Apple’s hotly anticipated iPhone event didn’t offer the bold design refresh that many (ourselves included) expected. The iPhone 4S is an incremental upgrade in terms of hardware, full of fat new tech specs—an A5 processor, a souped-up camera, iOS 5—jammed into a package that looks exactly the same as the last one. But iOS chief Scott Forstall still saved the best for last: Siri, a voice-controlled "humble personal assistant" app that points toward the omega point of Apple’s quest to perfect human-computer interface design. (Apple acquired the technology in April 2010.) Put simply, the interface will be gone. There will be no interface: You will talk to your iThing, it will understand what you mean, and it will do it. It’s the logical endpoint for a company whose design philosophy is "it just works." Siri itself may not be powerful enough to embody this vision completely and flawlessly right now, but it’s the first step—the beginning of the end.
Obviously, voice commands have been a smartphone staple for a while now. I’m an Android user, and it never ceases to amaze me how Google’s voice recognition servers can parse the name of my local French restaurant, despite the fact that I mangle the pronunciation and there’s a wall of street noise behind me. But barking search terms—or even spoken commands—is different from talking. If Siri works as advertised, it represents a quantum leap in interface design because actual, y’know, interfaces—those sometimes sexy, often incomplete, always indirect middlemen between what we want and what we get—will inevitably seem clumsy and thick compared to the effortless, transparent, everyday miracle of speaking in natural language.
Every decent futurist-vision of man/machine interfaces—from the nameless ship’s computer in Star Trek to the murderous HAL in 2001 to the gentle Gertie in Moon—has converged on direct, natural language input as the logical endpoint for mainstream, non-specialty computing. (There will probably always be a need for sophisticated uber-interfaces to accomplish specialized tasks—think Tom Cruise in Minority Report—but for the near-future equivalent of what most people think of as "computing," there’s little that a natural language UI can’t do most of the heavy lifting for.) Computers, especially mobile computers, are becoming our outboard brains—and we already "designed" the ultimate interface between brains tens of thousands of years ago, when homo sapiens started talking to each other around the campfire. Short of direct neural implants becoming common (don’t hold your breath), it only makes sense that eventually our information-processing appliances would start conforming to our most powerful way of communicating. Siri is based on technology originally funded by DARPA, the military’s bleeding-edge mad science arm, and you know they don’t mess around.
So, does it really work? Or will Siri be this year’s version of AntennaGate? Natural language interfaces, by dint of the ultra-high bar they set in terms of their accuracy, are essentially pass/fail experiences for the average user: either they perfectly, invisibly work, or they’re a total bomb. (What good is a "natural" language UI if you can never tell when it’s going to "get it right" or not? Stabbing with your thumbs is faster.) With Siri, Apple has put its user-experience neck out in a way that makes the radical redesigns of OS X Lion feel like small potatoes. It’s saying, "We know this is the only place for user-friendly computing to go. We might as well get there first." Whether it succeeds or not is anyone’s guess, but the ambition and vision is classic Apple. Once again, the company appears to have done what it always does—take something that’s already existed and refine it into what it feels like it should be. And it looks like other smartphone makers had better start playing catch-up—again.
[Kevin Purdy will have a detailed review of Siri and the iPhone 4S in the morning on FastCompany.com, so stay tuned.]