Imagine getting a text from a friend and your iPhone reads it to you in your friend’s voice. Rather than a recording, their speech would be simulated from a voice model.
That’s the goal of a patent Apple submitted recently.
Get texts from your bestie read in their (simulated) voice
Say “Hey Siri, read me my most recent text message” and the digital assistant will do so… but in Siri’s own voice. Apple wants to personalize that process.
It submitted to the U.S. Patent and Trademark Office a request for a patent on system for Personalized Voices for Text.
The filing starts by pointing out the problem: “Conventional systems do not include personalization of message playback, such as personalized audio playback of received messages.” Then it describes a system to “produce personalized voices for text messages based on a small amount of audio input.”
In basic terms, someone could to create a voice model for themselves with software on their iPhone or other Apple device. When texting with a friend, coworker, etc., they’d have the option to send that voice model to the message recipient (shown above). The second user could choose whether to receive the voice model, which would then be used to read aloud messages coming from the first person.
Apple has built many features into its Messages application to make it better than a “bare bones” text app, including the infamous blue and green bubbles. The ability to read incoming text messages in a simulation of the sender’s voice seems like a reasonable addition.
That said, whether Cupertino plans to take the next step and turn this idea into reality is anyone’s guess. Company engineers frequently patent designs that are never seen again, like this iPhone game controller or a MacBook with a popup keyboard.