Alexa ‘could be about to get even creepier,’ Amazon patent reveals

FILE - In this Aug. 16, 2018, file photo a child holds his Amazon Echo Dot in Kennesaw, Ga. Amazon met with skepticism from some privacy advocates and members of Congress last year when it introduced its first kid-oriented voice assistant , along with brightly colored models of its Echo Dot speaker designed for children. (AP Photo/Mike Stewart, File)
Could Alexa be about to get creepier? (AP Photo/Mike Stewart, File)

Amazon’s listening device-cum-speaker Echo could be about to get even creepier, according to a new patent application filed by the online shopping giant.

It would alter how Echo works - so that instead of waiting for its ‘wake word’, Alexa, it would effectively be listening all the time.

The idea is to capture audio when people forget to say ‘Alexa’ at the start of a statement and instead say, ‘What’s the weather, Alexa’, for instance.

The patent application says, “A user may not always structure a spoken command in the form of a wakeword followed by a command (eg. 'Alexa, play some music')," the Amazon authors explain in their patent application, which was filed back in January, but only became public last week.

Read more from Yahoo News UK:

Jeremy Hunt says no deal ‘political suicide’ for Tories

Albino panda caught on camera for first time

Spice Girls fans demand refund over ‘diabolical sound’

“Instead, a user may include the command before the wakeword (eg. 'Play some music, Alexa') or even insert the wakeword in the middle of a command (eg. 'Play some music, Alexa, the Beatles please'). While such phrasings may be natural for a user, current speech processing systems are not configured to handle commands that are not preceded by a wakeword.”

Instead, the patent application proposes, Echo devices will continuously record.

The data won’t be transmitted to Amazon’s cloud for interpretation unless the ‘wake word’ is detected.

Instead, when ‘Alexa’ is heard, the device will capture the speech just before and just after, and then send it for analysis.

The patent application says that the ‘system is configured to capture speech that precedes and/or follows a wakeword, such that the speech associated with the command and wakeword can be included together and considered part of a single utterance that may be processed by a system.’