Alexa Personalization: New Features Tailor AI to You

Amazon Enhances Alexa with New Personalization Features

Amazon is poised to introduce a set of three innovative features designed to empower users to personalize their Alexa experience through simplified AI training tools. Within the coming months, consumers will gain the ability to instruct Alexa to recognize specific sounds within their homes, such as the chime of an instant pot or a doorbell.

Sound and Visual Recognition Capabilities

For Ring device owners, the AI will be capable of detecting visual alterations – for instance, identifying when a door intended to be closed is left open. Furthermore, users will be able to explicitly guide Alexa in adapting to their individual preferences regarding favored sports teams, preferred weather applications, or dietary choices.

These advancements were unveiled during Amazon’s autumn event, alongside the announcement of the latest Echo devices and other hardware innovations.

Building on Alexa Guard

The new sound-identifying functionality expands upon the existing Alexa Guard feature. Alexa Guard currently identifies critical sounds like breaking glass or alarms for fire or carbon monoxide. This is particularly beneficial for individuals away from home or those with hearing impairments, providing alerts during potential emergencies.

With a premium subscription, users can even activate a simulated barking dog sound when motion is detected by a smart camera.

Amazon is now exploring how Alexa’s sound detection can be applied to non-emergency scenarios.

Custom Sound Event Detection

A forthcoming feature will allow users to train Alexa to recognize specific sounds important to them. Examples include the beep of a crock pot, an oven timer, a refrigerator alarm, a garage door opening, or a doorbell. The system requires only six to ten sound samples for learning.

This represents a significant reduction compared to the thousands of samples previously needed to train Alexa for other sounds. Training can be initiated directly from an Echo device or via the Alexa mobile application.

While the training process occurs in the cloud, sound detection will be performed locally on the device, ensuring user audio privacy after initial enrollment.

Once trained, users can trigger notifications or routines based on detected sounds. This can enhance accessibility or elder care, such as displaying doorbell notifications on a Fire TV. It can also streamline daily routines, like initiating a “I’m Home” routine when a garage door opens.

Custom Sound Event Detection is slated for release next year.

Custom Event Alerts for Ring Cameras

Similarly, users will be able to train Ring cameras to monitor a specific area and detect changes. Currently, this functionality, termed “Custom Event Alerts,” is best suited for binary changes, such as a shed door being open or closed.

This feature will initially be available to Ring Spotlight Cam Battery customers in the coming months.

Personalized Preferences for Skills and More

Another new Alexa feature will enable the assistant to learn user preferences related to food, sports, and third-party skills. Users can initiate the learning process by stating, “Alexa, learn my preferences.”

Learning can also occur implicitly. For example, specifying “Alexa, some of us are vegetarian” during a restaurant search will exclude steakhouses from suggestions.

Alexa will also incorporate highlights from favored sports teams when providing sports updates. Furthermore, the AI will default to preferred third-party skills instead of its native responses.

Currently, only third-party weather skills are supported, but Amazon intends to expand this functionality over time. This aims to improve skill usage by enabling a “set it and forget it” customization experience.

Amazon assures users that preference data is linked to an anonymized customer ID and can be adjusted. The data will not be used for customizing shopping suggestions on Amazon.com.

This preference learning will be available before the year’s end.

Towards Ambient Intelligence

Amazon emphasizes that these features represent progress towards its vision of “ambient intelligence.”

Rohit Prasad, SVP and head scientist for Alexa, explained that ambient AI “can learn about you and adapt to your needs, instead of you having to conform to it.”

He further stated, “Alexa, to me, is not just a spoken language service. Instead, it is an ambient intelligence service that is available on many devices around you to understand the state of the environment, and even acts proactively on your behalf.”