How often have you accidentally had Siri or Google Now activate because something was said in a song, on radio, in a movie, or by someone else that sounded like the activation word? Just last month it seems that a news broadcaster had accidentally set off quite a few Amazon Echo devices and asked them to buy dollhouses because of what he said.
Granted there wasn’t really any damage and to a certain extent it is pretty funny, but many have suggested that perhaps in the future, being able to distinguish between voices is something that companies should look at. We’re not sure where Apple and Google are in their efforts, but a report from TIME has revealed that Amazon could be working on something of their own.
According to the report, they quite sources who claim that Amazon is working on a feature that will help Alexa distinguish between different voices, such as those made by the owner of the Alexa-powered device, versus a sound from TV, music, and other users. The report claims that this feature is called “Voice ID” and has actually been in development since at least the summer of 2015. Assuming this is true, it means that Amazon has known about these limitations for a while.
Distinguishing between voices could also be useful for security purposes where trying to pull up private information will not be allowed unless the voice software detects that it is the owner. No word on when Amazon plans on rolling out this feature, but it sounds like it could be quite welcome.
Filed in AI (Artificial Intelligence), Alexa and Amazon.. Read more about