Using artificial intelligence to digitally replicate human voices | NPR

Share on facebook
Share on twitter
Share on linkedin

Given the intricacies of speech synthesis, it comes as a surprise to learn how simple it is to purchase one. All a client needs to do for a simple conversational build is record themselves talking a bunch of prepared lines for around one hour. That’s all there is to it. Siri, Google Assistant, and Alexa, as well as a bank’s automated customer care line, now sound quite real. Artificial intelligence, or AI, has advanced to the point that it’s often impossible to discern the difference between synthetic and human voices. The global speech and voice recognition sector, in reality, is valued at tens of billions of dollars and is rapidly expanding. 



Subscribe to our newsletter