Amazon Wants Alexa to Guess What You Want Before You Ask

Amazon’s Alexa home assistant devices can now reportedly guess what users want before they even ask as engineers work to improve the device’s algorithm.

ZDNet reports that Amazon engineers have tweaked the Alexa home assistant algorithm to help the device predict users’ requests and offer to resolve them before the user even expresses their needs. In a blog post, Alexa engineers Anjishnu Kumar and Anand Rathi explained some of the improvements they’ve made to the virtual assistant’s algorithm recently.

The engineers stated that speaking with Alexa should be as normal and natural as talking to a real person and enabling the technology to predict what is going to be said next in a conversation is key to enabling smooth dialogue flow.

The engineers wrote: “Now, we’re taking another step towards natural interaction with a capability that lets Alexa infer customers’ latent goals – goals that are implicit in customer requests but not directly expressed.”

A number of sophisticated algorithms are required in order to achieve that level of intelligence for a virtual assistant. Alexa must analyze multiple features in users’ requests and compare them to previous patterns of interaction in order to predict the users’ needs.

Kumar and Rathi explained some of the issues they faced when training the algorithms: “Our early experiments showed that not all dialogue contexts are well suited to latent-goal discovery,” said Kumar and Rathi. “When a customer asked for ‘recipes for chicken’, for instance, one of our initial prototypes would incorrectly follow up by asking, ‘Do you want me to play chicken sounds?'”

The algorithm was trained by the engineers to make an assessment based on a number of factors ranging from the text of the dialogue to the user’s previous behaviors towards the virtual assistant in order to better predict their needs. “We are thrilled about this invention as it aids discovery of Alexa’s skills and provides increased utility to our customers,” the Amazon engineers said.