Through the years Apple seems to have developed a knack for figuring out the next technology sizzle. The Apple II. The Mac. iTunes and iPod and iTunes Music Store. iPhone. iPad. The Mac with point and click and the iPhone with the touchscreen are examples of major leaps forward in user interface design.
What’s next?
Some of us hoped that the next great leap in user interface would be Siri, Apple’s intelligent personal assistant-cum-knowledge navigator, circa the iPhone 4s in 2011. That hope remains but seems to have diminished. Is Apple neglecting Siri’s potential? Siri’s original founders have long since left Apple and launched their own Siri followup called Viv which seems to outdo what Siri does.
Apple’s technology competitors haven’t been standing still but their efforts at a voice controlled assistant haven’t exactly caught the world on fire, either. Google Now does some things Siri does not. So does Microsoft’s Cortana. There are others but so what? Nothing has caught the ear of the masses.
Why not?
Apparently, artificial intelligence linked to voice recognition is very difficult to do, so all the technology advances to date have been, well, incremental innovation instead of disruptive innovation. Despite Siri working on iPhone and iPad and Watch people don’t walk around talking to their phones unless they’re talking to someone besides Siri.
What’s the problem?
From my perspective as a Siri user, Apple watcher, and a bona fide certified technology fan-girl who likes to figure things out, there are two main issues. The first is the actual voice recognition, which seems to work well for people who have a limited accent, but not so well for other voices and dialects. The second is integration; how well the queries and commands– in this case, to Siri– are integrated into the operating system as actions. Part of the latter is the AI’s ability to understand context; something even humans have difficulty grasping at times.
Operating systems are complex and complicated and adding a voice element which attaches to commands and information queries is no mean task. That explains why Google Now, Siri, Cortana, and Viv remain in elementary school and little more than parlor tricks with modestly useful functionality. In other words, they’re good sizzle but not much steak. Yet.
Finally, there might be another reason why Siri and others have not taken over our lives. Such technology is expensive to create, maintain, integrate, and expand, and, there’s no other way to put this, there’s no money in it. Siri makes for a good bullet point on an ad, a great demonstration of voice recognition tied to actions, but it’s technology with more promise than delivery, despite some of Siri’s personality.
Voice controlled artificial intelligence might be the future interface but for now it’s merely interesting, occasionally useful, but more than a little oversold.
Peter says
Honestly, this is an example of one of the things I don’t like about Apple.
Way back in Apple’s “troubled” days, Apple used to design APIs for developers to access their latest and greatest innovations. A big part of WWDC was selling these APIs to third-parties.
Now, don’t get me wrong–Apple was a mess back then and I’m not sure I want to go back to those days. But that is one thing that I miss.
Apple used to have pretty good speech recognition on the Mac. Combine this with Apple Events to applications and you could create some very interesting text-to-speech applications.
Apple doesn’t do that anymore. Siri may be great, but it’s Apple only and nobody else can touch it. If I want to ask Siri when the next bus will arrive or where the closest clean bathroom is? Can’t be done–I have to launch the app and look.
Why is there no API for my app to be able to tell Siri that I can deliver this information?