Voice recognition is one thing, creating natural sounding artificial voices is quite another. Lyrebird – a member of NVIDIA’s Inception startup program – uses deep learning to take this a step further, with a system that’s able to listen to a human voices and generate speech that mimics the sound of the original, human, speaker. We spoke with Lyrebird co-founder Jose Solero about the benefits of this technology, and why he feels the need to educate the public about what’s possible.