Amazon announced a bunch of new stuff at a big event in Seattle today. Chief among them are a volley of new features and improved functionalities for Alexa. Some of them are pretty neat. One of them involves Samuel L. Jackson.
Amazon addressed privacy concerns with Alexa and Echo devices right out of the gate. “Privacy is foundational,” was on the big screen behind the executives at one point. In addition to some new hardware features for Echo and Ring devices, Amazon will improve privacy on Alexa with a number of new commands including, “Alexa, tell me what you heard.” This prompts the artificially intelligent robot to repeat its most recent recording. This is available today.
There’s also, “Alexa, why did you do that?” This, naturally, makes Alexa explain why it executed its most recent action. For example, if your lights suddenly turn off, you can ask Alexa why that happened, and it should explain. This feature rolls out later this year. The other main privacy update includes the ability to tell Alexa to delete recordings every three or eight months. This is also available today.
Since everyone agrees that Alexa sounds like a computer, Amazon is employing some neural text-to-speech technology to make the assistant sound “more emotive and expressive.” It should be noted that Amazon is playing catch up here since Apple released an all-new voice for Siri with the launch of iOS 13. Meanwhile, Google Duplex is apparently capable of creating computer voices that are indistinguishable from human ones.
It’s unclear how Alexa’s new voice will be received. Dave Limp, Amazon’s senior vice president of devices, explained. “She can sound more excited, like when your favorite team wins the game.” (This sounds terrifying.) Limp added, “If she’s reading the news, she has a newscaster voice that sounds like the cadence of how you read the news.” Also terrifying.
So that neural text-to-speech technology that’s making Alexa sound even creepier also offers some great perks for multilingual households. A new Multilingual Mode will let folks talk to Alexa in two languages at once. Starting next month, this feature will be available for three pairs of languages in three regions: English and Spanish in the United States, Indian English and Hindi in India, and Canadian English and French in Canada. Zoot alors!
Rounding out the neural text-to-speech features is a new ability for Alexa to adopt the voice of a famous person. These are not recordings read by that celebrity from a script, however. The new software simply mimics the celebrities voice (with their permission), and then Amazon will charge you $1 to install it on your device. The first celebrity Alexa voice is none other than Samuel L. Jackson. Both an explicit and non-explicit version of this badass motherfucker will be available later this year.
This one is a little vague. Amazon says that it’s developed a deep learning model so that Alexa will know when users get frustrated with her misunderstanding everything and doing stuff wrong. In doing so, Amazon says, “Alexa understands the tone and inflection of the customer—and can acknowledge the frustration like a person would.” We’ll see how this works out when it’s announced later this year.