As much as I enjoy having a virtual assistant, I find the need to say “Hey” or “Ok” before I address Siri or Google annoying.
Some Google devices have the option of being set to recognize you on sight, which means you can just start talking without using the trigger phrase “Okay, Google.” But activating Siri requires that oddly casual “Hey” to wake her up. (Google Assistant also responds to “Hey, Google.”)
Why? Because it’s insanely hard to get a device to respond to a two-syllable word coming from billions of people who speak different languages and dialects and may have some kind of accent. “Hey” and “okay” are pretty universal in how people say those words, so that makes them very useful was verbal triggers. (Alexa just responds to “Alexa,” which is nice. Microsoft’s Cortana assistance has responded to just “Cortana” since 2018. That’s the advantage of a three-syllable word.)
Apple feels our pain. There’s a new internal project that’s working on a way to eliminate the “hey.” It will require a significant amount of AI technology to work in milliseconds if this is going to be successful. If they figure this out, it will also speed up back-to-back requests. In short, it will create more of an opportunity to have a conversation with Siri instead of yelling “HEY!” at her before each question or statement.
Here’s a question: How will this work for people with various Apple devices? What if you want to talk to your watch and not your phone? Will there be a way to specifiy which device you want to address?
More info here.