Okay, so last week saw Amazon’s Echo devices attract a air bit of ridicule as the Saturday Night Live featured a skit in which senior citizens would call upon Alexa using names like “Odessa” and “Alberta” and “Alisha.” There was a parody device called the Amazon Echo Silver, which would apparently respond to the most ridiculous names.
Well, the concept might work well enough on a TV show, but Amazon is not having it in real life. No indeed. Whether it is due to the fact that the company did not like people making fun of the Echo, calling it all sorts of weird names, or whether it was simply due to the fact that we cant have an Alexa powered device waking up for every random wake word — the company is now rolling out two-step verification of the “Alexa” wake word.
So here is how Alexa-powered devices roughly work: As soon as they hear the wake word, the device starts recording whatever the person is saying for eventual speech to text translation so that the command could be recognized and acted upon. The process is now being streamlined so that there is no scope for mistakes.
As per Alexa expert Ted Karczewski:
With this update, the wake word engine (WWE) on the device handles the initial detection of ‘Alexa’, and then a secondary cloud-based check verifies the utterance. If a false wake word is detected, the verification process directs the device to close the audio stream and turn off the LED indicator.
The two verification not only means that there would be fewer instances of Alexa waking up by mistake, it could also ease privacy concerns for folk who are worried about the device activating by mistake and then recording stuff when they clearly don’t want it to.
And of course, it would put paid to any chances of the parody translating into reality. Meanwhile, you can watch the video that has been creating so much buzz, right below:
