There is a time and place for everything. In the privacy of my own home, I have no problem saying “Hey” to Google, Alexa, Siri, Meta and even Bixby on occasion. But what about in public? Where can others perceive me? I’d rather crawl under a rock.
This is one of the biggest issues I’ve encountered with artificial intelligence devices over the past few months. They all seem to believe that the best way to interact with an AI assistant is by actually doing it speech To them, it’s like a movie she. In fact, I rarely see my friends and family using mobile assistants when we hang out in private no way in public places. So it felt like a little “aha!” During last week’s WWDC keynote, Apple mentioned that iOS 18 would allow you to type with Siri.
Technically, you can already do this through the iPhone’s Accessibility settings. (Go to Accessibility > Siri > Type Siri.) This brings up a fairly basic window and keyboard for you to enter commands. Bring up the Siri keyboard at the bottom of the screen. You’ll also see quick suggestions that you can just click on instead of having to type (or speak) the entire query.
There are many reasons why this makes sense. While digital assistants have gotten better at understanding commands, it’s still difficult to converse with them naturally. At home, when I use the wake word, I feel like I influence a certain tone and tone. I find myself thinking in advance about how to phrase my query. Despite my reluctance, I occasionally mess up when I ask Google to dim the lights in my living room to 25 percent brightness. I would feel even more uncomfortable if I had to do this in public.
It was also very noisy outside. When testing the multi-mode AI capabilities of the Ray-Ban Meta smart glasses, I often had the AI tell me that the glasses couldn’t hear my voice properly. Either my environment was too noisy, or I was subconsciously embarrassed to the point where I spoke too quietly for the device to clearly pick up what I was saying. This frustrated me enough that I took out my phone—the exact opposite of what the AI hardware wanted me to do.
This isn’t just some newfangled artificial intelligence device, either. If you were James Bond, talking into a smartwatch would look cool. Most of us are not. If anything, most people I see doing this look a little confused and frustrated. Is this in vain? Yes. But self-consciousness is a big reason why people are hesitant to try voice-controlled assistants when out and about. A 2018 PwC survey on voice assistant usage found that 74% of consumers preferred using voice assistants at home, and participants said using voice assistants in public “seemed weird.” In the same survey, a lack of trust was cited as another major barrier to using voice assistants – people just don’t trust their assistants to understand commands correctly. If experience tells you that an AI assistant may not understand you, then why would you bother trying to use it in a place where you are more likely to be judged? (Also, imagine saying “Hey Siri” and firing up another commuter’s iPhone. A new nightmare unlocks.)
Beyond the technical logistics, typing to an AI assistant can provide you with a greater degree of privacy. I don’t need others to know what I’m doing on my phone, even if it’s something as innocuous as playing a song or setting a timer. I especially don’t want to dictate text out loud when others can hear my voice. Typing these types of queries allows me to keep things to myself—and I’m happy to sacrifice some hands-free functionality for that.
I’m not denying that you might have reason to do this need Talk to your assistant, even in public. Voice commands are especially helpful if you don’t have the use of your hands or are driving. But having multiple ways to interact with AI assistants allows them to more seamlessly adapt to how we want to use our gadgets, rather than forcing everyone to adopt a new paradigm. Maybe one day it won’t feel weird to talk loudly to a chatbot while walking down the street. For most people, that day is not today. Until such time comes, I’ll happily type for Siri.
