Alexa will be able to talk or read you a story in your grandmother's voice even though she has passed away. It's bad news

Amazon held its re:Mars conference yesterday, and in it the head of Alexa development, Rohit Prasad, announced a new feature for this voice assistant: being able to imitate any type of voice with its voice synthesizer.

During the demonstration, it was possible to see how a device configured with this new version of Alexa read a story to a child with the voice of his deceased grandmother. The option is certainly unique, but the question is whether this is really a good idea.

Miracle or curse?

In the demo, a child was saying to an Echo, “Alexa, can Grandma finish reading ‘The Wizard of Oz’ to me?” After responding with her robotic voice, Alexa would immediately switch to a softer, human-like voice, apparently mimicking the voice of the little boy’s relative.

Those behind the development claim that their voice assistant is capable of producing high-quality voice “with as little as one minute of recorded audio.” The feature, still in development, has no estimated release date.

For Prasad getting artificial intelligence algorithms to converse and thus keep company has become a priority goal for Alexa, especially given that “so many of us have lost someone dear” in the COVID-19 pandemic.

The appearance of this option inevitably reminds us of the chatbot that Microsoft introduced in early 2021 that also made it possible to “talk” to people who had died. The controversy surrounding that development was remarkable, and eventually Microsoft decided that it was not going to continue developing it.