TECH TALK

Comment on this article

Artificial Intelligence: Info-Seeking with Voice Assistants

by Jenna Kammer

The holiday season recently ended and everyone is talking about their new Artificial Intelligence (AI) devices. Or, for me, I chose not to get an AI device (yet), but instead have been watching shows like Humans and Deadwood that show how AI can develop into human-like robots. Exciting stuff!
The state of AI is in the process of rapid development right now. For an overview of what is going on (and what is predicted), I suggest starting with this review from the MIT Technology Review, https://www.technologyreview.com/s/603216/5-big-predictions-for-artificial-intelligence-in-2017/, or this Economist article from 2015, http://www.economist.com/news/leaders/21650543-powerful-computers-will-reshape-humanitys-future-how-ensure-promise-outweighs.  In summary, an AI device is one that (through programming) can make rational choices, like learning and problem solving. These computers can understand speech and language, while also performing an action in response. An AI device can range from very sophisticated (like self-driving cars) to simple (like a voice assistant from Amazon or Siri on the iPhone).

The TV shows I mentioned above show how AI will eventually develop so that computers will someday outperform humans. In the show Humans (available for streaming on Amazon), the teenager characters struggle with how this will affect their future careers (why study engineering when a robot will be able to do it better). This is just fiction, but the concept is not far off. Once computers outperform humans in simple tasks like remembering, recalling and processing (let alone, seeing, speaking or hearing), society will be in for some huge changes.

Right now voice assistants, like Amazon’s Alexa (Echo) or Siri (iPhone), are very limited in what they can do. Cnet put together a list of commands that Alexa is capable of, and it’s not that long really: https://www.cnet.com/how-to/the-complete-list-of-alexa-commands/. There is a lot of development yet to happen, but when you look at this list and see things like (“Alexa, what’s in the news?” or “Alexa, what year did [band] release [song or album]?”), you realize that AI already has implications for information seeking and retrieval, which will undoubtedly transfer into librarian services.

AI uses a combination of natural language processing, pattern recognition, expert systems and robotics. Many of these things are done really well in library systems. For example, the OPAC uses natural language processing to retrieve appropriate results in a search. Other attempts to use expert systems to replace the reference librarian have been made, like ORA--the Online Reference Assistant. ORA simulated the reference interview by asking the user a series of questions to try to understand more of what they needed, then searched video libraries, databases and modules to find it. Some libraries, like the De La Salle University Library in the Phillippines, are using online reference assistants effectively. De La Salle uses LORA, the Library Online Reference Assistant, that can answer ready-reference questions, request interlibrary loans and allow instructors to schedule an instruction session. You can try it here:  http://www.dlsu.edu.ph/library/asklora.asp.
Devices like LORA, Alexa and Siri are changing the way that people seek information. How many of you have elementary or middle school children who prefer to ask Siri for information rather than Google it? I know mine do. The next generation is growing up with these devices and they are only getting more sophisticated. An article in Public Libraries Online from earlier last year explained how developments in AI create opportunities for libraries to rethink library services and how people are changing the way they access information (http://publiclibrariesonline.org/2016/02/libraries-in-an-artificially-intelligent-world/). Can AI make any processes in libraries more efficient? Better for the user? Probably, but how that will happen is still evolving.

Copyright 2017 by Jenna Kammer.

About the author:

My name is Jenna Kammer, kammerj@missouri.edu, and I am the author of this column. I work as an instructional designer of elearning at the University of Missouri. I teach a class at the University of Missouri called Technology to Enhance Learning. I am also a doctoral candidate at the University of Missouri in the School for Information Science and Learning Technologies where I research information policy, and the impact of technology in society.