AI isn’t the end-all-be-all relating to conversational experiences like Amazon’s Alexa. In reality, handbook options to issues are generally superior to automated, AI-driven fixes. That’s in line with Amazon Alexa AI director of analysis science Janet Slifka, who spoke throughout a session at present at VentureBeat’s Transform 2020 convention.
“If someone calls customer service and says something to the effect, ‘Alexa doesn’t understand me,’ in many cases, we can be faster to the customer with a manual fix,” Slifka mentioned, describing her crew’s work in triage. “In some cases, it’s almost required when you need words to enter the lexicon. You’re probably not going to wait until you rebuilt and deploy a statistical model when a new word gains prominence, like ‘Brexit.’”
This dedication to consistency on the expense of the bleeding edge is mirrored in Alexa Answers, a decidedly simplistic method to increasing Alexa’s base of data. Alexa Answers lets any buyer within the U.S. and Germany submit responses to unanswered questions. And whereas the platform isn’t good, Amazon says it permits Alexa to reply questions quicker than any automated backend might — particularly for breaking news.
“When you start with the customer, you look at all of your tech options. You may not choose the latest and greatest that caught your eye,” Slifka mentioned. “You might allow that the newer things that you’re doing are messier and more challenging. For that reason, you might have more manual processes in there until you really hone things.”
That mentioned, Amazon hopes to use extra autonomous, AI-driven methods to issues sooner or later. Scientists at Alexa AI are pursuing semi-supervised and unsupervised studying, during which AI methods study to make predictions with out ingesting gobs of annotated information. One manifestation is Alexa Conversations, an AI-based service that allows builders to create Alexa apps that seamlessly interconnect with different apps. Another is Alexa’s just lately launched good dwelling routine options. If prospects ask one thing like “Alexa, turn on the Sofa Lights” however the lights they’re attempting to activate are literally named “Living Room Lights,” Alexa may helpfully recommend “Did you mean Living Room Lights?”
“The idea is to remove the manual fix because the statistical model is the stronger option for the customer,” Slifka mentioned.