Google’s latest artificial intelligence experiment is a music-playing piano bot that digests whatever keyboard melodies you give it and tries to respond in kind. The neat tool is called AI Duet, and it’s part of an ongoing push from Google’s Creative Lab division to help the public familiarize themselves with AI and all the ways it can mimic human behavior — and even create art. A collection of music-focused AI tools were first shown off last fall, but now AI Duet in particular has been made available to the public.
AI Duet is an artificial piano player
AI Duet works by taking the notes you play using your computer’s keyboard and running it through a neural network that has been trained using machine learning with scores of examples. The neural net then looks for melodic and rhythmic patterns it can identify. Because it understands concepts like timing and staying in key from a hard data perspective — and not because it was ever explicitly told what to do — it’s able to then generate its very own organic melodies that, hopefully, sound like a direct response to your own. It was created by musician and computer scientist Yotam Man in collaboration with the Creative Lab and Magenta, Google’s open-source computational music and art project.
In practice, AI Duet doesn’t seem like it’s trying too hard to craft the next major hit. The melodies it spits back at you sound a bit random. There’s also no reasonable way that you could splice these segments together into an actual song even if you had wherewithal to do so, unless you were very musically inclined. But every once in a while, you do get a melody sent back at you that feels eerily as if it had to have been composed by a fellow human for how well it fits with your original.
Regardless of its effectiveness, AI Duet is yet another fascinating exploration of the creative side of AI, and one that fits well with the Google’s other musical experiments. Those include an “infinite” drum machine and this bizarre but incredible phone app that raps about what it sees through your phone’s camera by mixing image recognition and a speech synthesizer.
These web browser tools are all part of a growing number of initiatives at Google, including Magenta and Creative Lab, to bring its free-to-use AI tools to the public. The goal is to get everyday users to play around with these experiments to help inform how they’re built and how they can be improved, as well as to help independent developers and musicians create new experiences using powerful AI learning software.