The Robot DJ

AI is colonizing space….our space

Tools and apps like Serato Pyro, Pandora, Echonest, Algoriddim and Pacemaker currently let you automatically turn your Spotify playlists into ‘flowing’ mixes. Pacemaker recently had a ‘mix-off’ with Steve Aoki, and they scored fairly evenly!

There is also the AI DJ Project which won an ‘Honorary Mentions’ award at Prix Ars Electronica 2018. This involves the use of deep neural networks to enable the software to select vinyl records and mix songs back-to-back with a human DJ, albeit with the help of a human to put the records on the deck. For it’s music selection it uses models and datasets covering genre inference, instrument inference and drum machine inference (note the omission of the human word ‘drummer’!). It ignores any metadata in the music, and focusses purely on the music using a customized turntable and robot finger for beat-matching. The software also includes a deep learning based motion tracking technique to observe how many people in the audience dance to the music it plays.

Random Randomness

During the music selection process, the system tries to select tracks with similar mood, as long as the amount of the body movement is more significant than a given threshold. Once the index gets less than the threshold, random noise, inversely proportional to the amount of the body movement, was added to the feature vectors of incoming music, so that the system might be able to explore a new musical realm and (hopefully) stimulate the seemingly bored audience. Unsurprisingly, this randomness apparently worked as a feedback loop in the performance: the randomness brought more confusion to the audience, and it led to more randomness. It ended up proving the difficulty to maintain a subtle balance between regularity and unexpectedness in the DJ’s music selection process.’

Hang the DJ

One of the future directions the developers are taking is that the audience still needed a physical embodiment of AI DJ, in order to project an emotional bond with the music. At least it is a critical investigation into the unique relationship between humans and machines, at most…the beginning of the end? Will we soon see robots entering the various DJ mixing competitions around the world? Plus, a computer would never let you down, show up late or turn up drunk!

Click through to read Article: A dialog between human and AI through music

Then there is the various research based on Alan Turing’s 1950 paper ‘Computing Machinery and Intelligence’ investigating artificial intelligence. Dartmouth College Neukom Institute’s ‘Turing Tests In Creative Arts’ have conducted evaluations of machine generated music composition, in relation to their artistic creativity. In order to pass the test, the music is to be indistinguishable from human generated music composition.

Click through to read Article: Music Turing Tests 2018

Does any of this ultimately matter?

Young people who have grown up in the internet age are already used to interacting remotely with other people, with seemingly less need to interact ‘in the flesh’. As the gap between AI and humans gets ever closer, we are going to have to come up with ideas and experiences that technology cannot emulate, in order for us to carry on ‘being the DJ’. Or will the technology create more options for our human creativity, apart from simply being there to run the machines?

One thing is for sure, in the words of virtual reality pioneer Jaron Lanier ‘The world is changing rapidly under our command, so doing nothing is not an option’.