Why Siri Could Be the Killer App for Smartwatches

The question that no smartwatch has yet convincingly answered is: What should those gadgets really do?
Image may contain Wristwatch and Digital Watch
The real challenge with smartwatches will be figuring out what they're best suited for.Photo: Josh Valcarcel/WIRED

As we've said in this month's cover story, the wearable revolution is real and it's coming fast. For designers, though, putting computers on our bodies means tackling two very tricky questions. First, what should these gadgets look like? And second, what can they actually do for us?

We can make some good guesses about that first question. The current crop of wearables sorts out into two categories: things we wear on our arms and things we wear on our faces. Assuming we're not quite ready for the good cyborg life that Google Glass prescribes, that leaves us with the wrist.

>What can these gadgets actually do for us?

The idea of the smartwatch has long intrigued us: Just look at thepop culture supercut Samsung put out to promote the Galaxy Gear, its first foray into the field. In reality, reviewers found that the Gear tried to do far too much with its meager real estate (and did it far too clumsily).

The Pebble, an independent smartwatch, raised a staggering $10 million on Kickstarter last year, but it's hard to imagine its clunky looks and primitive UI getting much traction outside of the eager early adopter set. Companies like Jawbone and Nike have seen a measure of mainstream success with their fitness-centric bracelets, but what these devices have in style they lack in functionality. An activity monitor does not a smartwatch make.

But the continued interest in these devices—and, of course, the perennial intrigue surrounding the elusive iWatch—makes one thing clear: People are ready to wear a computer on their wrist. The question that none of the existing products yet answered, however, is this: What should a smartwatch really do?

When you consider where the existing crop of smartwatches have failed—trying to cram too much crap in too little space—you start to wonder if we haven't got the utility here backwards. What if instead of just supplying us with yet another screen, the true potential of the smartwatch lies in giving us an entirely new way to control our digital lives?

Beyond Notifications

The predominant use case driving the smartwatches we've seen thus far is basically notifications. It makes sense. Seeing incoming calls, texts, or emails on your watch, at a glance, means you don't have to pull your phone out of your pocket every time it buzzes. Over the course of a day, that could translate to a lot of time saved.

Combined with the right software, these devices could be useful for monitoring more than your existing feeds, streams, and inboxes. Many have pointed to predictive software like Google Now as an example of how other types of information could be packaged into bite-size, at-a-glance morsels. Google Now, paired with some sort of wrist-worn display, could bring sports scores, calendar alerts, and traffic updates to your person all without you having to ask for them. Again, useful.

.

But is a notification bangle the best we can do? And, more importantly, will it be enough to sustain an entirely new consumer electronics category? For a certain ultra-plugged-in, notification-obsessed demographic, it might be enough of a draw. But it's hard to see the widespread rise of wearables built on a bracelet that tells you when someone mentioned you on Twitter.

If wearables are truly going to be as big in the next decade as mobile was in the last, we should think beyond notifications and ask what these things might be able to do for us that our smartphones can't. That requires some hard thinking about the very big problem of input.

>Is a notification bangle the best we can do?

The Right Time for Voice

Smartphones brought about a true revolution in human-computer interaction, banishing cursors and mice and letting us touch the digital world directly with our fingertips. Many of the smartwatches we've seen thus far have basically tried to shrink that model down, put it on a wristband, and call it a day. That won't work.

No one's going to be tapping out text messages on these, for one thing. And it's hard to imagine that swiping between stripped-down apps on a miniscule touch-sensitive display is really the future either. And yet, if we want our smartwatches to be more than inbox armbands, we'll need some way of communicating with them. The obvious answer? Our voices.

On our phones, Siri exists outside and above the apps we're familiar with.

Image: Apple

If you haven't tried voice search lately—either Google's or Apple's Siri-flavored variety—you should. It's astoundingly good. Google trusts it enough that they made it the central method for operating Glass, and even on smartphone apps it's fast and accurate to a remarkable degree. But that's the thing: most people don't use it on smartphones

For one thing, it's dorky. In 2013, talking to a piece of hardware is still a largely ridiculous thing to do. And since we've gotten pretty nifty with our on-screen keyboards, the occasional inconvenience of using them is worth avoiding public ridicule.

Plus, in terms of input, our smartphones have always been text-first devices. That goes for search at large, too. Google has made some strides here, giving text search and voice search more equal footing in its apps, but the point remains. On our phones, we're comfortable searching like we always have, typing queries into clearly marked text boxes in specific applications.

Here, it's worth talking about Siri. Apple's personal assistant has failed to take root because it has never been a good personal assistant. From the outset, it didn't reliably do the things you asked it to, either because the servers it relied on were down or because it couldn't quite parse what you were asking for.

But Siri also suffers from the fact that it's cognitively at odds with how we've been trained to operate our smartphones. Instead of letting people navigate to their favorite apps and use their voices to control them, Siri arrived as its own isolated thing. It's a feature apart, not baked into existing applications or familiar workflows but existing totally separate from them. It's something you have to actively decide to use.

>Siri arrived as its own isolated thing. It is a feature apart.

On smartphones, that's a problem. We've grown to trust our apps. We know them. We like them! And the idea of letting this kinda-reliable digital assistant operate them for us just doesn't make sense when we can swipe over ourselves and be sure we were getting things done.

On a smartwatch, though, there isn't enough room to swipe anywhere. In fact, the smartwatch is the perfect home for a Siri-style assistant--one that can handle all sorts of requests, scooping up data from myriad sources and interfacing invisibly with other apps, services and websites to do our bidding.

Just like file systems were too messy for smartphones, it could well be the case that discrete applications are too unwieldy for the smartwatches of the future. For the last several years, both Apple and Google have been groping towards a future where we can get things done just by talking to our gadgets. The smartwatch might be the first time where that type of interaction makes more sense than anything else.

How It Could Work

The rub, of course, is that we're not quite there yet. In both Apple and Google's case, voice search requires a connection to the internet, which might be hard to do with a battery-starved wearable. And then there's the simple fact that dealing with Siri today isn't like delegating a task to a real-life assistant so much as asking a foreign acquaintance to do a favor over a stilted long-distance phone call.

Still, the newer version of Siri that shipped with iOS 7 is much improved, and it stands to reason that our ability to process natural language requests will only get better form here. In terms of device-to-device chatter, Apple recently filed a patent based entirely around the idea of using Siri to control one gadget from another--just the type of technology you'd need if you had a Siri-enabled smartwatch piggybacking off the smartphone in your pocket. So what would this all look like in a few years, if we're just having fun? Maybe something like this.

>So what would this all look like in a few years? Maybe something like this.

To start, this hypothetical smartwatch could have a feature where holding it to your mouth, Dick Tracy-style, activities it to listen. Think of it as a gestural, accelerometer-aided replacement for the "OK Glass" prompt that starts off interactions with Google's wearable. Wing the thing up to your face and it's ready to go. Then, you just talk.

Super Siri, perhaps bolstered in its language-parsing by the tech Apple recently bought from the start-up Cue for a reported $40 million, takes whatever you throw at it and makes it happen. Saying "Play Arcade Fire" could control the music app on your phone; saying "Text Dave, 'I'm on my way'" could send a quick message. These are things Siri can handle now—it's just that, today, why bother?

The Pebble smartwatch is based around at-a-glance notifications.

Photo: Ariel Zambelich/Wired

But in some cases, voice input could be vastly easier than our familiar smartphone workflows. For one, it significantly eliminates friction for creating reminders and notes. The best to-do list app is only as good as the things you actually log in it, and the process of opening an app and making a new entry for a simple, non-vital item like "buy milk" is often more trouble than it's worth.

But if you could hold your hand up to your mouth, say "remind me to buy milk when I'm walking home," and let your watch pass that reminder off to your phone to process, that's hugely powerful. All those fleeting ideas, observations, and to-dos that you have throughout the day—the ones that so often float away because you don't have a pen and paper at hand or don't want to deal with firing up a to-do app—are instantly archived and processed by a truly intelligent, watch-residing personal assistant.

Making appointments is also another powerful use case. A handful of calendar apps have incorporated natural language entry, using algorithms that can parse text like "lunch with Rachel at 9 on Tuesday" and make the corresponding calendar entry. But typing that sentence out is only marginally quicker than tapping over to Tuesday and putting "lunch with Rachel" there yourself. Being able to say "lunch with Rachel at 9 on Tuesday" aloud and have it register is something entirely different, and several degrees more efficient.

Still, a smart virtual assistant doesn't solve the problem of the smartwatch's limited output. You can only fit so large a screen on something people will want to wear. And even with a screen, you'll never want to scan through restaurant listings or read movie reviews on one of these things.

Yet, that fundamental just-talk-to-it ease could open new solutions for retrieving content, too. Even if a watch wouldn't be any good for reading the Wikipedia entry for "Jesse James" on your lunch break, what if, instead of dropping the matter altogether, you could utter "pull up the Wikipedia for Jesse James on my computer" and have it waiting for you when you got back from lunch?

***Of course, this is all fantasy. Asking Siri to do something today, when you can see exactly how she's messing things up, is scary enough. Maybe we'll never feel comfortable giving commands to a gadget without any sort of confirmation that it's getting them right. But if there ever was a device that was well-suited for a versatile virtual middleman, it's a smartwatch. On our phones, we're used to finding the right app for the job. And since we've got them in our hands so much of the time, there's little incentive for us to delegate those jobs out to any sort of assistant. If that assistant was just a tug of the arm away, though, it could be a much different story.