How we talk to technology today
For many of us, conversational technologies have become a part of everyday life. Virtual assistants like Alexa, Siri, Cortana, Google Assistant help get us through the day. Amazon Echos and Google Homes share our living rooms with us. Interacting with our favorite brands often take place through a chatbot, not a human.
Years ago, it was 2001: A Space Odyssey that first introduced us to Hal 9000. Knight Rider captured our imaginations with KITT. It seems, like in so many areas of our modern world, we continue to move more and more toward the stuff of science fiction.
But when we talk about conversational technology and AI, what can we really expect in the near and distant future? How are researchers at PARC and other organizations thinking about and designing these “invisible” conversations with technologies? How will the ways we live and work be impacted by it?
At our free PARC Forum January 18 in Palo Alto, PARC Research Manager Kyle Dent will be joined by Ron Kaplan, former VP of Amazon and former PARC Research Fellow and Research Manager, Natural Language Theory and Technology, Cathy Pearl, VP of User Experience at Sensely, and Abi Jones of Google for a panel discussion about these very questions.
Where user interfaces are taking us
Kyle Dent and his colleagues at PARC have been exploring conversational interactions with technology for years. Currently, their focus is on enterprise intelligent assistants. For his team, it actually starts with studying human-to-human communication and interaction.
“By understanding the conversational mechanisms that humans use with each other, we can understand how to create better intelligent assistants,” says Kyle. “We don’t want to pretend to be human, we just want it to feel like a normal, natural, personal interaction. If we can provide truly conversational interactions, what kinds of things might that enable that aren’t possible or practical today?”
For Kyle, conversational user interfaces (CUIs) represent a new paradigm in how people interact with computers, one that could be as transformative as the graphical user interface (GUI) was for the personal computer.
“Our smartphones are allowing us to do things with our phones, hands- and vision-free, that we couldn’t have imagined doing on a desktop,” says Kyle. “Just as the GUI helped propel PCs into widespread use by making interactions easier and more natural, the CUI could be the enabler of natural conversations with things, in ways we haven’t even thought of yet.”
As companies and organizations now begin to push into the world of conversational interactions with computing systems, PARC and others are interested in what the technology can enable employees to do in the workplace.
Focusing the conversation on business
Whereas Alexa or Siri can answer general questions by relying on widely available sources of information, like Wikipedia, businesses and organizations often have unique processes and information stored in unstructured forms.
Kyle and his team at PARC are working on enterprise intelligent assistants to streamline the process of gathering information, whether from documents, internal corporate wikis and webpages, or even knowledge stored in the brains of employees.
For example, instead of asking an IT or HR team member a routine question, employees can ask a conversational agent. If the system doesn’t have the answer, it will then ask that relevant expert, and after collecting the information, will store the data for future use. A conversational agent could save the employee hours each day answering questions.
Understanding how dialog works is key to enabling high-value interactions. “There’s a decision-making process happening at every turn of the conversation to figure out what should come next,” says PARC Researcher Jesse Vig, who focuses on using machine learning to improve the user experience in these systems. “The conversational system starts with a program that automatically recognizes speech and converts the voice signals to text. Another module examines and tries to semantically understand what the text is asking. That language understanding piece then goes to a dialogue manager that has to decide, in this context, what is the next best thing to do?”
Of course, developing high-value conversational systems for businesses requires a lot of work and investment. Kyle and his team focus on solving business challenges that have the biggest impact.
“We try to look at manageable tasks within organizations where the return on investment is most demonstrable,” says Kyle. “From there, we extend it to more things that it can do on behalf of the worker and organization.”
Plenty more conversations to come
As conversational technologies continue to get smarter and more natural, the things we talk to will feel more and more like something out of our favorite sci-fi movie. If recent Echo and Google Home sales are any indication, the technology will only continue its rapid evolution. And designing for business applications is one area that has exciting room for growth. But conversational technologies demonstrate promise in many other areas as well.
As one example, PARC is working with the Vista Center for the Blind and Visually Impaired to test its enterprise intelligent assistants with real users, and from those conversations came a long list of ideas for new ways conversational technology could help the blind and visually impaired. We’ll talk more about that in our next article.