[operatorius]

well you have pure data, max/msp, reaktor, etc for that. These experiments with ML can approximate the soundwaves generated by particular instruments where the soundwaves produced are very complex, but other than that there are tons of software that can help you design any sounds you like. Machine learning is like the micro thing in the 70's-80's companies are trying to put it everywhere they can despite there are no benefits to the existing convenient products/solutions or not incorporating ml at all but still labeling the product. i.e. quoted from wiki: Microwaves are a form of electromagnetic radiation with wavelengths ranging from one meter to one millimeter. I see nowhere the 10^-6 meter waves being used in our commonly used kitchen appliance

[TheOtherHobbes]

There are any number of ways to approximate the soundwaves generated by particular instruments that work better than this.

There are also any number of ways to make cool new sounds, even using old tech. (My current favourite is Aparillo by SugarBytes, which uses FM but sounds absolutely amazing.)

This Google project seems to be a technology demonstrator made by people who seem curiously unaware of the domain they're trying to work in.

There is a lot of competition in this space and this project is doing nothing remotely new, including the ML element.

All of which might be forgivable if the sound was unbelievably awesome... but it isn't.