Categories

Revolutionary User Interfaces

A few years ago, around the middle of the last decade, the mobile phone market was characterized by the rivalry between a few established vendors. These were Nokia, Samsung, LG, Motorola and Sony Ericsson. These incumbent companies had a broad portfolio of devices including smartphones and feature phones and basic phones. Many also sold networking equipment and were deeply engaged with their customers, network operators.

There was also a set of entrants who offered only smartphones.  They were quirky. HTC was a a prominent “ODM” or original design manufacturer who built phones for companies who added their brands and sold and supported the product. HTC made phones and PDAs for operator brands and for some large PC companies. It also began to sell phones under its own brand. RIM was also offering products that had evolved from pagers into email appliances with added voice capabilities. But RIM’s products were not very good as phones. Voice was so poorly integrated that many people carried both a BlackBerry and a voice phone. Then there was Palm with something called a Treo which promised many things but did not quite deliver.

In 2007 something happened which changed the industry. It took a few years to even realize it was happening but by the time it was obvious, it had changed to such a degree that huge companies found themselves in financial distress. This chart illustrates the effect.

In a few short years the rivalry between incumbent handset makers has evolved into a rivalry between mobile platforms where the incumbents were caught in an increasingly unprofitable role of hardware integrators. Motorola is now being absorbed by a search company. Sony Ericsson is being absorbed into a consumer electronics company. Nokia is going through a potentially existential crisis and even abandoned its platform. Acquisition talk surrounds LG’s phone efforts. Samsung is prosperous but sill at the mercy of its software suppliers. Even the entrants are starting to feel beleaguered. RIM, the oldest pure play smartphone company is trading below book value and has regressed in value to 2004 levels. Palm evaporated. A view into the changes is the rankings of the companies in terms of volumes, sales and profit as shown below.

How did large, powerful, smart companies let this happen?

It happened despite having a clear, front row view of the transition of the industry from mobile voice to mobile computing. The shift in the basis of competition from “connecting people” to “connecting people to data” ended up being a classic disruptive trap. Many will argue that it was the failing of individual managers. Perhaps, but how did they conspire to fail simultaneously?

When you see disruption happening, it’s natural to seek out a cause, a pivotal magical “force” or event that enabled the weak to humble the strong–the proverbial sling that enabled David to defeat Goliath.

My hypothesis is that The Primary Cause for the shift of profits from Incumbents to Entrants has been the disruptive impact of a new input method. I illustrated this with the help of the January 2007 slide presented by Steve Jobs at the launch of iPhone.

Each new input method led to not just a disruption but new platforms and new business models. Each new “sling” victimized a set of historic companies. Mainframe and minicomputers, consumer electronics giants and now telecoms.

Just like David’s sling, these technologies are not powerful in and of themselves, but rather, the way they are used makes them unpredictably sinister. The context of using capacitive touch on a handheld device rather than on a table-top makes it disruptive. Coupling it with high-speed mobile networks and powerful but efficient microprocessors made it into a force.

That is now ancient history. The consequences and repercussions are still being felt and weighed, but the smart money should be focusing on the next shift. From the time frame in the diagram above, it’s clear that the cycle time between “Revolutionary User Interfaces” is shrinking. It’s been five years since multi-touch. Is the next “RUI” already here? Is Siri the next RUI?

There are many things going for it:

  1. It’s not good enough
  2. There are many smart people who are disappointed by it
  3. Competitors are dismissive
  4. It does not need a traditional, expensive smartphone to run but it uses a combination of local and cloud computing to solve the user’s problem.
  5. It is, in a word, asymmetric.
My disruptive hypothesis for Siri is that it shifts the competition from platforms positioned on a device to a “coupled” super-platform dependent on broadband and infrastructural computing.
Just after collecting enough data and observing patterns in it that give us clarity, It looks like things are about to change all over again.

 

  • http://etn.se Jan Tångring

    There was more to Apple’s entry. For example, there was/is a cooperative relation between operators and suppliers of phones and suppliers of networks. This cooperation hindered development and hurt customers. Apple ignored that and formed a direct relation between itself and cellhphone users – who used to be considered the property of operators.

    • Anonymous

      Same reason Microsoft and Google can’t make good products. They built straightjackets out of dogma. Good products were not the priority.

      • Ab

        Interesting you care to expand? I’ve got kinect and its a good product. What products are you referring to in relation to Google. Sure MS has made some dud products the Kin comes to mind.

      • http://riverlaw.myopenid.com/ riverlaw

        google wave, the new reader release, the gmail mail app for iPhone, google music, Picasso.

      • kevin

        The dogma is to protect and extend Windows and Office for MS, and to protect and extend search/advertising for Google.

        Xbox and Kinect and Zune and Kin broke beyond Windows; they are the exceptions; and there were mixed results. But note Courier was allegedly killed for going outside, and many of those that led these exception-efforts have since left MS.

        Android and Chrome and Google+ are there to serve the ad master. Wonder if the automated car needs to serve it too.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        Yeah, because billboards should be INSIDE the car. I can hardly wait for HUD Adwords and the car’s “voice” telling me that there’s an Arby’s at the next exit.

      • http://twitter.com/qka qka

        Kinnect is the exception that proves the rule at Microsoft. They’ve had success with the XBox line, but at what cost? Over the life of XBox, have they recouped all their startup losses?

        Google is an advertising company. Their interest is in collecting information on individuals to make the advertising more targeted. People are coming to realize and don’t like it.

      • Anonymous

        And yet now Apple is doing targeted advertising as well.

        http://advertising.apple.com/brands/

        “Each ad is shown only to the audience you want to reach, in the apps they love and use the most. Our highly-effective targeting leverages unique interest and preference data that taps into user passions that are relevant for your brand.”

        Wonder where they get that data?

    • http://pulse.yahoo.com/_KEVQDCIMSOI6AFXXFFS5ZFFWBA PeterK

      Totally agree.

      Apple won the battle for ‘open’ innovation in the teleuniverse.

      Apple’s ‘closed’ begat ‘open’, and Jobs said, “It is good.”

      Now, we have the Android ‘Openistas’ screaming ‘closed’ as a reason to avoid Apple!

      They’re blind to Apple’s mission, which can be described as, “Bringing open and free radical innovation to the INDIVIDUAL in defiance of corporate ‘orifices’ everywhere”.

  • http://www.gerger.co Yalim K. Gerger

    Spot on article. We are at the beginning of another UI revolution. For what it is worth, I shared my thoughts about Siri and how it will change everything at http://gerger.co/yalimslodge/2011/10/13/siri-artificial-intelligence-revolution-has-begun/ .

    • http://twitter.com/dicklacara dicklacara

      Read your blog… Agree 100%.

      BTW, the answer to the poster’s question: How is Siri different than ELIZA (apologies in advance for the caps):

      1) Siri is LIMITED

      2) Siri DOES A FEW THINGS WELL

      3) and, as Horace says: Siri ISN’T GOOD ENOUGH

      • http://www.gerger.co Yalim K. Gerger

        You are right. The difference between ELIZA and Siri is so obvious. Siri is a UI. ELIZA is not. ELIZA is at best an interesting academic experiment.

        Siri picked its battles, and the ones it fought, it fought really really well. This trend will only continue. As Siri gets stronger, it will take on more powerful enemies.

        A new era has begun. Soon we will see every player join this race. If I were a VC, I’d be all over this trend.

      • Anonymous

        Siri has a male voice in Europe.

      • http://www.gerger.co Yalim K. Gerger

        True. Thanks for pointing that out.

      • Jzlatic

        Here’s a continuing theme in recent posts by Horace that should strike fear in Apple’s competitors: The cost of doing business, of competing with Apple, is going SKY high. The infrastructure required to produce devices priced against the economies of scale that Apple enjoys as well as the infrastructure required to do the Cloud, Siri, etc. coupled with patent hurdles is creating a very formidable nut to crack.
        Apples biggest threat in the near future will be the anti-trust dogs.

      • OpenMinde

        Not really. Dropbox relies on Amazon S3 cloud service for its storage solution. The only high wall is manufacture, in my view.

      • http://www.gerger.co Yalim K. Gerger

        I think if anything, this post by Horace underestimates Siri. I think Siri will be the dawn of a new technological revolution, not just a UI revolution. This is far far bigger than the mouse. This is as big as the Internet. I think we are in the 6th technological revolution.

    • Jzlatic

      Yalim, I concur with your posts, great job! I wrote similar prognostications when Siri was just a rumor, before I comprehended the significance of the AI aspect. My basis for crowning Siri a game changer several months back: each new UI advance before it had been a game changer. Also, I see few pointing out that each UI advance until now relied on human hands to communicate with the computer (punch cards, keyboard, mouse, touch). So, now, for the first time, we are moving beyond our hands as the only input tool… Now we have a new UI AND a new input tool: voice/natural language. Throw in self-teaching AI and… WOW!!! As you and I both noted, this is Steve Job’s parting gift and crowning glory.
      I saw doubters leaving messages on your board… They remind me of the Wright Brothers saying, “When we invented the flying machine we thought we had created a device that would make future wars impossible.” Or IBM’s President Tom Watson who, at one point in the early days of computers said: “The total worldwide demand for computers will be three.”
      As some people are good at math and some at English, we all have brains good for some things and not as talented in others. Some people like you and Horace have a gift for understanding the significance of disruptive technologies; others, not so much.
      As the Chinese proverb goes… We are all blessed to be born in such interesting times.

      • http://www.gerger.co Yalim K. Gerger

        Jzlatic, thank you for your kind words. The amount of negative comments kind of surprising to me too. Given that most visitors came from Hacker News, I’d expect a more positive reaction. To be completly honest, my biggest fear was that the article would receive a “Duh!” reaction. :-). As seen in the comments, that was so not the case.

        Finally, I cannot help but share this video made by Apple…in 1988. The vision was always there. Now they are executing.

      • Anonymous

        Somewhere in my garage I have a video (VCR) of that. Absolutely incredible!

      • GeorgeS

        In truth, that video showed what was wrong with Apple at the time. (That was Apple sans Jobs.) See John Gruber’s article at:

        http://daringfireball.net/2011/11/companies_that_publish_concept_videos

    • Christian Heidarson

      I don’t think it is accurate to call Siri AI, at least not at the level that you are talking of. When you ask Siri whether you should bring an umbrella, there is not an AI in the background relating concepts of clothing, comfort and raindrops. There is a pattern matching mechanism in play here, and an unprecedented amount of polish. But it is NOT a step-function in AI.

      I do thing that you are correct in your blog post in identifying this as an enabling step, allowing non-techies to do more with their phones. However, I don’t think AI is the most helpful label to put on that step. I’d suggest voice as a placeholder for now, but within two years we’ll probably be digging up ‘pervasive computing’ form the hype-jargon again.

      • jawbroken

        You are confusing artificial intelligence and “strong” artificial intelligence.

      • Christian Heidarson

        Siri can of course be called AI, as can the system that blocks my email spam.
        So to clarify: I was referring to Mr. Gerger’s use of AI in his blog posting, which implied that Siri represents a stronger form of AI than what we’ve seen before and an indication of a trend in AI strength.

      • jawbroken

        I don’t see anything in his article that particularly implies strong AI, or says that it is a “stronger form of AI than what we’ve seen before”. I can’t speak to the author’s particular intent but the suggestion in the article is that it will evolve as an assistant by being programmed to perform more tasks, not as the sort of peer intelligence that strong AI suggests.

      • http://www.gerger.co Yalim K. Gerger

        Hi Christian. No one is claiming that Siri is Data from Star Trek.
        However, Siri was recognized as an A.I. startup. Its roots are in A.I. research. Steve Jobs himself viewed Siri as A.I. So I think it is fair to call it A.I.

      • GeorgeS

        I’ve noted, over 40 years, that critics of AI just redefine the term anytime something gets too close. (The same applies to people rederining what it means to be “human” as science finds out more about animals behavior, emotions and cognition.)

        PEOPLE are “pattern matching mechanisms,” in large part. That’s why a computer can actually compete with and beat a top chess master, for example. People also go through stages of cognitive development and ability. By your example, a 2-year-old would not be intelligent.

        Alan Turing gave what should be kept as the definition of artificial intelligence–the “Turing machine.” His criterion was that a person interacting with the system would not be able to tell if it were a machine or another person. The interaction could be limited in method (he suggested a teletype, which was the technology of the time) and in scope. By that definition, for a set of topics, Siri comes very close to the definition of “intelligent.”

      • GeorgeS

        For what it’s worth, I used ELIZA in 1965-1966 at MIT.

  • Anonymous

    “It does not need a traditional, expensive smartphone to run but it uses a combination of local and cloud computing”

    For the moment, Siri only runs on Apple’s flagship phone, the 4S. Not even the iPad2 runs it. This might have more to do with the microphone and wanting a great end-user experience than a question of needing more computing power, however.

    I would be interested in seeing educated guesses on when Siri-type applications could run entirely on a smartphone. Might Apple’s massive investments in infrastructure become obsolete a few years from now? Is Apple being short-sighted by only offering Siri on the 4S when it should be getting everyone hooked on the wonders of its new interface?

    • SpeechGuys

      Doubtbul, The reason SIRI is is Beta is to tweak the interface by collecting lots of Use Case data and iterate the experience for 4s users.
      To have every Apple device using it might burden this collection and add another device variable to understanding UX.
      This is just a step to getting SIRI on the other leading-edge platforms, as you can run SIRI on lesser hardware effectively but makes little sense at this time.
      However, as Horace ( and I have said for years ), Voice Recognition will be most effective on lesser platforms without keyboards and fancy capabilities… and a network connection.

      Oh yeah, Don’t forget the TV and Microwave.

      • Ab

        I don’t think anyone can predict with reliability where voice recognition will be most effective. The best way to think about these interaction technologies is to see how they remove barriers between the user and the computer / device.

        In the future if cars can drive automatically we remove the barrier to learn how to drive. Is driving a skill we need? The same goes with the phone UI how long before we don’t need the skill to type?

      • Anonymous

        As in my reply to Walt French, I think this sounds quite reasonable. But what does it take for this to disrupt the competition? If voice recognition needs to be on the server for the foreseeable future, then I guess it comes down to Google’s reaction and what that might mean for Android in general. But if smartphones of 2015 can run Siri locally, perhaps no disruption will occur.

    • Walt French

      Remember that Siri is labeled “beta.” Which it is. It only works with about 16 built-in apps; more scheduled for 2012. Still no APIs to 3rd party apps; many built-in functions not linked. (“Siri, tweet ‘this is my first siri tweet’.”)

      I see it selling an incremental few tens of millions of iPhones in its very unpolished state and of course Apple can drop it into iPads via a software update just about whenever it wants. I don’t think of many millions of iPhones (where every 3 million phones is a billion of profit) as terribly short-sighted. Remember, they are not in the business of selling the service, it’s disruptive because it’s bundled with the hardware. Google’s business model, right in their face.

      • Anonymous

        I think that sounds reasonable, assuming that Apple plans on rolling out Siri to lower-end phones.. Still, I think we can only really call the service disruptive if a few years down the road we see that the mobile landscape looks significantly different. Assuming that voice functionality is here to stay, I still think Android (with Google’s back-end expertise) would be well placed to react. Should that not be the case, then I again would be interested in knowing when smartphones could be capable enough to run Siri-like applications without an internet connection. How many years lead might Apple need to disrupt the competition?

      • http://twitter.com/dicklacara dicklacara

        No hard facts here — just some suppositions and guesses.

        Siri may be limited to iPhone 4S models connected to Apple servers in order to:

        1) control the initial experience by limiting the server load

        2) gather information on what Siri does well, what Siri does poorly, and what the user asks Sir to do that she can’t do at all

        3) refine the voice recognition to improve the process (when the user has repeat or reword a request)

        4) tailor the voice recognition to the user’s voice, accent, slang, etc.

        5) form voice patterns (voiceprints ?) of common requests from the user

        6) As the voice recognition gets refined, tailored and as Siri gets used to you — many of your requests can, likely, be setup as a filter — quick test

        7) Then, Apple’s servers can verify this new filter by processing your requests in two parallel threads: normal Speech To Text and the quick filter

        8) At some point,as the requesting devices become more powerful — the validated “filter” test can be pushed down to the requesting device.

        9) When the filter on the device can’t decode your voice — it just fails through and passes the request to the server to be processed there.

        A similar information gathering, learning, refining, customizing process can be performed for the Siri’s parsing and AI analysis of the words comprising the request.

        Likely, Siri will find that many of the requests a user makes timers, reminders, texts, phone calls are limited to a few repeatedly-used patterns — and these too can e pushed down and performed entirely on the device (with no need for connection)

        Now our device knows our voice and usage patterns — If we’re lucky, some high percentage of our requests can be handled on the device, itself.

        The others, will just fail through the filters of voice patterns and usage patterns — and be sent to the servers.

        Or, an always-on connection to the cloud (cell, WiFi, etc.) may just become a necessity of our life style.

      • Walt French

        I think Google is the ONLY competitor that is positioned to react effectively to Siri. I think this is the knockout punch for Microsoft (!) and BlackBerry (duh).

        I was recently recalling the fun I had with my first computer, a Heathkit with 256 bytes of RAM — one two-millionth of the working memory of my iPhone. My SWAG is that Apple only needs a couple more ticks of Gordon Moore’s clock before Siri can migrate mostly into the phone.

        If it sees fit. We ARE becoming ever more connected. Siri requests only cost about a half penny of data at a likely 3G capped rate. Just as Amazon invisibly bundles in WhisperSync over Verizon into the Kindles, Apple could buy that data in bulk from somebody like Verizon, GoGoInflight and others to further extend their ubiquity.

      • Anonymous

        That’s an interesting idea about WhisperSync and buying data in bulk, even if I don’t see Apple ever going that route. They were the first (in my hazy memory) to oblige customers to have a data plan and thereby really unleashed mobile internet. I cannot even remember when they allowed an option to turn off mobile data, but it seemed like forever to me.

        If you are right about the power of tomorrow’s smartphones, then Siri might not disrupt the industry, even if it does signal an important UI advance.

      • Walt French

        I think the disruption to the current tap-oriented interface is independent of how Apple or others choose to implement Siri and its competitors. It is based on the ease-of-use of the voice control interface, not Apple’s particular implementation

      • Walt French

        I think the disruption to the current tap-oriented interface is independent of how Apple or others choose to implement Siri and its competitors. It is based on the ease-of-use of the voice control interface, not Apple’s particular implementation

      • davel

        Very nice point. I think Apple is looking forward to the day it does not have to deal with the telcos.

      • GeorgeS

        Minor correction: Amazon’s WhisperSync in the US is through ATT, not Verizon.

      • Anonymous

        I think we have a ways to go. Look at IBM’s Watson AI as a reference. It is a different approach than Siri, as its objectives are very different. Rather than using public APIs for queries as an assistant, Watson seeks to answer questions directly from a self-contained data set (On Jeopardy!, it was not doing online search for results).

        IBM has spent untold millions of dollars and years of development to get a functioning product, and it fills an entire server room. We have a long way to go before Siri can be taken offline. Also, many of the valet services it provides are dynamic. For instance, restaurant reservations are based on current availability. I think we will see network ubiquity before we see offline AI in our pockets.

      • Anonymous

        You make a pretty good case! The race between network ubiquity and offline AI will be an interesting one to watch. I am torn thinking about what the Apple/Google/Amazon server farms really mean.

      • Davel

        The problem with google is they don’t get it.

        I read somewhere that apple hires very good people, but these same people probably would not be hired by google. The resumes are different. Apple has always merged technology with humanities. In Steve Jobs’s address to stanford he mentioned calligraphy as being part of the soul of the original Mac. This one example clearly illustrates the difference between apple and almost every other tech company. They hire psychologists and linguists to work at a tech company because they value what they bring to the table.

        A computer with very few exceptions does not think and interact like a person. Google thinks like a computer.

      • http://riverlaw.myopenid.com/ riverlaw

        hi Walt I think you are wrong about the iPad siri software update. I think it needs the two microphones the phone has.

      • Walt French

        Of course, Sirii would USE the existing iPhone 4S hardware. That’s not the same as saying that it wouldn’t work well on the iPad. When we use Skype on the iPad, the audio is very clear. That may be because we use the iPad at home, away from street noise, etc., that’d be more likely for a phone.

        And I note having seen that the iPhone4S lacks an audio processor chip that was present in the 4; apparently the dual CPU now manages noise.

        (I’m referencing the iPad2. Yes, a model without a microphone wouldn’t work at all well.)

      • Foo

        Siri can work with the earbud mic, so no, dual microphones aren’t a requirement.

      • GeorgeS

        Siri can also use the Apple headset.

    • Anonymous

      It will likely always need the network, because it relies on massive databases that will only get more massive in the future as Siri does more. And each user only needs a very tiny piece of that database each day.

      Siri is a classic client-server app, like a Web browser. You don’t want to run the Web locally, you only need a very tiny piece of it today, and you want it fresh. Only the part you always need is local.

      • Anonymous

        I’d still like to hear some concrete numbers concerning processing and storage requirements.

      • Anonymous

        Not exactly what you’re asking, but here’s a study that goes in the direction of your question.

        http://arstechnica.com/apple/news/2011/11/how-data-heavy-is-siri-on-an-iphone-4s-ars-investigates.ars?utm_source=rss&utm_medium=rss&utm_campaign=rss

      • Anonymous

        Thanks. In particular, I’d like to know what the backend requirements are, mostly just to be able to hazard a guess as to when that sort of power is likely to make its way to smartphones. If we are talking five years or more, and Siri proves to be wildly popular, then only Google would be able to provide something similar (perhaps Amazon too). Would they? If so, would they for all Android branches? Might Google do so but only for its recently bought Motorola division?

      • Walt French

        I think that Siri could be implemented today in hardware not much more powerful than today’s iPhones. It’d be a bitch to manage, to distribute the knowledge database, etc.

        I wager that it won’t take Google more than a couple of years to come up with something roughly comparable to today’s Siri. Note that they recently acquired the airlines’ data service; they must be working assiduously on interfaces for it, for Zagat, and other properties of theirs.

        You could even claim that Siri is a pre-emptive defensive move by Apple.

        Yes, as neat and unique as it is.

      • Anonymous

        So do you think Siri will prove disruptive (and if so, to whom) or is this more of an advance that all/most current market participants will eventually copy in one way or another?

      • Walt French

        I believe that competitors have 18-30 months to build a credible work-alike to Siri, and that Google will likely succeed. This knocks RIM out; it is already struggling to catch up with the 2007 paradigms. Microsoft has a nice I/O and the integration that Horace mentions in today’s (Friday’s) post, so it might be able to hang in despite mounting costs to spread over a small share. (But they ARE prodigious s/w types!)

      • russell

        Walt,
        but isn’t that 18-30 months enough time for Apple to cripple/dismantle Google’s cash cow and potentially pass/match them in smartphone market share? Just mulling over the characteristic of time within a competitive advantage period Apple might have w/Siri.

      • Walt French

        Google having a cash cow in mobile is a myth. By my crude estimates, they’ve spent around $15 billion to date, with perhaps $1 billion of revenues that can be tied back to Android. (I’ve read that Google totally passes on its 30% of app revs to the carriers.) The old McKinsey schema would label that a “problem child” — fast growing but eating you out of house & home.

        I would call it a tiger by the tail. The world is moving to mobile and Google is challenged how to monetize it as effectively as they have with desktop. I don’t see what they’ll do, but I kinda think they’ll figure a way.

        A lot can happen in two years. Look back to the end of 2009 (when, e.g., Android Eclair 2.0 was brand new) and tell me that you could’ve foreseen much of today’s landscape. I will guess that Google comes up with a pretty good workalike in time to maintain their positioning as the hardcore users’ preferred OS. They might even bring out a more general inter-app communication that lets voice control the whole range of 3rd party apps, much as Win8 has some inter-op tools, and figure out how to keep it from being a resource hog that it could be.

        Don’t misunderstand my sense that Siri is awesome and the foundation for a whole new class of devices, and my hope that Apple is busting its chops to make it into a replacement for the world’s featurephones. But they don’t have the territory to themselves.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        I think you mean that you hope Apple is busting ass to make Siri a UI replacement (making a great effort). Busting chops is what I’m doing to you now. :)

      • Walt French

        chops |CHäps|
        plural noun informal
        1 a person’s or animal’s mouth or jaws: a smack in the chops.
        • a person’s cheeks; jowls.
        2 the technical skill of a musician, esp. one who plays jazz: when I’m on tour, my chops go down.
        PHRASES
        bust one’s chops informal exert oneself.
        bust someone’s chops informal nag or criticize someone.

        Looks like one of those auto-antonyms, like “cleave.”

      • Anonymous

        I don’t think 18-30 months is enough time because the AI part of Siri was started in 2003, and there were 300 researchers working on it. It began at SRI, and then in 2007 Siri was spun-off but they have an exclusive license to the AI technology (called CALO at SRI). It won’t be easy for Apple’s competitors to replicate a work-alike.

        1. http://www.sri.com/about/siri.html
        2. http://www.ai.sri.com/project/CALO
        3. http://en.wikipedia.org/wiki/CALO
        4. http://www.cbsnews.com/8301-505124_162-33447445/apples-acquisition-of-siri-hints-at-voice-activated-iphone-strategy/?tag=bnetdomain

      • Walt French

        Thanks for the details here.

        I wasn’t saying that workalikes would have the breadth of Siri. But dictation software is already in widespread use, and more structured commands would be implementable in much less time, perhaps even with a system-wide interface where apps could define what capabilities they had or grammar that they’d accept.

        From the very little that I’ve personally seen and read about, I DO believe that Siri represents some sort of a critical mass of responsiveness that is distinctive from what others will do. But that’s not to say that it’s the only such functionality possible.

    • Z Kariv

      A great article..
      not only of by itself but to guide you, the reader, toward the next step!!
      So, you should assume that Siri (and her next generations) are coming to all I..’s
      The question(s) is what beyond and how you can contribute/prepare to it

    • Just Iain

      I suspect the restriction to the 4S is intentional as releasing it to everything that could handle iOS5 would be truly daunting to say the least.

  • Anonymous

    You’ve nailed it, Horace!

    I’ve earlier tried to explain Apple’s disruption with a serie on the touch-cloud-mobile triple, but it became to broad. The picture of Jobs showing UI disruption just made it for me! It clicked!

    The post is worthy to be a classic. It sums up a lot of your posts and podcasts, and the way you use the graphics set the Apple profits from mobiles in a right context of quantity and rating.

    I cannot cite from this jewel of a post on my blog, It have to be read entirely. I’m still breathless….

  • http://twitter.com/dicklacara dicklacara

    Horace,

    Interesting article…

    Especially the part about what Siri has going for it: 1. It’s not good enough…

    I was fortunate enough to have a foot in both camps, mainframe and microcomputers, in the years between 1960-1990,

    As such, i observed (and even participated in) some of the disruptive technologies of the period.

    I would like to suggest some other notable revolutionary UIs that contributed to, and provide the infrastructure for what we have today,

    As with any disruption the actual events and the date are debatable…. Here’s my take:

    ~1977 the computer becomes the interface — the keybooard + display + microcomputer replaces mainframes with terminals and punched cards

    ~1979 the application becomes the interface — VisiCalc ushers in the era where 1 or more < $100 applications justify $2,000-$10,00 worth of hardware

    ~1981 sharing of information becomes the interface — The LAN allows local information sharing in the office, classroom…

    ~1994 information access becomes the interface — the world wide web (and underlying Internet) provides global access to, and sharing of information

    ~1995 mobility becomes the interface — pagers, PDAs and cell phones provide on-demand direct, and indirect (synched) access to information

    ~1999 connectivity becomes the interface — We've got the whole world in our hands… wherever, whenever…

    As I think back, one of the most significant revolutions was VisiCalc — the application became the interface.

    We are seeing that repeated today on the iPad,

    I believe that Siri is the beginning of the era where the assistant becomes the interface.

    I can't help but wonder what comes after Siri…

    • Anonymous

      The pods from The Matrix.

    • Davel

      Very nice list. I would add databases to your list.

      Databases changed the way companies dealt with data. We went from flat files to relational databases which allowed much greater flexibilty in accessing information.

      • Anonymous

        Yes I used VisiCalc as the “picture boy” of the era where the application drove the hardware sale… I could have used a WP or database program, but VisiCalc was unique — it was the first and only SS for a year or so (UNTIL 1-2-3). The customers came in to buy VisiCalc and whatever hardware (Apple ][, floppy drives, Monitor, etc.) needed to run it.

        I remember demoing a Pascal db app running on an Apple ][ to an IT type. After the demo, he was quite impressed.

        He asked how much the app cost.

        I answered $150.

        He said: “Only $150 per month?”

        I responded: No, $150 total.

        This was before relationals and the most popular flat-file DBDC system on mainframes was IBM CICS — costing a minimum of $3,000* per month.

        I saw some sales figures and IBM had over 35,000 CICS installations… you do the math.

        * to be fair, the customer got a whole lot of support for this amount.

      • davel

        Yes. Visicalc was very valuable. But Visicalc is also local.

        A modern RDBMS is more powerful. Multiuser, multitasking and if built right distributed. All modern companies depend on them. They are revolutionary because of the flexibility the data organization allows a very flexible way to access information in a way that flat files and spreadsheets just cannot.

        Like I said I like your list.

  • Ab

    I’m not so sure.

    At first we had simple phones and “clever networks” doing all the management and switching. As cpu’s got smaller it was ineviatble they would make their way into phones. The progress made by networks was always going to be slower than progress made at the periphery.

    Now we he phones being refreshed on an annual basis but you don’t see networks upgraded that fast and thats why currently the innovation will remain at the periphery. However with Siri which relies on backend services which do not evolve / upgrade as fast as services / hardware on the periphery are you saying the next disruption is going to be moving back to the core?

    • Anonymous

      Networks are updated daily. They are constantly growing. Nodes are being replaced with faster nodes, daily. There is a new kind of Wi-Fi with an extra antenna that acts as traffic cop and speeds everything up in the latest MacBook Pros. Before that, there was Wi-Fi n, g, b, a … it keeps evolving and getting faster even though we still call it Wi-Fi. Every single iPhone release ran on a new, faster network that the previous model could not run on, and that is expected to be true of the next model also.

    • http://www.asymco.com Horace Dediu

      Intelligence migrates to and from the edge depending on the prevailing problem that needs to be solved. Google benefits from a central consolidation of knowledge and the power needed to process that knowledge. Similarly, a voice interface may need the migration of focus from the edge to a combined model.

  • Ab

    Horace on the one hand you say how did all the incumbents fail to see the threat of the Iphone:

    Perhaps, but how did they conspire to fail simultaneously?

    But then you go onto say this failing to see the threat of Siri will again be repeated.

    2.There are many smart people who are disappointed by it
    3.Competitors are dismissive

    So you expect history to repeat itself? If that’s the case then have you not answered your first question with your points re Siri?

    • Walt French

      There’s a video now out of Christensen presenting to Gartner a few years ago in which he walks the audience thru how mini-mills disrupted the US steel industry. Repeats (for effect) a couple times about how the majors willingly ceded the low-quality territory, and that it was, in their eyes, the smart thing to do.

      You could imagine the audience nodding, “yes, we want nothing to do with that low-quality, low-margin re-bar business.”

      Worth watching for his other points.

      I’ve only heard allusions to Christensen having directly addressed Apple’s seeming ability to disrupt with MORE expensive devices. Apple is clearly NOT cribbing directly from The Innovator’s Dilemma. I’m intrigued by Apple’s response to Google of going up-scale, with things like the iPhone4S, but managing to capture huge new markets nonetheless. There’s plenty of insight still to be gleaned on how Apple succeeds so wildly.

      • Anonymous

        Apple’s success is self-evident. Make the best product, support it, improve it after the sale, treat your customers like they are personal best friends of the CEO. It’s kindergarten business methodology. Pure basics. People will always need a new computer in a few years. Building a rep for best product, best support, most improved product post-sale, and best customer service is a great sales strategy in any computing business.

        Remember that Google and Microsoft aren’t trying to make the best product. Google makes an ad platform; Microsoft makes a software platform. Somebody else always makes their products. They actually make it easy on Apple.

      • Walt French

        I don’t doubt your facts & observations. But Horace heavily features the disruption logic that he learned at Christensen’s knee, and Christensen goes out of his way to reject exceptionalism — the idea that company X is the only brilliant one — in favor of “they figured out (or didn’t) how to play the well-known asymmetric competition game.”

        As you say, Google DOES make the an excellent product for its customers. Just that those aren’t the people who walk into the Verizon store, but the advertisers who want to reach many millions of their eyeballs. (Hence, the large share of “free” phones.) Ditto, Microsoft is still selling primarily to Enterprise IT departments, where they do a great job that has nothing to do with today’s buyers of smartphones and mobile devices.

      • Anonymous

        Clayton Christensen’s commented on Apple’s unique disruptive behavior (and unique ability to avoid being disrupted) in August, 2011.

        http://blogs.reuters.com/great-debate/2011/08/29/jobs-made-apple-great-by-ignoring-profit/

        Read the whole thing, but here are some great pieces.

        “But what may be most notable is how Apple is in the process of disrupting itself right now. You almost never see this happen. The iPad — Apple’s most recent success — is disrupting the PC industry, and by extension, its Mac business. The tablet computer is going to do to the PC what the PC did to the minicomputer.

        But it hasn’t affected Apple because that’s not how the company sees the world. Profitability isn’t at the center of every decision. Apple’s focus is on making truly great products…”

      • Walt French

        Thank you. Perfect.

    • http://www.asymco.com Horace Dediu

      Simultaneous failure indicates that something else is at work. It can’t be an individual’s mistake. Disruption is a process whereby regardless of how much effort or intelligence or insight or knowledge or data or money or power you have, you lose.

      Not only does is the process repeat, it’s so common that it almost never fails to happen.

  • Ab

    Horace

    Just on the point of individual managers failing – you have worked at Nokia and in your experience would you have thought it was possible taking into consideration the culture prevalent at the time?

    • http://www.asymco.com Horace Dediu

      Managers make decisions given not only culture but the “prevailing wisdom”. Many of the decisions which might seem necessary in retrospect would be considered dangerous failures of judgement at the time they should have been made. Had a manager tried to “do the right thing” they would have been denied and blocked and perhaps dismissed. They really have no choice. Even the CEO could lose his job if he did the right thing. I cannot blame the decision makers because I know the constraints they face.
      There is a story in the media now about how Kodak is failing because they did not enter into digital photos early enough. They did not enter because it would have cannibalized film. We can shake our heads and say what a stupid mistake that was, but back then how could a manager have chosen to put a bullet in their profits and kept his job?
      The solution to the dilemma is not to kill the sustaining business but the nurture the new in a very complicated and fragile way. It is like the process of raising children: It takes a long time and seems uneconomical and a poor use of resources. Business cycles and careers are such that this is typically impermissible. And so companies die.

      • GeorgeS

        “There is a story in the media now about how Kodak is failing because they did not enter into digital photos early enough. They did not enter because it would have cannibalized film.”

        I hate to disagree with you, Horace. You might want to read this article:

        http://inventors.about.com/library/inventors/bldigitalcamera.htm

        and the WIkipedia article

        http://en.wikipedia.org/wiki/Digital_camera

        and this one from Kodak:

        http://www.kodak.com/global/en/corp/historyOfKodak/1990.jhtml?pq-path=2703&pq-locale=en_US

        – Kodak developed the first megapixel sensor (1986)

        – Kodak made the first professional digital camera (1991) in collaboration with Nikon

        – Kodak sold the second consumer-oriented digital camera (March, 1995)–Apple sold the first (February, 1994). Kodak had a massive PR campaign to promote digital cameras.

        So, Kodak didn’t “fail” because they didn’t enter into digital photos early enought: they entered before anyone else, other than Apple.

      • http://www.asymco.com Horace Dediu

        The story, which I should have linked but can’t find now, was quoting former management which described how the initiatives related to the inventions you mention were terminated or underfunded.
        Kodak was a pioneer in digital photography just like Xerox was a pioneer in user interfaces and Digital Equipment Corporation was a pioneer in microcomputers.

  • Pieter

    Horace thank you for your insight. Your best article to date!
    Two notes:
    (i) Siri may be extreme bad news for the current Google mobile search model
    (ii) A future disruption will be “mind” control. It is only a matter of “when?”

    • http://twitter.com/dicklacara dicklacara

      The bad news for Google (Bing, Yahoo) is much more than mobile search — no reason that Siri can’t run on the desktop and use Apple’s servers to provide better results with all the dross of ads and click-thrus removed.

      Siri will become the “go to” source to find information and perform tasks. The search engines will be the avenue of last resort.

      “Siri: Book me a flight on the next flight to Chicago, open return”

      as apposed to….

  • http://www.informationworkshop.org Mark Hernandez

    I’m looking forward to hearing about your thoughts regarding the “disruption of conferences.” Every last one of us has that “left out” feeling multiple times per year when there’s a conference we find essential to attend, but cannot.

    I would for now simply refer to the “iTunesification of Conferences” as putting a finger on where the disruption might be centered — a way of distributing the voice/video of the speakers to a planet-wide audience hungry for it, and making session videos 99 cents and permanently eliminating the problems of parallel tracks, fatigue, and the inability to pause and rewind what was just said (and missed because you were thinking about was said just before it) at a real-life conference.

    This is a scary disruption for some, but they need to be brave if they want to reach the people who will benefit from [the useful life of] the information presented. There will always be those who want to attend in real-life in order to meet the speakers and other attendees.

    Apple kinda already does it with WWDC. Apple charges $99 for an annual developer subscription, and there were 109 WWDC11 session videos so in a sense we’re already at that 99 cent price point (but you still have to buy the album and not individual tracks. :-) ).

    (I just now noticed that this article is not directly linked to your recent 5by5 podcast (but the topic was the same) in which you mentioned the disruption of conferences. Some may be wondering why I’m mentioning this.)

    • Anonymous

      What a great idea for iTunesU to implement.
      Virtual conferencing – most of the event’s value minus the travel, taxis and hotels.

    • http://www.asymco.com Horace Dediu

      Apple did away with conferences because they have a store network. That is how they have a venue for customers and can showcase their products.
      If I were to look at trade conferences as places where attendees can learn and have fun, I would think about the job to be done and build around that. For example, I would do them on the weekend so that it will be seen as recreation and won’t take people away from their work or meetings. I would make the audience a part of the presentation by directing them in a case-based discussion (so there would be pre-requisite reading). I would have a variety of participants, not a monoculture of experience. I would use tools that allowed everyone to share their screens not just the presenter. Etc. There is a huge amount of innovation possible.

      • GeorgeS

        “For example, I would do them on the weekend so that it will be seen as recreation and won’t take people away from their work or meetings.”

        Ah, but many people WANT to get away from their work and meetings, not eat into their minimal recreation time doing work. They could just as easily go to the conference during the week and make up work time on the weekend. The effect would be the same.

        Plus, if travel is involved, it would still take up Friday and Monday.

      • http://www.asymco.com Horace Dediu

        The job to be done of the conference would be primarily entertainment. It would be for fun, rather like the toy-like products I like to talk about.

    • Just Iain

      Mark,

      How do you see this working as compared with the TED confrences? The model there as I understand it is finding a sponsor for the videos so a few seconds of ad for the price of what could very well be very insightful.

      Another thought. Some of the attendees at the iPhone 4S launch not only live blogged but responded to the questions from their audience. Could you turn this into a show blog commercially?

  • Walt French

    I agree that Siri is huge.

    Your 10/18 piece on Christensen’s 2004 insight into Siri — that dictation couldn’t compete against a professional typist but that teens would love it — hit just as my wife reported getting texts from a friend apologizing for running late to a meeting. The friend turned out to be not a passenger, as my wife assumed, but driving + texting with Siri.

    One datapoint, a 20-something Millennial instead of a teen, but sufficient proof for me. Who in this connected cohort, as well as nearby demographics who text a lot, won’t want it?

    I look at the 4S + Siri as a prototype for the iPhone pico, a device that utterly loses its screen. Imagine a perfectly smooth lipstick tube that charges by induction and communicates locally by bluetooth, with the obvious goal of the femto going entirely into the earpiece. Or maybe it’ll be a dockable module in a small iPad.

    Mr. Rubin is about to be surrounded by LOTS of people talking at the air.

    • Anonymous

      Yes, the goal should be a device that no longer appears to be a device. Even the current Siri ads appear a little awkward as the people speak into the iPhone.

      • Anonymous

        You articulated so well what I’ve been trying to put into words. And the fact that Apple is working to make a device that no longer appears to be a device, all the while Microsoft, Google, Nokia, Samsung and everybody else are working on incremental improvements to make their devices more powerful and easier to use, is going to be the real killer. I wonder if even this device-as-non-device is the real extent of Apple’s ambitions, though, or if they have even more in store.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        It’s possible that Apple has more in store, but I think a technology like Siri is going to keep them very busy for the next 5+ years, as they explore/exploit all the fertile areas. Autos, appliances, utilities, customer service, pretty much any area where humans interact with machines.

        I think Apple will eventually get into the licensing business. There’s just too big a space for one company to fully exploit. At the same time, I think Apple will try to exert as much control over the licensees as possible (adding to or emphasizing the disruption) much as they disrupted ATTs and other carriers’ business models.

        I’m not sure Apple would want to meddle in, say, the overall outward design of an automobile, but they would want some say in the implementation of the Siri-tech. However, this might include dashboard displays and interfaces!

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        On a whim, I just had a looksie at Daimler AG: Market cap of 36.80 Billion. I’m not saying Apple would or should acquire a car maker, but they certainly could. ;-)

  • Xavieritzmann

    Horace,

    What a great article. I’d say one of your best, if you didn’t already have so many home runs!

  • Eli

    Just listened to the most recent Critical Path where you discuss the disruptive potential for a screen-less Siri phone to replace “feature” phones for people who don’t want a full computing-in-their-pocket experience. It sounds like an “iPhone Shuffle”! This is how the current iPod Shuffle works (minus the cloud & assistant).

    • Ab

      A screen less phone? What’s the cost difference in a phone with a screen and one without at the low end? $10-20? Then consider the advantages of the screen and user experience. Apple prioritises the experience and focuses its products. Adding phone and siri to a shuffle but not a screen does not make sense.

      • Anonymous

        Without a screen, the phone would be much smaller. Many other components would be ripped out as well. We’re talking about a radically different device, not iPhone-lite. Who knows if one is in the works, but it sure would be more than a $10-20 price difference and would be a different enough product that it wouldn’t likely cannibalize the iPhone.

      • Anonymous

        I agree. Without a screen, you won’t need a GPU, and the CPU most likely needs less horsepower as well. Take the screen and GPU away and put in a lesser CPU and you have cut the power usage drastically, so you won’t need such a big battery. Also, without the screen you’re much more free to design the physical device: less constraints, no easily breaking parts so less rigorous requirements for device frame stiffness etc. The total implications to the device design would be huge, and the possible cost savings quite immense.

      • http://twitter.com/nursegirlt Thea

        The rationale behind screen-less phone wasn’t solely about cost, but was also about usability for people with visual disturbances and for the illiterate. I doubt that’s where Apple will go, but Horace was suggesting that a smart startup could move into China with this sort of technology and disrupt all sorts of incumbents before the incumbents even knew what was happening to them.

      • Walt French

        Fortunately, impaired vision and illiteracy are fast disappearing from this planet.

        There are, however, some 7 billion people on the planet, the majority of whom are feature phone users for cost reasons. A Siri phone — with a small or nonexistent screen — would allow very low hardware cost while giving them Apple smarts.

        I wonder how quickly Apple can move to serve this market.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        Drunk dialing on the iPhone shuffle would be hilarious.

    • Yacko

      Siri has to display some results as JPEGs, like Wolfram, as that’s the way they are provided. Also lack of a screen would not work well for a multiple choice answer, like the answer to “List of Indian restaurants within ten blocks”. I don’t see it happening.

      • Davel

        Why? Before computers people would conversationally give this choice.

        There are 3 restaurants within 3 blocks, there are 10 within 10 blocks. I like these two the best, etc.

      • Walt French

        Today, Siri has to display some results as JPEGs…”
        Fixed that for ya!

  • http://www.noisetech-software.com/Home.html Steven Noyes

    All I know is the more I use Siri, the more I like it. I almost never use the Calendar app any more but I am managing my schedule more and more.

    I am finding Siri an enabling technology.

    • http://banapana.com Banapana

      Yes. Siri should be the fourth in that list of user interface technologies, under voice, and the industry it is about to up-end will be the search industry. Moreover, all of the complaints listed about Siri could have been said about the mouse when the Mac came out (and they were all true, too). The most important factor not listed about any of these RUIs is that they are democratizing. They make computing more available to more people, and they make computing more FUN (see GeorgeS below—he gets it).

  • Pingback: BI Research Express 11/3/2011 « Tax Australia()

  • Billykunz

    Horace:
    In a word:
    OUTSTANDING!
    A separate thought: We don’t have many women in tech development. I read Isaacson’s “Steve Jobs” book and “tech” women just are not there. They certainly have liberal arts skills.
    billykunz

  • Guest

    Nice article… just one thought. While Siri is the best voice command service out there at the moment, other computing OS providers (e.g., GOOG, MSFT) and 3rd-party apps can already provide similar experiences and could likely ramp up pretty quickly. Voice command might disrupt but everyone might find themselves swimming in a red ocean in no time.

    • kevin

      Um, you’ve truly missed what Siri is. The voice recognition part is a commodity that Apple actually just licenses from Nuance. Google, Microsoft, and lots of others have similar technology.

      It’s the back-end intelligence, including natural language understanding, that’s in Siri that all the other voice-driven apps don’t have. It’s that intelligence processing that is loading up Apple’s servers in North Carolina. So try again.

      • Guest

        Hmm… I’m sure I’m missing something but probably not as much as you’ve indicated. The questions really are a) how fast we think competition can catch up and b) how fast apple can evolve this to a point where it is disruptive.

        Are competitors collectively sleeping at the wheel and being dismissive? Are competitors going to completely miss the next shift (if we accept the premise that Siri is the next RUI) and give Apple the chance to disrupt them all over again?

        http://www.tested.com/news/siris-search-capabilities-vs-google-voice-search-on-android/3032/

        http://gigaom.com/mobile/speech-smack-down-siri-vs-android-voice-actions/

        http://www.theverge.com/2011/11/2/2533891/4s-siri-vs-original-siri-assistant-vs-vlingo

        http://gigaom.com/2011/10/17/vlingo-enjoys-a-siri-boost-voice-actions-up-50-percent/

        http://www.microsoft.com/presspass/features/2011/aug11/08-09Serafin.mspx

        Siri has the lead and is well executed, but I doubt everyone will be caught completely flat-footed by hints of shifts to voice. (…or the hints of shifts “from platforms positioned on a device to a coupled super-platform deponent on broadband and infrastructural computing” as the author is arguing.)

        So I tried again. I still don’t think I’m very far off. ;)

      • Secular Investor

        Gueast,

        As PC Magazine would say “What are you smoking?”

        Comparing Google and Microsoft voice recognition software with Siri with AI is like comparing a dumb feature phone with a Smartphone! Sure, both can make phone calls, both can receive emails, and a feature phone with WAP can kind of browse the internet, but a Smartphone can do a whole lot more. And so can Siri with AI do a lot more than Android or Windows Mobile voice recognition..

        Also both Google and Microsoft are in denial (as are you), as PC Magazine points out:

        “As I watched The Wall Street Journal’s All Things Digital Asia interview with Android’s Andy Rubin, I was highly intrigued by his comments about Apple’s Siri. Rubin told Walt Mossberg, “I don’t believe your phone should be an assistant.” He said, “Your phone is a tool for communicating. You shouldn’t be communicating with the phone; you should be communicating with somebody on the other side of the phone.”

        Furthermore, when questioned about Siri, Microsoft’s Andy Lees said it “isn’t super useful.” At the same time, he noted that Windows Phone 7 has a degree of voice interactivity in the way it connects to Bing. Thus, it harnesses “the full power of the internet, rather than a certain subset.”

        What are these two guys smoking? They both seem to ignore the fact that Apple has just introduced voice as a major user interface. Its use of voice, coupled with AI on a consumer product like the Apple iPhone, is going to change the way consumers think about man-machine interfaces in the future.

        But I think their responses were rooted in jealously and the fact that, based on what it will soon become, Siri will ultimately threaten their businesses.”

        http://www.pcmag.com/article2/0,2817,2395533,00.asp#fbid=DbgBYROw7dy

      • Davel

        To add to your response Microsoft and to a greater extent Google are not good at ‘the intersection of art and technology’. This has always been Apple’s strength. So while I have no doubt that Google and Microsoft are working on a response. I expect that the response will be as second class citizens.

        1) Because both do not play in this arena – not withstanding Microsoft’s decades long research in this area.
        2) Apple has a head start and so as first mover will have an advantage.
        3) Apple has a vision of what it wants, while the others will just have a checklist of features.

      • Anonymous

        I hope they were just feigning disinterest, because if they really believed what they said, then Apple really has a couple years in the lead.

      • http://twitter.com/dicklacara dicklacara

        Ha! For lack of a better reason, let’s call the competitors’ assistant “AndyAndy” after the two dismissive executives.

        you: “AndyAndy where was julio iglesias born”

        AndyAndy: “You talkin’ to me?”

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        Skip to the last graf for the tl;dr.

        I suspect that both Google and MS are behind in this technology, but the question is how far behind? MS is well known for its many secret R&D projects that never see the light of day because they don’t help sell or support the company’s core product. But if MS has such a project and has woken up to smell the bacon, they might not be too far behind. (On the other hand, I’m not sure how well their corporate culture could sustain such development efforts.)

        Google, too, might have some ongoing development in this area, but there are several factors that will have impeded such an effort. The main factor is that a Siri level technology does not support and might even undermine Google’s main business, advertising based search. (Perhaps Google’s ongoing acquisition of Motorola is a hedging move, in case they need to shift away from advertising and into the hardware business!)

        I can’t point to anything specific, but surely there is ongoing academic research in this area that can be or has been spun off into start ups or can be monetized thru licensing. Siri might be the breakthrough technology and first-to-market product to disrupt the industry, but I would think that it is not alone in the field. (I’m looking forward to the patent wars over assistive AI! :-P)

        tl;dr – Google is screwed unless they drastically change their business model. MS does have a chance if they can avoid falling too far behind. There might be other companies or academic institutions not that far behind.

      • kevin

        You’re still quite far off. The Siri technology is not about moving from device to a coupled device & cloud, though the practical implementation clearly requires that. The Siri technology is about computers understanding language as it’s normally spoken between people, not just the limited sets of commands that have been around for years. It’s not about dictation accuracy or transcription, which they all can do fairly well today. Human language is just not that easily understood by computers.

        So who else has similar technology? IBM is a likely possibility, though I don’t know how they’ve implemented it. I don’t know of anyone else that has a production-quality system (as opposed to an university-level prototype). Surely they’ll be people coming out of the woodwork. But all in all, technology-wise, they’ll likely be 2 years behind. Two years is a long time in the mobile world.

      • Alan

        I agree with much of what you said, kevin. But Google has shown itself to be a very fast follower and has the ability to generate systems quickly. They have the server infrastructure and so forth in place already, and the voice recognition part, and, perhaps most importantly, they have a working example of a well implemented system from which to derive the specifications on performance and interface.

        So they could copy Siri as they copied the touchscreen interface without too much trouble. It may take a year to get their beta out but if they are committed I don’t see it taking longer than that.

        The question though is if that move makes sense for them and their business model. Andy Rubin’s early comments indicate no, but perhaps he’s just using misdirection to throw off attention from what they are planning.

        My suggestion would be to monitor the Google recruiting boards for AI and natural language positions and see if there is a spike compared to their normal level.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        I think you’re partly right, at least. Apple has no monopoly on Siri-like enabling technology (although it might, probably does, have a buttload of patents). So, let’s say some other companies have developed or are developing a similar tech. Let’s say they sell or license it to Google for inclusion in Android. Because this enabling technology bypasses most search, Google has just pointed a gun to its own head and pulled the trigger. Android suddenly is not an mobile advertising revenue stream.

        All the disruptions that Siri type tech will cause are unforeseeable. However, the scenario of Google blowing its figurative brains out and perforating the small intestine in several places is not hard to picture.

        In military language, Apple is using Siri to cut off Android’s supply lines. Without Google’s search based advertising revenue, whither Android?

      • Anonymous

        Great point, and one that I’ve been also thinking about. Sure, technologically and resource wise, Google might be able to come up with Siri-like services (even though it seems Siri tech has been in development for a decade or so). But as you said, the real issue for Google is that in order to do so, they’d need to disrupt themselves and figure out a new way to do business.

      • davel

        Why? Surely Google can build a voice assistant that retains the personalization required to sell data as is its model.

        The issue for them is if Siri is as good as the demo and expands to encompass the Apple domain then that population is excluded from Google and hence a big hit to revenues.

      • Anonymous

        Well said Anon. A mature response to an immaturely rude challenge.
        Suddenly, everybody and his uncle (including Kevin, it seems) is THE expert predictor on Siri).
        Siri is a beta release and for many it seems a good enough beginning. It will not take long for its banks of knowledge, gleaned from the enquiries of millions of users, to become a dynamic and comprehensive resource that enables better and better answers to users’ questions.

      • http://www.asymco.com Horace Dediu

        Technology is not a hurdle. Neither is the ability to execute. Touch was an off-the-shelf technology when introduced in 2007 (as was the scroll-wheel and the thumb keyboard and the mouse). The disruption happened when these were coupled with business models that allowed an entrant to grow while the incumbents spent efforts in protecting their cores. Incumbents did not see it valuable to develop computing platforms and are still avoiding doing so even now because of various asymmetries with their core business models.
        AI interfaces become disruptive if the entrant uses them in ways that are asymmetric to incumbents. To put it in context, an AI-based UI could enable an entirely new way to make money from the internet. Let’s assume it can without defining exactly how. The prevailing internet business models today depend on advertising. Then if a new UI offers no sustaining benefit for advertisers then anyone whose business model depends on advertising will find reasons not to use it.

      • http://dleppik.wordpress.com/ David Leppik

        Good point. I would add that Google was free (zero revenue) for years while they tried to figure out how to make money on search without annoying banner ads. At first, Larry and Sergey seemed to be saying that they’d never take ad revenue because they hated ads. Then they found a form of advertising that didn’t annoy them, and the rest is history.

        At its best, Google can be powerfully disruptive by taking good care of users. But the best example was before Google got rich; it remains to be seen if they act that way when core revenue is threatened.

        What amazes me is that Siri has taken the upper hand in a field where Google should have all the advantages. The sort of real-time, distributed natural language processing that makes Siri work is extremely similar to the algorithms which power Google search. And it leverages a lot of the research that Google has been doing (search YouTube for “Google Tech Talks AI.”) I don’t think it will be hard for Google to catch up technologically.

        Of course, Apple ultimately has the advantage if it owns the OS. It doesn’t matter whether Siri or Google has the best voice search if everyone is carrying iPhones.

        One final note: when it comes to natural language search (web or voice), how good the results can be depends on how much input data there is– how many queries. Google is the best because it has enough data to analyze very tricky edge cases. This virtuous cycle produces a natural monopoly. Siri may well follow the same path.

      • GeorgeS

        “The sort of real-time, distributed natural language processing that makes Siri work is extremely similar to the algorithms which power Google search.”

        Are you sure? I typed into a Google search box, “I am locked out of my car.” Google came up with mostly irrelevant link. The best were actually ads. Excluding the ads, the top article was on answers.yahoo.com and included that phrase.

        That is NOT “natural language processing.” That’s word/phrase matching. it’s the difference between a concordance for the bible and a text on biblical criticism.

      • russell

        I think the challenges they(microsoft) face is IF they could match Siri capabilities, then the OTHER problem to overcome would be to get folks to buy their phones. The mometum has shifted away from them on so many fronts w/ the mobile internet. Saw a stat that stood out in presentations by Roger McNamee(Elevation Partners, VC) and Mary Meeker (Kliener Perkins, VC) that Microsoft’s marketshare of connected consumer devices to the internet has dropped from 90% to less than 50% in the last 48months. Losing share can be very easy but getting it back is very hard in technology.

        Otherwise, great video though. I used TellMe for a period of time in mid 90s and was not impressed. It’s looks like they have improved a lot based in this clip, even if it could reliably do half of what they are showing here.

      • k.mun

        Good point. It is the contextual nature that will be differentiating. With the control over underlying OS and to some extent third-party apps, it is a matter of aggregating user input (touch, voice, gesture) in the context of app usage and search queries that will distinguish Siri service.

      • Jamie=

        I think it’d be too early to write off the competition, but getting something like Siri to work is very similar in principle to getting the “I’m feeling lucky” button on the Google home page to work, and a decade hasn’t really brought any movement on Google’s front. And Microsoft has been spinning its wheels on voice command, despite loudly proclaiming it’s future, for at least as long.

        It’s a little wonky and management-consulty, but I really don’t think either of them has a quality culture. Google’s really innovative but they don’t spit-shine anything. If you go into the Google restaurant and order a chicken, it’s liable to come out tasty but still have lead shot in it, and if you go into the Microsoft restaurant and order the chicken, you’re going to get nuggets and sweet’n sour sauce.

    • Walt French

      I agree with you that thinking about voice command is more important than thinking about “Siri” per se.

      But while the competition will also recognize the potential of it, there IS such a thing as clever bundling and tuning for real people. For the last couple of decades, usability and intuitiveness have been Apple’s defining strengths. No reason to expect that Google (and certainly not others) will somehow snatch away that crown.

      I see lots of people dismissing Siri, especially those who have Google speech capabilities and don’t use it. But I think the industry has evolved a LOT since firms dismissed the mouse for a decade. (Helps to see RIM being eviscerated for not recognizing the iPhone’s disruptive potential.)

      I expect to see lots of competition in this area. And I expect Apple to benefit wildly from having brought it first as what customers can perceive as a usable feature, and to benefit from their lead in the area.

  • Anonymous

    Great insights. Heard the nub of this yesterday on 5by5. Voice may be the new interface. How far can this go? In the sixties, Joseph Weizenbaum wrote ELIZA, a program that responded in the manner of the psychologist Carl Rogers with, in primitive form, his client-centered, non-directive approach. The resulting “conversations” were a bit eerie. As I recollect, researchers found that users liked ELIZA. As they do Siri, clearly.

    And voice removes the language barrier. If I can understand your Spanish and you can understand my English, we are good to go. Ergo, if Siri understands your Spanish and my English and delivers usable translations, that’s fine too.

  • Anonymous

    I also think there is one more reason why incumbents dont invest in the disrupting product. From an incumbents point of view, they see many such potential disrupters everyday that it gets hard for them to choose any one disrupter and invest in that channel.

    Also, there probably are many such potential disrupters that failed for every successful disrupting product. Even if the incumbent is interested in investing in all the potential disrupters, they aren’t willing to do so since the success % is way too low to explain it to the stakeholders.

    • Spike Ennis

      Great point but you forgot one thing. Companies are just sitting around waiting to see the next disrupting product….. They should be creating the next disruptive product.

      Apple creates, fails, tries again. Apple creates the iPod version of an MP3 player, the iPhone version of smart phone, they become the disrupter..

      Then the disrupt their own product. Why, As Steve Jobs said, “If someone is going to take sales away from us,,,,,,,, I want it to be us.”

      Ah to the Crazy Ones…. they think different (disruptive).. LOL

      en

  • http://www.facebook.com/people/Daniel-So/19902953 Daniel So

    imo, the next disruptive User interface technology is advanced haptics. To be more specific, it will be haptics that allows a phone to be used “blind” — think Matt Damon in “The Departed” furtively texting Jack Nicholson from his pocket. Right now, this is perhaps the only area where feature phones > smart phonesm, as the current generation of UI requires too much visual oversight. Technology like those introduced by the latest Nokia prototype — such as the ability to input through bending a phone — is the next big thing.

    • r.d

      you may want to update your talking point. It seems you don’t know the
      capabilities of iphone.

      Stevie Wonder recently said the following

      “There is nothing that you (concert audience)
      can do with the iphone that I can’t.”

      • Anonymous

        The point being, iOS has very deep accessibility features that allow blind people to use iPhones, iPod touches, and iPads with ease today. Haptics is completely unnecessary for accessibility, as Apple has already proven in the market.

  • Anonymous

    It’s very simple… incumbents are no longer hungry to learn, no longer hungry to evolve, no longer hungry to experiment, no longer hungry to ask bigger questions…
    Microsoft, on the last article I read, spend about 10 billion a year in R&D, Apple spend a bit over a billion…
    So who’s spending the money more wisely.. but more importantly, who’s turning R&D into products, and profits?
    A big patent portfolio is fine, and patent revenues are top tier in global corporations, but in order to continue creating new ideas, and products, new and exciting products need to be put to market. Those product revenues pay for the R&D for the next big product…

    • Seran89

      This is easy to understand. Disruption means “change”. The incumbent has settled, his market share and revenue is secure and stable. The last thing the incumbent wants is anything to change. On the contrary, the strategy of all incumbents is that everything should stay the way it is, forever.

      Until the gap between the lowest common denominator that allows the incumbent to sustain its position and what’s possible becomes so big, that a disruptor can enter the game.

    • Just Iain

      “but more importantly, who’s turning R&D into products, and profits?” A lot of MS work is invisible to the public with constant updates to the Corporate products. It’s the average consumer that is underwhelmed by Microsoft products.

  • http://www.facebook.com/lee.penick Lee Penick

    Thanks Horace. Wonderful work. I think there is a lot of money to be shifted around and made based on what you pulled together and summarized so well. The brilliance of Apple is being demonstrated. Will be an interesting 5 years as this plays out.

  • Christian Heidarson

    Horace, this post validates your position of how no-bs data analysis can elevate the discussion. I think many people have felt in their guts that Siri is something huge and potentially a step-function in the industry. But with your charts and figures you provide a way to monitor the impact concretely. Thanks!

    Nevertheless, a word to the wise: skate to where the puck is going! I would bet lots of money that within two years, Apple will introduce its next UI revolution. So, the question I am waiting for you to address is:
    How is it that Apple can continually introduce game-changing UIs? It must be more than the tech/humanities intersection thingy. Isn’t there a difference in business model here, where Apple has found a way to MONETIZE user interface innovations where other fumble?

    • Anonymous

      I feel it is quite simple, actually: Apple’s customers are the end users, and the end users are the most likely to appreciate and value better user interfaces. It is therefore integral to how Apple does business to maximize the end user value. The same is not true for Microsoft or Google, at least to a degree.

    • kevin

      One of the ways Apple succeeds is because they are willing to disrupt their own products. Sometimes, it is through the introduction of a UI that is better suited for the jobs at hand. The UI gets monetized by that product because it turns out to be a better product for the consumer with those jobs to do.

      It also succeeds by having few, well-fitted UIs, and by constraining its product line to fit its few UIs.. Apple has a Mac UI, iPhone/iPod Touch UI, iPad UI, iPod nano UI, iPod shuffle UI, iPod classic UI, and AppleTV UI. For example, the iPad UI is designed for a 9.7″ multi-touch screen, and the iPhone/Touch UI for a 3.5″ screen. So there are no 4″, 5″, 6″, 7″, or 8″ phones/tablets because Apple doesn’t believe the current UIs are well suited for them, and doesn’t see any additional value in having another UI for them..

      For another example, in recent years, Apple refused to put its Mac UI on products whose screens couldn’t fit the width of a typical web page and whose keyboards couldn’t be full-sized. There was no computer with a screen smaller than 13.3″ until recently when screens became good enough for Apple to release an 11.6″ Air but still with a 1366×768 native resolution, full-size keyboard, and sizable trackpad..

  • Fred

    My thoughts on various ideas raised here.

    Graphs are interesting but I think growth of Android isn’t shown in them. It would be interesting to see the graphs as of last quarterly, particularly with Samsung now biggest selling mobile manufacturer.

    It interesting when discussing UI innovation that nobody mentions how superior Android OS is with the application widgets, true multitasking with app preview. These features enhance GUI both in functionality, organisation and experience.

    When I think about Siri as a UI compared to others I go to a rule my favourite of thumb for UI design “Will a non tech user find it easy to use”. At first you might think yes because you just ask a question, deciding what question to ask and trying to control Siri is hit and miss. And fast responses are highly dependant on a good internet connection. This equals sinks the technology for a non tech user since internet connections are often not up to scratch and the results of Siri aren’t always relevant or correct.
    The point of relevance also answers why google is in no danger in the search business.Google is no. 1 because its search results are the most relevant, unless Siri can provide better search results than google, than there is little competition.

    Is Siri an UI or actually an assistant, I think its the 2nd some some UI features.
    A better UI will let you do almost everything you can do with existing UI but in a way that is more intuitive and equal in speed or faster. Again Siri fails here.

    This is partly because the GUI helps the user organise their work/life or interests into areas and this organisation acts as memory trigger when useful links and applications. This functionality is slow with voice control and is impossible without a GUI.

    The idea that talking to a machine is a better UI also flawed and this is simply because of the design of the humans senses. For most people the visual sense is dominant, hence all GUIs are superior to non UI.

    Communication between people is mostly visual, as so it should be with a all things. This is what devices must innovate towards.

    Is the GUI easier to control via voice, which is what a mouse, click wheel, and touch screen provide. The answer is no, voice control is clumsy, slow and doesn’t allow for fine control.

    The real future for UI is innovating on how humans can interact with a device in a more intuitive, faster, and finely controlled way.
    There is best method then using hands and fingers.
    The only better way to control something would be to directly control a device via the brain.

    Where current devices fail is that touch screen doesn’t provide much tactile feedback, doesn’t allow for true 3d control.
    Talking to a device is fun and gimmicky but not a serious way to interact with a device.

    The real future for UI with devices is when human gestures control the device without touching the screen in a nature 3D manner but still displaying the results in a 3D GUI.

    • Anonymous

      First off, the article was about revolutionary user interfaces, not merely innovative. I agree that there is a lot of room for improvement within the confines of a mobile multi-touch user interface. But in order to be revolutionary, the innovation(s) need to change our mode of operation. Widgets, multitasking UIs etc. are incremental improvements that, as you say, make it easier, more intuitive and efficient to use our device but we still continue to use it in the same way as before.

      I obviously disagree with your view about Siri and what its effect will be. You are looking at what are the shortcomings now, and why it won’t/can’t work. But Horace listed 5 reasons why Siri is disruptive, and the very first one was “It’s not good enough.” You’re listing many of the same things as Horace did, which is anecdotal evidence that Siri is, in fact, disruptive. Imagine how far Apple can go with Siri in just a few years when they have all the data of tens of millions of people using it.

      Obviously there are certain significant areas where Siri won’t be all that useful, for example games and drawing/art apps. For those, the touchscreen UI will remain. But for a number of things Siri already is easier and faster than any traditional UI, for example setting alarms and reminders and calendar appointments is probably an order of magnitude or two faster using Siri than via other methods.

      You say the future of UI innovation is in interacting with the device faster, more intuitively and with more control. I say the future is making the device do what we want, and Siri has the promise of doing that.

      • Just Iain

        “Obviously there are certain significant areas where Siri won’t be all that useful, for example games and drawing/art apps.”

        I suspect that for graphing and games, voice input would be fantastic. For example the interaction of human and computer assistant Jarvis in the Ironman Movies.

    • jawbroken

      “It interesting when discussing UI innovation that nobody mentions how superior Android OS is with the application widgets, true multitasking with app preview. These features enhance GUI both in functionality, organisation and experience.”

      How do you explain the large disparity in user satisfaction between the operating systems?

      “And fast responses are highly dependant on a good internet connection. This equals sinks the technology for a non tech user since internet connections are often not up to scratch and the results of Siri aren’t always relevant or correct.”

      Analysis of the network usage shows it is currently quite minimal, which makes sense as it is designed to work well over mobile connections. I don’t see how your assertion is supported.

      http://arstechnica.com/apple/news/2011/11/how-data-heavy-is-siri-on-an-iphone-4s-ars-investigates.ars

      “A better UI will let you do almost everything you can do with existing UI but in a way that is more intuitive and equal in speed or faster.”

      I don’t see why UIs can’t be complementary. When the mouse was introduced we didn’t throw out the keyboard. A device can easily support touchscreen and voice interactions and the user will pick what is appropriate and efficient for the task, just as they do currently with a keyboard and mouse. Siri already shows efficiency and ease-of-use advantages for things like calendars, alarms, reminders, playing music, etc and has advantages in situations where touchscreen input isn’t convenient, such as when driving.

      I’m not sure why you believe 3D UIs are particularly important but that seems tangential to the article anyway.

    • Anonymous

      You wrote, “My thoughts on various ideas raised here.”

      But, rather than give your thoughts on the ideas of RUI, you mention this, “Graphs are interesting but I think growth of Android isn’t shown in them” What exactly does this have to do with the “various ideas raised here”?

      Further, if you are so interested in the growth in Android, it is visible in all 4 charts by looking at the lines of Samsung and HTC.

      Then, you wrote, “It interesting when discussing UI innovation that nobody mentions how superior Android OS”. Yet again this is not what is under discussion. Android is just another multitouch OS that with iOS disrupted the incumbents starting back in 2007. The topic isn’t whether you like Android better than iOS, but why the incumbents didn’t see the new touch interface disrupting their business back in 2007, and whether Siri is a harbinger of a new disruption.

      Then, you wrote, “Will a non tech user find it easy to use”. At first you might think yes because you just ask a question, deciding what question to ask and trying to control Siri is hit and miss. And fast responses are highly dependant on a good internet connection. This equals sinks the technology for a non tech user since internet connections are often not up to scratch and the results of Siri aren’t always relevant or correct.”

      While reading this, I could not help but think you have either not used Siri, or you haven’t used it much. Further, this is exactly the kind of thinking that is ripe for disruption, as pointed out by Horace’s first three points: “It’s not good enough. There are many smart people who are disappointed by it. Competitors are dismissive”

      Congrats, you helped make Horace’s point!

    • sleiii

      FYI: In your statement, “Is Siri an UI or actually an assistant…”, the standard usage rule is to use “an” before a vowel sound, not automatically before every vowel. In this case the sound in “UI” is “yoo”; so ” a UI” and, as you correctly note, “an assistant.”

      (A professional courtesy from a parallel “technology.”)

    • kevin

      “Is the GUI easier to control via voice, which is what a mouse, click wheel, and touch screen provide. The answer is no, voice control is clumsy, slow and doesn’t allow for fine control.”

      Not always. There are many tasks where voice control is much faster than any other method. Siri is slower if I need a see a list of multiple items or a long text-based or graphical response, since seeing is faster than having something read to you. Excluding that, Siri is just one action and some spoken words, so anything that requires more clicks (including typing) or multiple gestures will take longer.

      Besides the ones Apple’s already shown in their ads, here are some other examples where Siri can eventually beat any app:
      – what is 22% tip on a bill of 45 dollars and 82 cents?
      – split the bill of 87 dollars and 36 cents across 7 people.
      – where is Barack right now? (where Barack is in Find My Friend)
      – when will Barack get here? (where Barack is in Find My Friend)
      – which store has the lowest price on 2% milk? (where Siri knows to look locally based on where I am)
      – when is United flight 247 from New York arriving today? (where Siri knows where I am so knows which airport I’m referring to)
      – when should I leave to get to the airport as soon as it arrives?
      – is there on-street parking for XYZ restaurant? where is it?
      – do I have any library books due today?
      – is my Best Buy purchase ready to be picked up?

    • davel

      While I will agree that people are highly visual communication is mostly by sound. Sign language is used by a minority of people. Mostly because of a handicap. When watching a movie do you turn off sound and turn on the subtitles?

      Kevin gave a nice list of examples on how natural voice is.

      However I am not saying that voice will replace visual, only that it adds a powerful alternative to the existing visual interface which Horace points out in this post. Also the examples you give are limited. They are strictly limited to standardized computer interfaces to inter-operate with a computer. The Siri demo shows how natural an audio question and answer sequence can be without using your hands and eyes to focus on a device. This can be very useful if you are driving where both senses are required to safely control the vehicle.

  • Hossein

    Guys who read this blog and the comments regularly, do you remember my two predictions?
    http://www.bgr.com/2011/11/03/siri-suffering-from-extended-outage/

  • http://twitter.com/phil_hendrix Phil Hendrix

    Horace, your readers might be interested in the following:

    Leveraging capabilities from Siri… and other capabilities, notably speech recognition, voice UI (user interface), location and the cloud, Apple is poised to deliver an innovative platform that once again separates the iPhone from other OEMs’ mobile devices… If the benefits to users are as compelling as suggested by Siri co-founder Norman Winarsky, who called it “world changing,” consumers will be clamoring for a new iPhone for Siri as much as for the device itself.
    Is the iPhone Assistant Apple’s next “Blue Ocean? http://bit.ly/nNuOYR

    Advances in speech technologies will fundamentally alter the way in which users experience mobile devices and apps: devices secured with voice authentication; individuals — including the sighted and visually impaired — enjoying mobile content and apps without ever having to touch or view a device; new applications, from search to language translation and others, enabled by speech recognition; and in many other ways.
    (Report) How Speech Technologies will Transform Mobile Use (http://bit.ly/l2Ax9J; request copy from author at @phil_hendrix

    Is iPhone’s Voice Control the Sound of Things to Come? http://bit.ly/iGNBqk
    Starting with the Mac but most evident with each new generation of “i” products — iMac, iPod and iPhone — the company has demonstrated time and again what so many other device makers and mobile operators have failed to understand: It’s the UI, stupid! So when Apple features Voice Control in commercials for the newest iPhone 3GS, the mobile industry should sit up and take notice.

    Prioritizing Opportunities for Speech-enabled Apps
    http://bit.ly/skXPSY

    The Ideal Speech-Optimized Mobile Solution
    http://bit.ly/rqjxFU

    Dr. Phil Hendrix, immr and GigaOm Pro Analyst
    http://www.immr.org; @phil_hendrix

    • http://twitter.com/Reiserechner Stephan

      And yet, Horace never put links to his stuff in other people’s comment sections. Just something to think about…

  • Pingback: OnlineMagazine » Blog Archive » Artificial stupidity: Siri suffers a 5-hour outage()

  • James

    Siri is an absolutely genius move by Apple. The thing is, Siri *completely bypasses* search (except for instances in which it can’t answer your question at all, which will surely decrease as time goes on and the technology improves). This puts Google in a very difficult position with Android – if they implement a similar feature, they’ll only end up shooting themselves in the foot, since bypassing search would severely dampen the effectiveness of their advertising platform, which is the whole reason why Android exists.

    • Troymclure

      Completely agree with you.. A few days ago, I was using SIRI in my iPhone, then I leave from work and it was a hot day out there, so I look into my car’s dashboard and I was thinking “I wish I can just ask my car to turn on the A/C..
      Siri is the most amazing user interface and if apple gets to implement it into other devices, I’m pretty sure it will revolutionize any industry it get on.

      • Anonymous

        You can ask your car to turn on the A/C. It’s called Microsoft SYNC and it’s in many current Ford cars. Look it up.

      • Anonymous

        I have SYNC in my Ford, and Siri as well. SYNC is no Siri. Unlike SYNC, Siri does not require specific syntax to work. The computer and network do the work, not the human. You know…the whole idea.

      • http://www.multiplexcomic.com/ Gordon McAlpin

        You have that backwards, I think: Apple doesn’t need to implement Siri into other devices. Other devices’ manufacturers need to implement iPhone docks in their cars and create APIs that iPhones can access, such as the Nissan Leaf’s iPhone app, which already does give you some measure of control over the AC — not with Siri, of course. Not yet.

        It’s just a very short matter of time before Apple opens up Siri’s API to developers. That’s when Siri will get REALLY interesting.

    • davel

      Yes, but even if Google clones Siri ( I don’t think it will be as easy as you think ), they will still be cut off from a significant source of mobile searches which will hurt revenue.

    • Ab

      What if Google stops Apple from connecting to their API for the search? Would you use Siri with Bing?

      Google still provides the best search results.

      • jawbroken

        In what way would that be advantageous for them, to give up even more ad revenue?

      • Kan

        Currently when you search on Google you get a list of results and a list of adverts. If siri only pulls back the organic lists then google isn’t losing any ad revenue if they block siri?

      • kevin

        You are correct that Google wouldn’t lose any more click-through ad revenue because they wouldn’t have been getting any from Siri anyway, but they will lose the opportunity to “learn” based on which link the user chose. That learning will now go to another search engine elsewhere, which will enable it to improve. In addition, users will start to become more familiar with another search engine.

        If Google blocks Siri from its search, I wonder if Apple would keep Google as the default for the remainder of its products. Maybe, maybe not.

      • http://www.asymco.com Horace Dediu

        But if they only deliver adverts then they are not answering the question. The user is asking “I need to find a flower shop” not “which flower shops are paying to find me?”

      • Ab

        I don’t follow. Who is only delivering adverts?

        Googles spiders crawl the web, the parse the webpages for keywords and then rank the pages using their pagerank algorithm. We do a reverse lookup ie we tap in search terms and Google returns a list of webpages sorted by pagerank plus a list of adverts.

        Who has an Index as wide and deep as Googles? Yes we know Googles model but just saying Siri simply disintermediates this is too simple a conclusion.

        If Siri cannot access Google Index then its results will be from Bing, Wolfram Alpha? Yahoo etc I assume and even thought I have tried Bing I have returned to Google.

        @Kevin

        As to learning people only tend to click on links on the first page which are the ones with the highest pagerank so it really just reinforces behaviour and pagerank. The other search engines dont have as wide index as Google.

        As to your second point I see Apple and Google try to decouple their linkages and reliances as much as possible.

        Apple controlled content distribution through itunes and Siri could potentially be a way to control Google “content” but if Google looks at the past letting Siri do this would be detrimental for them.

      • jawbroken

        Sure, but nobody said that and that’s not how it currently works at all. I don’t know why you would presume this would happen.

      • GeorgeS

        How is Google going to know that a query comes through Siri and not through some other application? (Siri could imitate a web page entry, for example.) Would Google block ALL iPhones?

      • Anonymous

        I’m not sure Google could. How could they know it was a different search vs. a user typing it in?

  • Andy

    When apple came out with OS X, iPod & iTunes they change their business model especially when they changed the company’s name from “Apple Computers” to “Apple Inc”. And Steve Jobs professed to the world that Apple would be a digital hub. I don’t think anyone had any idea what he meant or to this extent. For example, iPhone, iPad etc… Now we are all bearing witness.

    • http://www.asymco.com Horace Dediu

      Ah Apple Computers. I just bought this yesterday for 1 EUR at my local Salvation Army store.

  • Kevin

    To my view, SIRI is silly, as are the Android variants. Why would I tell my phone to make an appointment, when I can just hit the calendar and do it myself, in about 20 seconds? But that’s just one user’s preference.

    Nor is this a brand vs brand issue, as so many here have made it. The question for me is one of technology, as in when will voice interface actually become usable for something more than parlor tricks. Whether it’s SIRI, or IRIS (the Android variant ginned up in 8 hours by some MIT whizzes), the real question for me is when features such as face recognition phone unlock, etc., will become part of a truly game-changing package.

    Brand loyalty ignores development. So Android people say that SIRI and a 3.5″ screen is a joke, Apple loyalists say “Who needs facial recognition software, and a 4.7-inch screen is gaudy.” Fact of the matter is that both platforms are pointing the way to a new, much more interesting future in the devices that are about to replace our home computers.

    So it isn’t a question of Apple somehow suborning Google’s strategy by freeing the user experience from search engines. Google doesn’t seem to care about that, so invested are they in other means and platforms. Like Apple, Google wants to control its market. Ask SIRI a question, and it will search the Web, just as any of the Android voice apps do. That isn’t the question. The REAL question is learnability, as in the future when we can have a phone that will unlock when we look at it, ask us what we want to do right now, and respond to that task. At present, EVERYTHING on offer, including SIRI, falls far short of that promise.

    Both platforms are evolving and adapting, as as they adapt, they move closer to similarity. It doesn’t take much time with a Samsung Galaxy S2 for an honest Apple to find things to like, and vice versa. We need both platforms, and their competition, for mobile technology, whatever platform one craves, to reach its fullest potential.

    • jawbroken

      “Why would I tell my phone to make an appointment, when I can just hit the calendar and do it myself, in about 20 seconds?”

      20s sounds like an eternity compared to the time it would take to say the details of a simple appointment.

      I’m not sure what point you are driving at with your comment, or how facial recognition unlock is particularly relevant to a post about UI.

    • davel

      I think Google cares a lot about Siri. If they don’t they will when the ad/search revenue they gain from Apple goes to zero because the searches are anonymous or not done through them.

    • Anonymous

      suborn |səˈbôrn|
      verb [ with obj. ]
      bribe or otherwise induce (someone) to commit an unlawful act such as perjury: he was accused of conspiring to suborn witnesses.

      Probably not the word you were looking for.

      • Anonymous

        suborn has more generic meanings in vulgar (common) usage.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        Uh, like what? “I’d like to suborn her!” Please, this is not a criticism of your intelligence. You’re just factually wrong about the definition of a word, which is no big deal.

        The more precise we can be with language, the better we can express ideas. An intelligent person such as yourself is open to learning from mistakes, no? Even small ones, like this!

      • Anonymous

        Like in “to corrupt”, except that it also implies specific actions/behaviors.

        But “factually wrong about the definition of the word” as a rebuttal of my it’s-a-vulgar-as-in-common-use-of-the-word is a little off, don’t you think?

        Of course I’m factually wrong when you stand it up to the definition of the word from a dictionary, but if I said the word “pervert” was synonymous with “corrupt”, you don’t think the vulgar definition of “pervert” wouldn’t pop into your head?

        Seriously.

    • Walt French

      “To my view, SIRI is silly…Why would I tell my phone to make an appointment, when I can just hit the calendar and do it myself?”

      You’re probably used to making thoughtful analyses in some domain, and know how well you think about things. And this logical ability overlaps with your personal decision-making, as well.

      But I propose that when something is touted as BIG and it seems silly to you, that you go meta for a moment and consider whether your perspective is somehow limited in understanding what’s going on.

      I recall a coworker returning from the 1980 West Coast Computer Faire and telling me breathlessly about VisiCalc. To me, it was silly: why put in 70% of the work of learning FORTRAN, in order to get 10% of the computational power, and be stuck having your answers be locked into a single grid, to boot? Yet VisiCalc and its successors defined the Era of the Shrinkwrapped PC Application that is only now being succeeded. Despite his high IQ, Mike was much less adept at programming than I and could easily see people with even less skill than he had, getting lots done.

      So now to Siri specifically:

      Siri would’ve allowed me to set up my next dental visit in 10 seconds, versus the 30 or so it took to open my calendar, switch to May 2012, etc. And I could’ve done it before I’d finished walking out of the dental office. It could’ve let me stay more hands-on, in eye contact and conversation when, during meal prep, friends wondered (off topic) about the nutritional values of foie gras. It DID allow a friend to text while driving, with an ETA on meeting for a get-together.

      These uses morph from extensions of 1950’s-era computing per se into 1980’s-era PC apps into 2010’s-era assistance. Especially that last example is simply not possible with a typing/reading model, and I suspect that Generation TXT — anybody from about 10 to 30 plus a fair number up to their 60s, including their parents — will soon be ALL OVER Siri because it is faster, more convenient, more flexible and costs essentially nothing for all those plusses.

      I see Siri as about Apple’s long-standing emphasis on making “computer” functionality ever more available to people (which to them, seems incredibly “powerful”), versus others’ emphasis on Watson, AI, more “powerful” tools that are not available. That may not be your, or my first item on the wish list, but it certainly will have a big impact on many millions around the world.

      The perspective of what new uses and new users, is incredibly significant given the 6+ billion people who don’t yet own smartphones. No, it won’t do quite as much for those adept at manipulating information such as most readers/commenters here — I, for one, am not rushing out to break my old contract — but even we will find it very nice in its “limited” way.

    • Anonymous

      “To my view, SIRI is silly, as are the Android variants. Why would I tell my phone to make an appointment, when I can just hit the calendar and do it myself, in about 20 seconds? But that’s just one user’s preference. ”

      I’d bet it would take more than 20 seconds, just that muscle memory for it makes it feel a lot faster.

      But on the other hand, it would take Siri (it’s not an anagram) a lot less time (probably less than 20 actual, real-world seconds) to do it.

      I guarantee you that something “ginned up” in 8 hours by MIT was not built from scratch and/or did not have the contextual understanding that Siri does.

      I’m not quite sure you get what Siri does. It’s not just a voice command replacement for typing.

    • Anonymous

      “Ask SIRI a question, and it will search the Web, just as any of the Android voice apps do. That isn’t the question. The REAL question is learnability, as in the future … ask us what we want to do right now, and respond to that task.”

      Siri does limited learning, it learns your preferences within a few domains like restaurants, movies, events, services. For example, with airlines reservations if you live near two airports and ask Siri to schedule from one of them, the next time you fly, it will default to that airport. If you have a frequent flyer program or you prefer aisle seats, you’ll just say this once and your preferences will be saved for the future [1].

      There was an interview with the company founders last year, and they didn’t sound like they planned for Siri to initiate conversation and ask you what you want to do. But you will be able to say “I want something to do” and as time goes on, Siri will learn about you and the responses will be tailored for each individual [2].

      1. Robert Scobel. Interview with Siri’s founders Dag Kittlaus and Adam Cheyer. “A new personal assistant on your mobile phone.” February 5, 2010.
      http://www.youtube.com/watch?v=ts4yr3qti9w – slide timer button top 13 min 15 sec to hear about flights.
      2. Go to 11:30.

  • Spike Ennis

    Sorry but I think you are wrong sort of.
    “My hypothesis is that The Primary Cause for the shift of profits from Incumbents to Entrants has been the disruptive impact of a new input method.”

    Older companies just try to make profit. Make money. Improvement is nothing we really want.
    Apple looks to improve for the user. Improve interface, improve operation, improve use… Make an insanely great product and people will buy it.

    When you are greedy and the salesman run the company, you just sell what you have, hyping it to the max. When you make great products, eventually people find out. Then they beat a path to your door…

    Its just that simple.

    • http://www.asymco.com Horace Dediu

      When an improvement is consistent with the existing business model then that improvement will lead to increasing revenues and profits. Incumbent companies are always motivated to improve along the prevailing basis of competition. The disruptive impact of of a new user interface is that it creates new business models which, being asymmetric, compels incumbents to ignore it.

      • Anonymous

        Horace: I find it amazing that the key to disruption that Apple has employed time and time again is eternally missed by its competitors. That key could be best stated as the “the power of simplicity.”

        At every step in the history of Apple’s new products, competitors (and fans of the comptetitors) dismissed the advantage of Apple’s new interface as “being nothing new.” Every time they would say: “We can do that already on the (fill in the blank) platform…all we have to do is download a couple of new applications, write a python script, upgrade the computer and voila! Same as Apple! I should probably write a story about the history of this.

        But there is a theshold level of ease of use that makes the difference between something that works and something that disrupts an industry. A new interface that allows just a tiny improvement in how people use their technology will not disrupt the industry; but once you provide methods of use that are “just enough easier” it becomes a game changer.

        iTunes wasn’t the first music download service, but it was the first that enabled twelve year olds to manage their music libraries. The Apple II wasn’t the first small, cheap computer, but it was the first that somebody could take out of the box, turn on and run a program. The Mac wasn’t the first computer with a simplified interface (GEM Desktop anyone?) but it was the first that allowed the mass market to do useful things without training. The iPhone wasn’t the first touchscreen phone, but it was the first that allowed grandma to comfortably use it. Apple retail stores weren’t the first computer stores but they were the first to make shopping for your technology a pleasure instead of a chore.

        In each case, the criteria that Apple applied to the product design was “is it enough simpler than what exists?” Its competitors would be happy just including a Wizard to help the user through a convoluted process.

        Google can upgrade it’s voice actions but it is unlikely that it will be able to hit that “just enough easier” for quite some time. Siri is already there and it will disrupt and define the industry as a result.

    • Anonymous

      Microsoft is trying to create the next great new interface in their Microsoft Research division. They’ve been doing that for 20 years. They’re just not very good at it. Apparently Kinect came not from them but by Rare, a small game studio subsidiary of Microsoft Game Studios.

      Also, you can’t deny MS is trying to improve the interface with their Metro/Zune style. I don’t like it, but many people find it refreshing. If you’re tired of looking at a grid of icons when you look at your mobile device, I guess that’s the best alternative at the moment.

  • Anonymous

    Regarding Siri, I suspect that Google will behave much as they did with Android. After Apple’s initial success with Siri, Google will copy the primary features that users find the most useful. All the extra features and polish of the Apple implementation will be ignored. Then the Google offering of solid, basic functionality will be provided and adopted by all the competing handset manufacturers who will maintain their market share against Apple.

    The recent concentration of tech innovation in an area that a sole consumer can afford to buy (as opposed to just large corporations) is great. I don’t know what the future holds for mobile technology. It’s been on an unbelievable run lately. What I do know is this: it’s a great time to be a consumer.

    • Jet

      The problem with duplicating Siri is that – especially if an interaction is voice-only, it’s hard to slap ads on it. Text ads don’t consume your time, radio-style audio ads would. Google essentially already has a siri competitor in google voice, it’s just much harder to monetize than ads on a webpage.

      • Ab

        Say I am searching for some information on calf tears – yes I like to run. I get returned a few forum postings – you think I would wait for the voice system to read me everything out or rather more likely I will scan the posts and see if I can find what I need.

        Voice systems might be great at providing instructions and commands but as yet fall flat when you are returned data.

      • jawbroken

        Sure, as it is with any interface, there are things that are more natural to do with one interface than another (e.g. mouse vs keyboard). What if, in the future, you could ask “what’s the best way to treat a calf tear?” or “how do I avoid a calf injury while running?” or whatever, though.

      • jawbroken

        And, of course, data can be returned as text or voice or both, the input and output channels don’t have to be of the same type (as they aren’t when I use a mouse to navigate the internet).

      • Ab

        Search has always been moving from simple word search to becoming more semantic and understanding context. When I am searching for Chelsea football club I do not want any reference to Chelsea Lately (awful program ). So results can be returned in voice, text, video, images or any combination.

        The issue for Google is how do they continue their ad business in this new paradigm. If results are returned as text, video or images then it requires a screen to view so ads can be pushed similar to the way they are now. However if the interaction is voice only what is stopping Google from sending you back organic and paid results without differentiating between them?

      • http://www.asymco.com Horace Dediu

        Is search a product or a tool? Do people want a tool to solve their problem or do they want their problem solved? Put another way, does the buyer of a drill want a drill or a hole in the wall?
        If you think people want to search then a new input method may be sustaining. If on the other hand people want to use search to solve a problem then a new input method may be disruptive because it offers the solution to the problem without requiring the use of some tools.
        I remember a time before people searched and I hope to see a time in the future when people will no longer search.

      • Ab

        I think you are mixing a means to an end and the end itself.

        Lets take your example – yes I need an end ie “hole in the wall” but I also need the drill to do it this time and next because I am after a product or service that leads to an outcome.

        I could extend this to I need my ACL repaired – I know the desirable outcome but how can I get there? Well I need to find an able surgeon with a good reputation. I have to go find one through multiple means.

        People are always using search to solve a problem – if I have a problem with some coding issue I can hit the forums.Medical problem – search the web about

        We will always be searching – I am looking for a good plumber, buy a new car etc. The idea that we will not search seems rather strange.

        Currently Google indexes unstructured information which we search. This information is explicit – its writen down. The disruption is when we can get people to expose their tacit information which they hold ie their knowledge which has been done to some extent by Quora and Twitter.

        By the way Horace, I find it refreshing that you take the time to respond to posts. Keep up the good work.

      • http://www.asymco.com Horace Dediu

        The phrase about the drill comes from the famous quote by Theodore Levitt “People don’t want to buy a quarter-inch drill. They want a quarter-inch hole!”
        His point is that what people buy is often not what companies sell.

        When I said people did not search I meant that they did not use search engines. I remember it was around 2003 and many people were surprised when I suggested they use something called Google and type “search terms” into a browser window. Over time people have gotten used to the idea of search and queries but it does not mean it’s natural or cannot be improved as a way to find whatever you’re looking for (including love, affection, peace of mind, etc.)
        There are underlying jobs that search is hired to do right now but I think it’s a very poor performer. I’ll write more about this later but I expect at some point we’ll have to explain to our children what “search” using something called “a browser” meant.

      • Ab

        Yahoo answer to the problem of unstructured data within websites was with lists and categorising pages and building up a catalogue of websites.

        As the number of websites exploded we needed something more automated and along came Google.

        The irony of search is a) we are creatures of habit and visit the same sites often.
        b) We only tend to click on the first page.

        Also search is a means to an end and not the end in itself.

        I agree it can be improved but I don’t currently see how Siri will improve the process of indexing of explicit data – its a UI at the end of the day.

        Siri can draw on a wide range of data sources but the data sources themselves remain undisrupted.

        Take for example Iron Man and his interaction with his digital assistant – there is no explicit use of search its we need a solution to xyz and the assistant will scour its data sources for possible answers it can collect together to provide a coherent response. If Google is the best source of data and its not used then the likelihood is that the response is not as complete as it should be and will diminish the interaction no matter how fluid and seamless it may seem. If I want upto date stats I don’t want those from a year ago.

      • http://www.asymco.com Horace Dediu

        I should mention that my career began in 1989 doing research (at GTE Laboratories) in the area of information retrieval–something that came to be known as search. So I was familiar with the concept before the World Wide Web even existed. We searched full text data and even considered using hypertext markup before there was HTML. We dreamed of a future where people did not use database front ends to find information and our approach of free association search (based on a weighted criteria) was considered a bit weird. (You can search papers I wrote in Google scholar…)
        I’m very fond of search and my background doing research in the area before AltaVista and before Google and before Yahoo convinces me even more that search is but a stepping stone in our quest to get computers to be helpful.

      • Ab

        http://www.engadget.com/2011/11/07/siri-may-pose-competitive-threat-to-google-eric-schmidt-tells/

        Schmidt says Siri is a threat. I think this has more to do with him using any justification against an anti-trust investigation into Google monopoly power on search.

        He is talking up his competition but then dismissed it a year ago.

        Going forward what you will see is Facebook and Apple encroaching into Google territory similar to Goole encroaching into FB and Apple territory.

        The real question is – is this good for consumers and are these companies effeciently using their resources?

      • El Aura

        Depends, some people want the hole, some people want that feeling of being able to drill a hole whenever they want. They buy the tool because they like owning tools.

      • GeorgeS

        “They buy the tool because they like owning tools.”

        How true. The guys on Car Talk (NPR program) often suggest that a caller fix something himself because it’s an opportunity to buy more tools.

        The next time you’re in a “home improvement” store–a poor substitute for the true “hardware store,” watch different age people. The younger folks, especially couples, buy paint, light fixtures, tile, closet organizers, etc, obviously improving their homes. The middle-aged men go slowly through the tool section, fondling the routers and reciprocating saws and playing with the tape measures. The old farts, like me, are found in the “hardware section,” looking totally befuddled as we try to find a match to the screw/bolt in our hands.

      • http://www.asymco.com Horace Dediu

        Indeed. This is what tool makers count on. Trouble is there are more people interested in the hole and they are not being addressed by tool makers.

      • Ben Rosengart

        I didn’t know cows cried.

        Oh, that kind of calf tears.

      • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

        This is a moo-t point.

        Sorry, I couldn’t resist.

    • http://www.asymco.com Horace Dediu

      I have no doubt that Google will attempt to copy the functionality of Siri but what would be the impact on Google’s business model for an interface that takes users away from advertising. Perhaps that is why Google bought Motorola. To have some way to make some money from mobile computers.

      • Walt French

        Google has doubts about the ad-supported model? Unlikely. It was the major disruptor that allowed radio and TV to dominate the mid- and late-20th century. It’s totally supported them for their entire existence. It was their disruptor that allowed their entry into mobile. I wouldn’t think they COULD go upmarket to paid services even if they wanted to, unless they first get monopolies, so I could envision their dropping ads only with an even more powerful disruptor, one that I don’t have up MY sleeve.

        Perhaps they could have ads on notifications, some of which might appear unbidden. As long as a user is interacting with a device, there’s a chance to steal a few CPU and eyeball cycles for a brief commercial interruption. Or perhaps they work a deal with the carriers to share revenues on every ad that runs in front of a YouTube video, allowing lower-cost data budgets on Androids.

        These don’t directly address Siri, but then again, it’s not clear that a direct assault would be especially effective. Android already has a market position where they’re perceived as the lower cost provider of all of the good enough mobile services; they can simply leverage that and let users think that Siri costs $25/month just to ask it what the temperature is.

        Hard for me to gauge how well that’d work. I’m of the elitist persuasion; I’d much rather pay a couple more $ per month to be spared the aggravation of wasting my precious time on ads. But my preferences and the majority’s preferences could be pretty far removed. There must be no less than sizable minority that’d vote for free.

      • http://twitter.com/andyidsinga andyidsinga

        hmm. google has had voice search for a while in the google app.
        I’ve been wondering if google will try to meet apples siri challenge by adding logic to make the google app talk back to you. they certainly had some of that with goog-411.

        cheers

      • Ab

        You are confusing interaction ie voice with data. I may inquire by voice but actually want results returned as a data set which I can quickly scan. Imagine having the results read out – that would be time consuming and inefficient. The data set returned will still have a list of organic results and paid for adverts.

        Google bought Motorola because of the patents than any overiding need to get into the hardware game.

        Voice interaction removes the barriers between me and what I want to do but it doesn’t improve the data I need.

      • http://twitter.com/tim221175 Tim W

        The answer is:

        User: How much wood can a woodchuck chuck?

        AndroidSiri: Just a minute I will find that for you, in the mean time please listen to these words from our sponsor … recorded audio ad …

      • Anonymous

        I agree that copying Siri could undermine Google’s business model, but I don’t think that is the fundamental obstacle in Google’s way. The big question is IF Google CAN copy Siri; at least in any reasonable length of time. I doubt it.

        Siri did not spring full grown from Apple’s loins, but rather was the product of one of the longest running, best funded and most advanced AI program in history. Had DARPA and SRI not invested many years and hundreds of millions of dollars in developing her (Notice: I unconsciously called Siri a “her”), then Apple could not have created it.

        It is highly unlikely that Google can clean room one of the most powerful AI’s on the planet, bolt it onto Android and make it into a product that is both competitive to Apple and appears on the market within this decade. Sure Google has a damned good voice recognition product, but it is NOT an AI and and there is a huge difference. Of the two, a voice recognition system is easy compared to an AI.

        The killer factor for Siri is the back end training capability that comes with an AI. It will learn idioms, dialects, slang, non-standard usage and will figure out the nebulous contexts that they all swim in and it will learn it in real time from a hundred million users. Simple voice recognition cannot scale like that and without that training capability Android will be hanging onto the tail end of the exponential curve describing iPhone sales.

        I suppose that Google could do the same as Apple and buy an AI progrom somewhere…but where? I sure don’t know of any that could work. IBM’s Watson is a first thought but Watson’s architecture is so far from Siri’s that making it work in a telephone environment would take years in itself.

        I don’t mean to underestimate Google or it’s abilities, but the question of Google sacrificing its ad revenue for a Siri knockoff is a moot one. I don’t think it can be done, apple is too far ahead with a product that improves itself with evergy iPhone sold and there is no known competitor that can be drafted into action in market relevant timeline.

  • Per

    I believe Siri is the start of something even bigger than another disruption of the mobile/computer industry.

    I just want to point you to an article by futurist John Smart, who has written an essay called “The Conversational Interface: Our Next Great Leap Forward”, where he suggests that what he calls the “Conversational Interface” (CI), is likely to be ”the most important enabling information technology development and collective intelligence advance on our planet in the next thirty years.”

    He also writes:

    “Let me again go on record proposing that the conversational interface
    will be the single most important technological innovation the average
    person alive today will witness in their lifetimes, out of a very long
    list of competing innovations, like personal computers, the internet,
    automated supply chains, credit cards, ATMs and cell phones.”

    Note that this was written before the release of iPhone 4S (the essay is dated 2003, updated last time 2010), but he refers to Siri and the company SRI International, and suggests that Google should buy Siri. He also thinks that Google is on a good way of creating the new conversational interface, based on the increase in the average number of words used in a typical Google search.

    My own two cents is this: I do not think that Siri necessarily should be seen as a real conversational interface, but it may very well grow into one. And I also think that Apple accelerated the development in this direction by integrating Siri into the Iphone and that Google and Microsoft and others will surely respond, creating intense innovation towards the conversational interface. Finally, it would not surprise me if Google already have been working in this direction, and may soon have something to show.

    It sure is an interesting development.

    John Smart’s essay can be found here:
    http://www.accelerationwatch.com/lui.html

  • http://twitter.com/andyidsinga andyidsinga

    I love love love your list of what siri has going for it!!! especially #s 2 and 3

  • GeorgeS

    The discussion has been interesting, enlightening, and intriguing. However, most people seem to have overlooked something: how Siri affects the user. Not what data or functionality it provides, not even what job it’s hired to do, to use Horace’s favorite concept, but how Siri makes people FEEL. People seldom buy or use a device (or make almost any significant decision) based upon calculated utility but because of how it makes them feel, especially about themselves. See Simon Sinek’s TED talk and book, “Start with Why.” Sinek uses Apple as one of his primary examples of a company/organization that understands WHY it exists.

    If you want concrete evidence, just view the videos about Siri on YouTube, posted by users. My favorite is:

    Listen to the young woman as she interacts with Siri. She laughs, she is impressed, she is enthralled. She sums it up at the end: “Oh, my gosh! I LOVE this phone!” THAT will sell many, many more iPhone 4Ses than creating events in the calendar or sending text messages.

    • Davel

      Thank you for the Ted reference. I learned something today.

    • Walt French

      Spot on. I was listening to a Christensen speech recently about “the job that milkshakes were hired to do” and along your lines, realized that people will hire Siri to entertain and quasi-befriend them. While I think Siri, in “beta” form is already pretty amazing, I think that Apple will have to work hard to expand its domain quickly enough to catch up with users’ expectations and wants.

    • 4s Fan

      George, I agree completely. My wife hated my old iPhone 3G, because it seemed “cold” to her. I talked her into getting a new 4s when I did, and Siri is by far her favorite feature. The delightful nature of it is not to be underestimated. Plus, she hated the old keyboard, and while its much better now, being able to circumvent it is huge. She will never switch to another platform now.

    • Anonymous

      Eh. The girl in your video found the list of “humorous things you can get Siri to say” then went through the list on video, laughing at each one, complete with obvious cuts, ending with “I love this phone.” Meanwhile, she appears like she goes through life laughing at everything and generally being positive. Nothing wrong with that. But for you to pretend it’s related to the phone is kind of silly. Perhaps you should look through the “related videos” about Siri on the right some time. Though I’m sure you have. You cherry-picked this one of course.

  • Steve Rosenberg

    Makes sense Horace. Pulling from the same speech, the problem with circa 2007 smart phones were:

    “not so smart”
    “not so easy to use”
    “the buttons and controls cannot change”
    “each application requires a slightly different user interface”
    “who wants a stylus? Yuck!”

    The solution required a large screen, a touch interface and great applications. Eventually the first two took off but the third made the difference amongst the copyists. I would argue a revolutionary user interface is a minimum requirement for disruption. But other feature are EQUALLY required to maintain the lead. The bandwagon will follow quickly with Siri and I am sure te competition get it. What added layers of defense can Apple deploy? Exclusivity for 1-2 years with each of their database partners. Another great use of cash.

  • http://skypejournal.com Phil Wolff

    typo: “deponent”, should be “dependent”?

  • http://dantech.tumblr.com Daniel Tung

    I would just like to mention one minor point: is SIRI best used on a phone?
    I mean, if you’re talking to friends over the phone, and you want to immediately ask SIRI to check a date for appointment. This is obviously impossible according to the current implementation of SIRI in iPhone 4S.
    The phone is a communication device, SIRI is primarily an input device, when these two functions collide, one of them has to be sacrificed.

  • MOD

    siri is an interesting toy.

  • chiefthinker

    How long do you think before Apple will buy a few dozen or even hundreds if IBM “Watsons”..or have they already? Siri + Watson would be a giant leap in AI and UI and certainly remove any “toy” labels.

    Any thoughts on this?

    • http://twitter.com/Marcos_El_Malo Marcos_El_Malo

      For breeding purposes?

      • berult

        Artificial Intelligence through Artificial Insimination. AI times AI = I’SIRI’BM bliss …”elementary my dear Watson!”

  • Jeff Dutky

    I suspect that the Apple Timeline is a little self-serving, and concentrates too much on recent “advances” at the expense of the broader historical perspective (e.g. I’m not sure that the click wheel really merits inclusion in the timeline, and with only two other data points, in severely distorts the apparent progression). For a better historical view, let’s consider the development of major user interaction modes on computer systems since the advent of interactive systems: the first interactive computer systems were developed in the mid-60s (e.g. TOPS-10 in 1967) and used paper TTYs for the user interface with a command line interpreter. The next big step was in the mid-70s with the introduction of “glass TTYs” like the IBM 3270 or DEC VT100. Then we have the introduction of the GUI in the mid-80s (though the basic elements of the GUI date all the way back to Doug Engelbart in 1968). Pen computing appeared in the early to mid-90s. Finally, multi-touch interfaces appear in the mid-2000s.

    With this timeline it appears that a new user interaction method appears about once every decade, with no discernible change in frequency over the past 40 years, or more. We might argue about WHY the period remains constant (e.g. is it a problem of human psychology and the ability of people to perceive the shortcomings in current paradigms? Does it have something to do with a built-in logarithmic bias in the human perception of performance improvement?) but the assertion that the rate of change of interaction methods is increasing is, at best, questionable.

    • Anonymous

      “…let’s consider the development of major user interaction modes… Pen computing appeared in the early to mid-90s. Finally, multi-touch interfaces appear in the mid-2000s.”

      I think there have been more UIs. One is automatic speech recognition that’s widely used in call centers where agents receive calls from customers needing assistance. AT&T started using these in 1992, and they routinely process 1.2 billion voice transactions every year with these systems [1]. Another is optical readers; one application is fingerprint sensors for unlocking computers, cars, bikes, etc [2]. Laptops had these in the late 1990s. In the mid 2000s, research started on near field communication. I’m not familiar with this, and I don’t know how widely it’ll be used.

      The last is the conversational interface (CI) where natural language speech (not keywords) is used. With the CI, people can speak sentences that are more than 5 words. English speakers use 5-11 words in spoken sentences and 8-14 in written when we ask each other complex questions [3]. The computer analyzes the speech and understands the user’s intent. The conversation continues until the request is explained, and then it’s performed [4].

      With Siri, you can begin by saying, “I want a romantic place for dinner tonight,” and it will understand that your intent is: a reservation at a highly rated romantic restaurant. In addition, it knows that you might have a preference for which rating agency is used. You can continue the conversation by adding more qualifiers like the type of food and the location, and when you’re done you’ll have a reservation [5].

      I don’t know if optical readers and NFC are revolutionary UIs, but if they’re included them in the timeline then we see new interfaces created faster than once every decade, and the rate of new ones is increasing. It makes sense that this would happen because processor and memory chips are so much cheaper and smaller now than 30 years ago so the building blocks are there for this new technology.

      1. http://www.ece.ucsb.edu/Faculty/Rabiner/ece259/Reprints/354_LALI-ASRHistory-final-10-8.pdf
      2. http://fingerchip.pagesperso-orange.fr/biometrics/types/fingerprint_products_notebooks.htm#NEC
      3. http://www.accelerationwatch.com/lui.html This source is from Per. His comment is about 12 down down from yours.
      4. http://www.quora.com/Adam-Cheyer/answers
      5. http://ontolog.cim3.net/file/resource/presentation/Siri_20100225/Siri–An-Ontology-driven-Application-for-the-Masses–AdamCheyer-TomGruber_20100225.pdf

  • http://twitter.com/drostyboy steven

    horace – on a slight, hopefully relevant tangent: playing around with siri, i just realised that siri has the ability to translate what it learns about you. switching from english to german, siri understands that “my brother” and “mein bruder” refer to the same person. hitherto for multi linguists (who really use different languages constantly), switching keyboards brought you the advantage of language specific alphabets and spellcheckers, but siri’s “fuzziness” (“brother” is more fuzzy than a specific name, say) coupled with implicit translation means that it is “asking less” from you and “doing more” – surely a ui boost?

  • Pingback: The Future (Holger’s Talk) // UX Café()

  • Pingback: The Global Smartphone Market Landscape | asymco()

  • Pingback: Two Visions of Input Devices - Behind Companies()

  • Pingback: Re-imagining Apps for Ultrabook™ (Part 1): Touch Interfaces | ServerGround.net()

  • Pingback: 2012 - The Year of the Interface | Digital Tonto()