Alexa is smart, feature-rich, and pretty innovative, but it has trouble spots, particularly with differentiating different voices to respond to. That may change soon.
Amazon is reportedly developing an advanced voice-recognition feature that will allow Alexa to tell different voices apart, according to sources familiar with Amazon's Alexa strategy.
Will Alexa Be Able To Tell Voices Apart Soon?
The report comes from Time. Alexa, like many virtual assistant, including Apple's Siri or Microsoft's Cortana, are designed to respond to voice prompts such as "Play some music," "What's the nearest gas station?" and even "Order Starbucks." Amazon's new reported feature, according to Time's report, would match a person's voice with a voice print for verifying voice commands.
This means an Amazon Echo, or any device that supports Alexa, would only respond to a specific voice print and not to others. For instance, the technology could be used to only respond to a parent's voice when making purchases and ignore that of a child's when making a similar request — such a flub has happened before, to largely comedic results.
Alexa already allows multiple accounts and PIN verification for purchases, but adding a layer of security by employing voice-specific verification will certainly aid the process altogether.
Amazon Alexa Voice ID
The said feature is reportedly known internally as "Voice ID" and has been in development since 2015. Of course, the new voice-specific recognition technology brings further security concerns to Alexa, which is already in a tussle over a legal controversy regarding private data.
Time's sources claim that the technology is already technically ready and that Amazon is simply trying to figure out how to roll out the feature to Alexa users without compromising privacy.
At any rate, a technology such as Voice ID is pretty interesting, and it'll help solve one of Alexa's foremost issues, of course, that is if it works as described. One can only imagine the extent of Amazon's exploration with voice-recognition technology. It could give Amazon's smart speaker line of devices further boost. Since 2014, Amazon, along with third-party developers, have been cooking up different skills for Alexa, the number of which has recently passed the 10,000 mark.
It could also be Amazon's edge over Google's smart speaker Home, which is powered by Assistant. At present, Home can't distinguish between different voices yet. Such a feature could prove practical for households with multiple Echo devices and multiple accounts across them.
Voice Recognition Will Make Alexa Smarter
The concept would play well into Amazon's announced plans for Alexa. Speaking to CNET during CES, Mike George, VP for Alexa, said that it was working to teach the virtual assistant how to better understand context, such as putting in mind prior requests. It's also working to make Alexa's responses more conversational. Coupled with voice recognition, that could make for some pretty interesting results. Imagine if an Echo owner's guest spoke to Alexa but gets rebuffed: "I'm sorry. You aren't my owner," or something along the lines.
"She will become smarter and use that information to have much more natural interactions with you," he said last month.