Human or Machine: Why Do We Want Robots to Sound Like Us?

Machines that talk have existed in our lives for years.

In 1939, Bell Laboratories changed everything by unveiling The Voder, the first electronic vocal synthesizer.

Featured at the New York and San Francisco World’s Fair, it was played using 10 keys, a foot pedal and a bar controlled by the operator's wrist. The demonstration marked the first time the world heard what a modern machine voice could sound like.

Today, most of us find ourselves surrounded by many machines that talk (Alexa, Siri, Google Assistant, to name a few) and they increasingly sound a lot like us. With advances in artificial intelligence (AI), it’s getting harder to distinguish these synthetic voices from human ones.

So why do we even want robots to sound like us in the first place, and what makes machine voices still sound robotic (albeit decreasingly so)? KQED explored these questions in a two-part project: "Send in the clones: Using artificial intelligence to digitally replicate human voices”

Chloe, meet “Chloney”. Follow KQED reporter Chloe Veltman’s journey to get her own voice cloned by a San José-based natural language speech synthesis company called Speech Morphing.

'Human or Machine'

Put your own ears to the test. See if you can tell the difference between human and machine voices in “Human or Machine,” our interactive voice game for Alexa and Google Assistant. The game, which sounds deceptively simple, was created by Lowell Robinson, KQED’s Voice and AI senior producer.

To play Human or Machine on Alexa, say “Alexa, open Human or Machine.” You can also find it here on Amazon.

To play on Google Assistant, say “Hey Google, talk to Human or Machine” You can also find in here the Actions Directory.