Last week I attended a seminar in a nearby hotel You know how that works, right? Long tables face a podium. Attendees sit behind the table. In this seminar many of the attendees had Mac and PC notebooks, but a surprisingly large number– perhaps 35-percent– had an iPad or iPhone to take notes (or to check Mail and Messages during the talks). Something happened midway through the morning session.
Somewhat slowly, and hopefully not on purpose, one of the speakers said something coy, and then followed it up with, “Well, hey, s-e-r-i-o-u-s-l-ly (not spelled out, but spoken more slowly than normal. Immediately a bunch of iPads and iPhones went ding, ding, ding, ding, ding, ding, ding, ding as Siri on each one attempted to respond. Yes, Siri dings even when iPhone is on Mute.
One voice triggered about a dozen dings. What happened?
Siri does not know your voice. That’s wrong. Technologically speaking, there is no reason that Siri should activate to the wrong voice. Instead, Siri– like Amazon’s Alexa and Google’s stupidly named Assistant– just answers when called.
I remember some kind of setting on my iPhone where I was allowed to train Siri to my ‘Hey Siri…‘ voice but obviously that training doesn’t mean much. Siri has no clue who I am.
What if Siri could be trained to recognize our voice. Let’s say, you know, via voice recognition; technology which has been around for years. Siri has options on iOS, including an automatic Listen for ‘Hey, Siri…’ Or, for iPhone X, press the Side Button to activate Siri. There’s even an option to allow Siri to speak and be spoken to when iPhone is locked, and a few options beyond that, most related to specific applications.
What Siri does not know well is who we are by recognizing our voice. That’s wrong.
Since voice recognition technology has been around awhile, why doesn’t Apple have a setting whereby we can speak a few phrases, sentences, or even a paragraph or two, so Siri can analyze and utilize our specific voice and respond to that instead of just any old Joe Average who yells out, “Hey, seriously…?”
First up, such personal recognition would make Siri a bit more endearing so we may not worry as much about her shortcomings. Second, Siri should be set different on different devices. For example, Siri on my iPhone should listen for ‘Hey, Siri…’ only to my voice but be activated by anyone else with a click to the side button. However, Siri on the Apple TV remote should be activated to handle multiple voices. HomePod, too.
There really isn’t a reason why Siri could not be set to recognize half a dozen voices on different devices and respond according. For example, ‘Hey, Siri…‘ and Siri responds, ‘Yes, Dave, what can I do for you today.’ Yet, if my wife says ‘Hey, Siri…‘ to HomePod or Apple TV, Siri should know her voice and perhaps make recommendations for music or TV viewing based upon her profile, not mine.
These mostly stupid talking assistants might seem a bit smarter if they knew who we are when we ask a question, right?