Amazon today announced that AI assistant Alexa will now automatically begin work to recognize a user’s voice when they say “Alexa, play music.” Automatic voice identification that occurs when saying this phrase will serve up personalized selections from Amazon Music. Automated voice ID can be disabled using the Settings section of the Alexa app.
Setting up an Alexa voice profile allows users to receive personalized results when playing music, sending messages, and allows users to complete shopping transactions without the need to enter a four-digit pin code. Automatic voice recognition when using the “Alexa, play music” command is limited to personalized songs from Amazon Music.
The ability to teach Alexa to recognize your voice so you can receive personalized results was first announced in October, and since then Amazon has been trying to make it easier for people to create unique voice profiles. Creating a voice profile used to require use of the Alexa app, but voice profile creation with an Echo speaker was first introduced in February.
The previous procedure for creating a voice profile required a person to say 10 unique phrases. An Amazon spokesperson declined to provide additional details about the number of recorded interactions necessary to train Alexa to recognize a voice.
In addition to personalization for music and the like, work is also underway by Amazon’s team to better understand a person’s emotional state when speaking to Alexa, knowledge that could lead to additional personalized experiences in the future.
In related news, developers can now add up to eight unique humanlike voices from AWS Polly that sound different than Alexa, and earlier this month in-skill purchases for voice apps were introduced so that third-party developers and businesses can accept payment from millions of Alexa users in the U.S.
Amazon’s Alexa is able to recognize up to 10 voices per household.
Update 1:54 p.m. to include additional comment from an Amazon spokesperson.