User (Hears a song on the device): When is their next concert?
System: Coldplay’s next concert is in Seattle on March 25. Would you like me to get your tickets?
User: Yes.
System: That concert is sold out, but there is a concert in Portland on March 24. Would you like tickets to that one?
User: Sure.
System: Tickets are US$35 to $100. Here is the seating chart.
This is an example of the kind of interaction a user and his/her device can engage in through voice artificial intelligence, a technology that Bellevue, WA-based VoiceBox Technologies focuses on. “It catalyzes a multi-user, cross-device, cross-app and cross-ecosystem experience capable of intelligently assisting you wherever you go – including car, home and work,” said Phil Cohen, Senior VP of Advanced Technologies at the company.
At the heart of voice AI is natural language understanding, with which the user is able to get answers to complex and conversational questions from structured and unstructured web, corporate, and personal knowledge sources. “For example, systems will be able to read and understand textual data, such as medical research articles, and answer questions about the results described therein,” Cohen said. “Users will also be able to provide commands and instructions to their devices and robots, to be executed now or at some future time subject to the user’s stated conditions.”
And according to Cohen, voice AI extends that natural language understanding capability and helps achieve a higher level of accuracy. “Voice AI employs machine learning and deep neural networks to process natural language using the user’s physical, social, and linguistic context, and incorporates communication using multiple modalities, such as voice, gesture and vision,” he said. “It broadens natural language processing by using a semantic parser to turn language into meaning representations, which are then input to reasoning components that generate an intelligent response that uses large knowledge graphs to address the user’s plans and intentions.”
According to Cohen, voice AI has opportunities in a variety of markets including medical, industrial and infotainment. The company recently announced a R&D partnership with Macquarie University, one of the leading research hubs for natural language understanding worldwide. Under the partnership, Macquarie professor Mark Johnson has been named the chief scientific officer of VoiceBox Australia and will lead the VoiceBox office as a center of excellence for voice AI.
According to Cohen, the future potential of voice AI can’t be ignored. “In the next 10 years, voice AI will enable human-computer interaction via speech to be available for nearly every application. As larger knowledge bases are made available, enabled by computers reading text, systems will begin to reason and acquire common sense.”