Communications of the ACM
Researchers Aim to Make Digital Assistants Like Siri Less Annoying
Researchers are working to refine digital assistants to make them more useful.
Credit: Reuters
Researchers continue to work on developing digital assistant technology that will help run people's lives more smoothly.
For Justine Cassell, an artificial intelligence (AI) researcher and associate dean at Carnegie Mellon University's School of Computer Science, the key is the technology's ability to read people's conversational clues--not just their words--and draw on an ongoing relationship for context.
The digital assistants from Apple, Microsoft, and Amazon always talk to people as if it is the first time they have met them, Cassell notes. Instead, she says they should remember what they recently searched for online and let them know the price has declined.
As an example, Cassell says digital assistants should be able to remember how much trouble someone got in last year for forgetting their significant other's birthday, suggest that they look up the item today, and buy it now, while the price is low. She notes such "socially aware" AI would respond more like a colleague or friend.
Cassell's team has researched human interaction and is building its findings into its "socially aware robot assistant."
From Public Radio International (MN)
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found