Apple Adds 'Hey Siri' Voice Training in iOS 9
The new iPhone 6S and iPhone 6S Plus, which become available for pre-ordering tomorrow, add a feature for "voice training" when Siri is used with automatic activation, according to a report today in the Apple user site 9 to 5 Mac. Before that feature can be enabled, iOS 9 runs users through "a few training exercises," the site reported.
iOS 9 is scheduled to be released on September 16. The iPhone 6S and 6S Plus will hit the market on September 25.
Sensitive Siri
Apple said the latest update to its mobile operating system for iPhone, iPads and other mobile devices will bring new capabilities for Siri. The digital personal assistant has been redesigned to provide more intelligent searches and contextual reminders, even taking into account the time and a user's location.
"Siri can do more than ever and new proactive assistance helps you get more done before you ask, all while protecting users' privacy," said Craig Federighi, Apple's senior vice president of software engineering.
Recent discussions on Apple's Developer Forums have called attention to the increased sensitivity of the updated Siri. One user noted in June, "I was watching the keynote on my phone last night while plugged in. When they got to the Siri demonstration and said, 'Hey Siri,' my phone cut the video and started listening."
Users have also commented that Siri has responded unexpectedly by correcting them for swearing, and -- in one case -- activated after hearing the word "Syria" on a televised news broadcast in the background.
Context Is Important
We reached out to Amy Schade, who is UX director and mobile expert at the Fremont, California-based usability consulting firm Nielsen Norman Group, to learn more about what the possible improvements to iOS 9's voice control might mean.
"Voice control can be great for convenience, but the context of use is hugely important," Schade told us. For example, it can create user issues if employed in an environment that is too loud or too quiet, she said.
"If an environment is quiet enough to use voice successfully, is it too quiet to feel comfortable talking out loud to a machine?" she asked. "If an environment is too noisy or crowded, is it too difficult to relate information successfully via voice?"
Another aspect is how well a device actually responds to any particular voice command. "The quality of voice recognition is of course important -- whether and how well the device can understand the voice command at all," Schade said. "But another aspect is the feedback that is given back to the user -- how does the device respond, what information can it present to the user, and in what format?"