We will see a lot of new developments in the near future around new sound generation and synthesis based on machine learning. Specifically, neural networks with a lot of layers (deep learning) can be applied to 'learn' new types of soundwaves, based on learning through feeding the neural network with existing soundwaves. Google has done just that by developing an instrument called the NSynth Super. Basically a sample player with new sounds learned and generated offline by the neural network and fed into the NSynth Super. You can set the output waveform on a touchpad by mixing the characteristics of 4 selected basic waveforms. The synth is available as an open source build-it-yourself project. Take a look at the video to see what it is all about.
I like music, especially making music. I listen to a lot of different stuff, but I like soul, jazz and fusion music most. I like to develop and experiment with tools to make music performance and improvisation on the spot more exciting. It is fun when unexpected musical things happen during a performance!
In recent years I developed the MIDI real-time Harmonizer, a tool for generating harmonized chords when playing a solo line. Inspired by the work of saxophonist Michael Brecker, I started to develop this tool to run on Mac and PC. Brecker played the Electronic Wind Instrument and used the Oberheim Xpander synth to generate random chords from an EWI solo line. Now you can accomplish the same thing with the MIDI real-time Harmonizer driving your own favourite (plugin) synths.