Amazon’s Alexa Will Now Ask You Follow-Up Questions for Integrated Conversational Experience

On Wednesday, Amazon announced that in its recent update they have made Alexa, the digital assistant model, even more talkative. In the updated version of the assistant, it will be able to infer a user’s latent goal and use those to responding to demands from consumers with follow-up questions.  

For instance, if a user asks Alexa for the time taken to fully cook an egg. The assistant tells you that it takes roughly 1 1/2 minutes, then she will ask you if you’d like to set a timer.  

Amazon says “the update involves algorithmic tweaks and a deep learning model that will allow the device to evolve based on its relationship to the user. If you’re asking Alexa about how to cook eggs every morning and always opting to set the timer, Alexa’s discovery model will use active learning to improve its predictions and more accurately conclude whether or not you want to know when those 1 1/2 minutes are up.”  

Also Read: FDA Approves Apple Watch App That Uses Vibrations To End PTSD Caused Nightmares 

In a blog post, Amazon wrote “Amazon’s goal for Alexa is that customers should find interacting with her as natural as interacting with another human being.” “While [apps] may experience different results, our early metrics show that latent goal [inference] has increased customer engagement with some developers’ apps.”  

The company seems to acknowledge that latent goal discovery might become annoying for users. For instance, in early prototypes, when users requested “chicken recipes,” Alexa would follow up with- “Do you want me to play chicken sounds?”  

So, to mitigate unwanted suggestions, the company has implemented a deep-learning model that factors in different aspects of the dialogue context. With its help, Alexa will be able to read the customer’s current session and determine whether the customer has engaged with Alexa’s multi-skill suggestions or not.  

The news of new inference comes as Amazon launches Alexa Conversations with a series of deep neural networks that aim to help developers easily integrate a natural conversational experience into custom apps.