Google AI creates sounds that no human has ever heard

Google's artificial intelligence research team Magenta has created software capable of generating entirely new sounds. The software uses thousands of instruments to create sounds that are otherwise impossible to produce. The Neural Synthesiser, or'NSynth', invents audio using deep AI neural networks, which blend the sounds of two instruments to create a novel, hybrid sound. The technology uses AI neural networks to blend two sounds together. This is not the same as layering two sounds on top of one another.