Siri Thoughts and Opportunities (or "We missed it by THAT much, Chief!")
I've had an iPhone 4S for about two weeks. I like it and I'd choose it as my mobile communicator and wearable computer again. But besides the hardware improvements that made the 4S a compelling replacement for my last cell phone, the prospect that excited me most was Siri. Pre-release, based on early reviews and my previous experience with the Siri app on earlier iPhones, I and many others predicted descriptors like "thrilling" and "game changing" to the synergy Apple created by embedding Siri into the guts of the phone's functionality. But I harbored the caveat that, while cool and useful, what we would experience with the 4S would still be very "first gen." I predicted (or guessed, if you prefer) that the iPhone 4S + iOS 5.0 + Siri would only presage the next UI paradigm rather than plop us happily into its voice-activated lap. Alas, I was more right than I wanted to be.
If you have an iPhone 4S, you've probably played with Siri, Apple's "Virtual Personal Assistant." If you don't use an iPhone, but you've heard the hype, you may be wondering how much you're missing out on. Here's my take: One day, tech like this may be your preferred means of interacting with and communicating through your smartphone. But for now, it's not, and here's why.
Mainstream adoption of any new technology is either retarded or liberated by its UI (user interface), the hardware from which it must derive its power and its usefulness, and accessibility. Supported by those three pillars, new tech moves from clunky to sexy en route to ubiquity. Take any one of them away and new tech falls over like a broken Weeble. It faces mainstream adoption chances lower than a naked mole rat in a Beverly Hills pet store. New tech gains mainstream adoption rapidly only if it's an order of magnitude easier to use, more convenient, or enabling than old tech. It helps if it's overwhelmingly cheaper. Or if it unlocks pent-up demand. Or advances gaming, or porn.
What I underestimated in my Siri prophesying was how unfortunately inaccessible Apple would make this slick technology, the continued limitations of the new hardware, and (hugely!) the dearth of accessories needed to unlock Siri's power. Those three technological gaps limit Siri's true usefulness and subsequently doom this version to naked-mole-rat levels of adoption. Though wonderfully easy to use, as long as the common-case interface with Siri is first via touch and THEN via voice, *and* as long as the latencies associated with invoking Siri are irritating or distracting it will be relegated to fringe adoption, trotted out mostly at social gatherings for "ooh's" and "aah's" more than for daily usage by Joe and Joan Consumer.
What will it take to close the gaps? When will Siri become your preferred method of interacting with your smart phone?
(1) When you get receive improved hand held hardware capabilities and network speeds. Response latency is a critical element in the function of any UI. Humans want speediness between action and reaction. Siri currently relies on both (a) handset processing power and (b) network access, because much of the language interpretation (and integration with remote systems), happens in an Apple server farm in California or North Carolina. Lack of handset processing power is why Siri won't work well on previous versions of the iPhone; the iPhone 4S's dual-core A5 chip is *just* fast enough to make the voice recognition software work well enough to be usable (and, make no mistake, the voice recognition on the iPhone 4S is stupifyingly good). But there's still a LOT of speech processing happening on the Siri servers, which takes old-school voice recognition and adds the magic that makes Siri so promising. That means in most cases the time between a spoken command and Siri's interpretation and handling take just long enough to be annoying once the novelty wears off. Rule One to the Theorem of Consumer Adoption is "Mainstream people tend to avoid things that annoy them." Siri works better if you have a local wireless access to the Internet rather than a cellular network's higher latency, but that difference makes Siri's response times inconsistent. Corollary One to Rule One is that Inconsistency Is Annoying. As chip speeds, memory, and battery technology continue to improve, Siri will do more on the handset, making the functionality speedier and more satisfying.
(2) When you see a stronger and more thoughtfully designed integration with the phone and 3rd party apps. This problem surprised me, because the entire Apple brand is built around elegant design and streamlined UI's. But they tubed it with Siri's implementation. The user shouldn't be required to change their method of interacting with their smartphone in mid-process. Changing the interaction method from touch to voice or vice versa means that one method or the other isn't trusted. That's a failure and failures are annoying (unless of course you're reading about them on a blog somewhere, in which case they're hilarious!). The 3rd party app integrations will come along; the APIs (application programming interfaces) for Siri have been published and are being extended rapidly. So the paradigm "If you can do it by touching, you can do it using Siri," shouldn't be far off. But for now, violations of that mandate are the rule rather than the exception. Consequently, some things you can do with Siri, and some things you can't. That violates Corollary One to Rule One. Speech should be able to invoke any UI element on any screen, with no touches, taps or swipes needed. Every violation of this UI tenet slows mainstream adoption because somewhere, someone will try to use some piece of functionality, have it fail, and return to their old ways.
(3) When some enterprising company reduces the dearth of Siri-enabling accessories. Here's where I see immediate business opportunity for an agile, forward-looking company (perhaps opportunity even accessible to a small firm! ). If you have to have your iPhone in your hand to really use Siri, the tech ain't finished or ready for prime time. Outside of accessories appealing to geeks or power users, the wearable computing and hands-free accessories landscape is ridiculously barren. One might envision the myriad square feet of peg space visible in every electronics store in the world and think that previous statement is crazy. But recall those displays and ask how many of those accessories are equipped with microphones *and* do not dominate or detract from the appearance of the person wearing them? Most people care about their presentation to others, but the market is flooded purely with accessories that only an engineer or a teenager could love! If Rule One to the Theorem of Consumer Adoption is "Don't annoy people." Then Rule Two is "Most people won't begin using something that makes them look or feel stupid." The average person does not want a piece of technology to glaringly dominate their presentation or appearance, no matter how functional it might be.
And herein lies opportunity. Because Siri is going to be a game changer. Voice is (finally!) on the cusp of becoming a highly adopted (if not full replacement) method of interacting with our electronics and dispatching messages and commands, especially in environments where visual/touch interaction is problematic (or dangerous!). But for this to happen accessories must be available that give you full, simple access to your smartphone (your wearable computer) and the information conveyed BY your smartphone, without access to the display. Widespread adoption of that kind of accessory requires meeting three market demands: (1) A hands-free connection to the phone, because wires are annoying an unsightly; (2) the accessory must be lightweight and long-lived, because you'll habitually put it on with your clothes and expect it to last all day and through the evening, despite constant usage; and (3) your accessories must be fashionable and/or surreptitious. The business person or female consumer is only going to buy and wear Siri-enabling accessories, all day every day, if those accessories complement the individual's personal presentation and don't fail when invoked (say, because the battery ran down). Accessories have to fit in with casual wear, business wear and even evening wear for both men and women. Without the availability of accessories meeting these requirements, Siri won't be convenient enough in an "anytime, anywhere" context. But with that kind of accessory and the ability to easily and surreptitiously invoke and use the organization and research power Siri can deliver, this new tech grants the holy grail of wearable computing: nearly infinite knowledge and capability in a way that's immediately accessible, no matter where you are or what you're doing. It makes you *smarter.* That kind of advantage that will indeed drive mainstream adoption...one day.
Posted by khiggins at October 29, 2011 05:06 PM