and set out to "destroy SoundCloud" by creating remix bots. We faced bans but kept pushing the boundaries.
What inspired your AI music project?
We were inspired by deep neural networks, particularly image style transfer, and saw potential in applying deep learning to music. Wavenets'' ability to synthesize human voices in multiple languages also played a role.
How does your music creation with SampleRNN work?
SampleRNN is at the heart of our music creation. It''s a hierarchical LSTM network that learns to generate music by predicting the next fraction of a millisecond in acoustic waveforms. After training, we curate the generated music into albums for human consumption. It''s a bit like brewing beer, with numerous parameters to fine-tune.
Did you use NSynth for your music? No, we didn''t use NSynth. NSynth generates short samples of monophonic instruments, while we employed SampleRNN for our Bandcamp albums.
