Apple’s New Siri is Playing Catch-Up to Google Voice Actions. But in a Very Human Way.

Apple’s New Siri is Playing Catch-Up to Google Voice Actions. But in a Very Human Way.

As far too many people have pointed out already, Apple’s event this morning was a bit of a disappointment — as a spectacle. 16 months is a long time to wait for an incremental hardware improvement, which the iPhone 4S unquestionably is. But there was one announcement amid all the recap of iOS 5 and iCloud features that should have been tremendously exciting to anyone who cares about the future of interactions: Siri, the voice-activated assistant exclusive to the iPhone 4S.

As always happens when Apple rolls out a major technology (in this case, e-concierge services), critics are happy to point out that Cupertino is late to the party (can you believe that it took them 5+ years to respond to the Treo?!). Specifically, they’re calling Siri a catch-up effort to match Google Voice Actions technology that’s been available on Android for well over a year. Having used Voice Actions for awhile now, I can confirm that this is half-true. On a feature-by-feature basis, Siri looks me-too. But from an experience standpoint, it’s totally different. As usual, Google’s implementation is process-oriented. Apple’s, unsurprisingly, is human and friendly. And this is why Siri has the potential to be revolutionary.

Before we go any further, take a minute to watch how the two companies pitch their respective products:

Google

Apple

What do you notice? Superficially, the biggest difference is that Google emphasizes that what you’re talking to is a piece of technology. Your commands have a unified structure they need to follow. For example, you would say, “Call <name of user as written in address book> at <location>” in order to place a phone call to anyone, regardless of who that person is to you. All people are treated the safe. Siri, on the other hand, makes things as personal as possible. Depending on the person’s relationship to you, you can use more friendly language, like “Call my wife.” It feels a lot more like talking to a person, which makes the conversation both more natural and more magical.

Or take something significantly more straight-ahead and functional: weather. On Google Voice Actions, you talk like you’re dictating a Google search: “weather San Francisco Oct. 7’. And then it displays the google search results that appear on a web page. That’s about it. On Siri, you say, “What’s the weather going to be like when I’m in San Francisco? How about Napa.” And the app responds in fairly natural form while also displaying the information it’s talking about in visual form.

From a functional standpoint, this is identical. From an experiential standpoint, it’s worlds apart. Google wants to turn your voice into a keyboard and mouse. Apple wants you to have a different kind of relationship to your phone that looks a lot like your relationships to people.

There is good reason for this. Virtually anyone who might use Siri has experience with (terrible) voice recognition software, including Apple Voice Control, the inept tool included in iOS 3 that was basically Google Voice Actions except wildly inaccurate (especially for recognizing Indian names) and capable only of placing phone calls and playing back music. I think it’s safe to say that most of our experiences look a lot like this. Apple has chosen to make Siri respond less like typical voice recognition and more like a helpful human assistant in part to avoid some of our negative associations with the underlying technology.

It’s a good approach. Tellme, the startup that Mike McCue founded prior to Flipboard and a Microsoft technology that seems at least partly responsible for the voice side of Kinect’s magic, used human-sounding voice interfaces with pauses and complete sentences to make customer service lines way less terrible. That kind of power can now be processed on a phone in near real time.

That’s what makes me so excited about Siri — it isn’t set up to be an improvement over the crummy consumer voice recognition that we’ve all been exposed to over the last decade. It’s taking technology and intelligence that previously required brawny hardware and putting it in the palms of our hands. It’s the Knowledge Navigator. Finally.

Related
  • Adam Kozlowski

    I’m really trying to use Google Voice Commands, but it’s so flaky, even on a brand new HTC phone. Really, the mistakes it makes are way funnier than iPhone autocorrection jokes.

  • Anon Developer

    I just cant see myself talking to a machine any time soon. Is there an option to type in the questions / commands or does it have to be vocal?

  • Marcel Ve?ký

    very good article..the siri tech has its own potencial to be experienced..

  • Dion

    I agree it looks much better than google voice actions but I don’t agree that it represents any significant difference in approach. It seems fairly obvious that the end game of any voice action software is to become more like human interaction and I’m sure that’s what google have always planned for google voice actions. Kudos to Apple for appearing to get closer, first, but there’s nothing revolutionary about the principle or approach.

  • MacHead

    Yes its called google inside safari, the point of this is not to type. You can already do everything it does by typing or searching your calender and set reminders yourself. Its obviously not something youll use in a crowded room but when your at home itll be a massively convenient feature

  • ddmcd

    Your  “fair and balanced” comparison of these two approaches is not going to make people who want a “tech food fight” happy.

  • yoyogipark

    “16 months is a long time to wait for an incremental hardware improvement”
    That’s idiotic. The improvements are no less significant than other upgrades, and if you expect Apple to reinvent the phone every year you’re naive. There’s irony in that while the criticism is all about the insignificance of the improvements, 90% of those criticizing would be satisfied if Apple had put the same improvements in a new plastic shape and used the name “5″ instead of “4s”.

  • davidk

    Dissapointing?!? What were you expecting, a free unicorn with each phone? You’ve got a phone with a dramatically faster processor, twice the storage, top notch camera, significantly improved antenna design that allows 4G range speeds over 3G, the best looking screen on a mobile device AND Siri. Speaking of Siri, it’s not like Google Voice Commands, something even classic MacOS could do, a fixed, very exact voice control set, it’s a leap beyond! Natural language voice control with context recognition. Apple announced the iPhone 5 yesterday, they just didn’t call it that.

  • cliqsquad

    Maybe not to the full extent of the demo you won’t but I bet you will use it for basic things. 

    Need to set up an appointment, or an alarm. Are you really going go to the app now that you don’t have to? Just hold the home button and tell Siri to set the alarm for 8AM. Get home late and very tired really want to set the alarm on your phone? Now it’s a piece of cake. 

    I can’t wait for iOS 6 when the APIs will most likely go out to developers to integrate into their Apps. Really not gonna love this feature on Navigation or email?

    Imagine one day voice recognition to unlock your phone. 

    William: iPhone unlock, 
    Siri: Hello William, iPhone unlocked 

    or 

    Siri: Hello William, you have three app updates and a new voicemail from John & Eric, would like to listen them now. 

    William: Play both or Yes, Eric’s first, John’s only 

    This is could be the early days of something truly epic, things we seen in movies. Like Marty’s smart house in Back to The Future ll, but in the palm of your hand. 

  • Michael Ingraham

    “That kind of power can now be processed on a phone in near real time” – according to the Siri guys, it’s not being done on the phone but rather on the cloud. This is how the voice recognition part of Android takes place as well. It will be interesting to see if Google improves Voice Actions to become more aware of “natural” language.  My question is whether the iPhone/iPad context has the ability to be told or learn about relationships (“my wife”, “my dad”) or is this the same voice recognition as having a contact in my address book titled “Wife” or “Dad”? 

  • MrSarcy

    There is no way in hell this is going to work (and I love Apple, don’t get me wrong). But just like in the ads where people never place a finger wrong on the screen and they edit out wait times for apps to load, this is going to be nowhere near as easy as the ads make it seem. We’re at least 20 years away from a world where computers understand what we say and then do something useful. Best we can hope for is voice recognition that is marginally better than having to type. Mark my words: this will be a huge disappointment. ‘Call Dave on his cell’ indeed… And then the damn thing calls the wrong Dave on the home phone number because you didn’t meticulously categorize all numbers for your contacts…

  • dwplay11

    The more i look at it, it seems like google comes up with half-baked features just to be first. There implementation on almost all there features are junk. Apple set there iPhone 4S release 4 months late because of Siri. Google’s way would of been, lets integrate it now and maybe update it later!

  • VGISoftware

    I’m marking your words as needlessly skeptical and negative.

  • VGISoftware

    Who cares if it’s “revolutionary” or not? What matters is both the impact and iPhone sales AND the impact on people’s lives.

    I know our lives have been significantly impacted my our iPhone 3GSs. Now we’re ready to experience these significant improvements.

  • VGISoftware

    Conservatism is overrated. It says, “Stop!” while being “clothed” in “acceptable behavior.”

  • Enough

    I’m marking them as just needless

  • Mahesh

    Good one Inghram , I just thought the same. How come a mobile know who is my wife? If I haven’t stored any one under the contact .  I felt that, it is just a fan boy’s expectation rather than the real technology that is in hand.

  • Bbq007

    Siri should ask you which dave if there are more than one.  You say the last name and it calls the right one.  Even if you didn’t use speech recognition and have multiple dave’s you always look for last names don’t you?

  • God

    You would want the facts to get in the way of the fits people feel they deserve to throw.

    ‘but i wanted an iphone 5…i like the number 5 and i won’t be happy until Apple gives me exactly what I want, when i want it. If they don’t they are just shit and I’ll continue to say this until the rumors of iPhone 6 come along. Then I will only be happy if they call it iPhone 6. Don’t even think of calling it iPhone 5 because I say so.”

    The fanboys don’t care about function. They are just angry that if they get the 4s people on the street won’t be able to visually see they are the ultimate fanboy with the lasted phone. Someone might think they still have an iPhone 4. Gasp! Move on to other devices like you are all saying you will. I for one will be happy that I won’t have to hear your constant complaints anymore.

  • dagamer34

    All this pointed out to me is that Apple’s synthesized voice just sounds wrong to me. We are still several years away from voices that don’t sound like a robot. Until then, it’s going to be creepy to use.

  • Hampus

    True, but at least it isn’t MS Sam :p
    Apples synthesized voices are some of the best out there right now.

  • Hampus

    Indeed, that’s how Google work, the release as soon as they can and then they tweak it over time, Google just love releasing beta services :p

  • Hampus

    Well, according to yesterdays keynote it can indeed learn such things.
    During the demo Scott said “Remind me to call my wife when I leave work”, the reminder created by Siri said “Call Molly”, after that Scott explained that Siri had asked during an earlier conversation who his wife was and now it knew.

    It’s probably sensitive to several relation related words, when it detects mom, wife or girlfriend and doesn’t know who that is it’ll ask and you can probably just specify a contact.
    Heck maybe you can just say for example “Molly is my wife” who knows :p

  • Dreezy

    This is NOT a “fan boy’s expectation” as they used the wife example in the Siri demo yesterday at the announcement.  I am assuming your relationships with your contacts will be defined through a field added to the contact information.  If not, it will ask who your wife is the first time you make this request, and store the information for future reference.  You will not need to save the contact as “wife”, I can almost guarantee that.

  • Guest

    I simply don’t get it. Why can’t Apple fanboys just relax and say, okay fine Apple copied a couple features from Android/Windows Phone 7/etc.

    Why all this nonsense on how the Apple way is more user friendly? They copied the basic functionality right? Nuff said …. end of the story.

  • ElVox

    You didn’t read the article, right? He *is* accepting that Apple is copying the basic functionality…but that is not the end of the story, there’s more…Apple doesn’t just copy, they make it their own…in this case, by making it all natural-language instead of computer commands…and that’s the killer feature.

  • Tim_at_IMM

    > We’re at least 20 years away from a world where computers understand 
    >what we say and then do something useful

    I think you’re wrong about this. You’re looking at things from a linear growth perspective regarding computers while they are actually going through an exponential growth phase. I’ve been following the predictions of futurist Ray Kurzweil, and this is very much in line with what he’s been saying. If you’re not familiar with IBM’s Watson, I suggest you check out the Nova episode which is available free on the Nova’ website. 

    Because of their small screens, voice control of our smart phones and other electronic devices is becoming more and more important. Of course, Apple would be a major innovator and trendsetter in this regard. Next, watch for all phone makers to come out with “intelligent assistants” deeply integrated into the operating systems of their products, followed by a wide variety of other devices.

  • Kezz

    Does anyone know why Siri will not be available in Canada?  Language is not the barrier because Siri is being made available in US, UK, France.  So there must be another reason.  Licensing? Carrier restrictions?  Does anyone know?

  • pangeomedia

    Why is iPhone 4S an ‘incremental hardware improvement?’ Seriously, why? Are not, by definition, all hardware changes merely ‘incremental’ until the next hardware change? iPhone 4S is completely new inside. World phone. A5 CPU. And so on. If Apple had simply renamed the phone iPhone 5 would people be grumbling? No. Apple gave us exactly what we expected except the iPhone 5 name, and thinner design. And I don’t want thinner. 

    The disappointment in the iPhone 4S by pundits and followers is self inflicted. Get over yourselves. Name another phone on the market that is better (keep the comparison to both hardware and software) than iPhone 4S? It’s been almost five years since the iPhone was introduced and the latest version, iPhone 4, is the best ever, sold more than all other iPhones put together, and is loved en masse by users. The iPhone 4S improves on that. Double speed. Faster downloads. Faster graphics. More memory. Awesome OS upgrades. 

    Get over it already.

  • TheDigitalPimp

    You know, we’re all some kind of pundit savants, having not used the damn feature yet, and making wild predictions of success or failure. Somewhere in the middle is the truth. If you watched the live demo from the presentation yesterday, anyone could see this seems very different and very real. But time will tell if it helps open up the device for a new class of customers. You know, the ones that don’t care about specs & crap, who just want to get things done.

  • TheDigitalPimp

    Totally agreed. Would the analysts & fanbois be happier if they put the same specs into a new case? I have been having that argument with all sorts of people who think this was a let down, when it’s a big leap in core hardware. The damn phone is already #1 by a wide margin, so what’s the rush? Not like any other OEM will catch up before they release an iPhone 5678.

  • Greg Smith

    my biggest complaint is why siri still has to sound like a robot.  With all that technology, shouldn’t they have been able to make the voice sound human? …or is that harder than it sounds?

  • TheDigitalPimp

    Yep. It’s hard to simulate a natural sounding voice, and Apple’s are very good. Takes more CPU than we think (read an article on it long time ago when OS X debuted)

  • John Neumann

    I had the chance to play around with Siri (the original app, since removed from the App Store) on my 3Gs and even in its hobbled state it is surprisingly quick and user friendly. Last night I was trying to find its flaws and when I asked for a cigar store, I got the attached exchange. 

    I had never input my name into Siri, it figured that out for itself after a few searches and offered some healthful tips in regards to my choice of store. Scary? Fun? certainly impressive.  

  • John Neumann

    The only implausibility I see with the video posted above is the use of the monster-ear-canal-sized Apple earphones. I find them useless and I never see anyone wearing them. Other than that, I like the mic on the wires but if you cannot wear the ear pieces, what good are they? 

    Ives, c’mon! Make some ear phones that WORK!

  • Pauly

    Actually I don’t think that is it.  I think it was telling you the type of language it understood.  Which is England (both American English and British English) as well as French.  I don’t think it means it’s not available in Canada.  Or at least that is how I understood the information and what I hope. 

  • markrlangston

    Apple also makes an excellent (albeit expensive) set of in-ear buds with the same functionality as the ear buds.

  • markrlangston

    Amen to that!!

    I’m glad someone out there doesn’t expect Apple to reinvent the wheel every 12 months. The fact that some people are calling the iPhone 4 design archaic is 100% insanity. So the most well crafted and best sold mobile phone in the world is already old news after 16 months. 

    By that logic had they released the iPhone 5 people would be expecting Apple to release and even thinner, faster, lighter, whatever-er device the following year and again, and again, and again.

    No other tech company has the expectations that Apple has. 

  • Daniel Harris

    I just find it hard to believe Siri will really be able to consistently understand people just talking like people. That’s an incredibly difficult thing for a computer to do. Demos won’t convince me it really works because obviously they are only going to show the times that it works. I will have to put it through the paces myself with real world use, or at least hear from others who have.

  • Kezz

    Seems there is a lot of confusing information on the web right now about Siri’s availability / functionality in Canada. Apple Canada’s promo material (inc. video) on their website has had all Siri references stripped from it, which I find incredibly odd if this is listed as one of the key selling features of the new phone on the US site. If what’s available in Canada is simply not a customised Canadian version, but still works if you choose “US English” or “UK English” as your language, why would the selling feature be completely stripped from all Canadian promo materials? Makes me question it’s availability in Canada when we get the 4S on Oct 14th.

  • Barton Lynch

    so it’s the google one only a little bit better. not really a killer feature to me. 

  • Pauly

    Hmmmmmmmmm.  That IS confusing.  And if true rather disappointing.  For a phone loved the WORLD over, it sucks to keep finding content locked to the United States.  I mean I get it from services like Pandora and the like, but from Apple….

    I guess we’ll just have to wait and see what pans out. 

  • sir1jaguar

    Hey stupid,

    Go to best buy and if they have tmobile version of NEXUS and download the 2.3.6 gingerbread then use the google voice recognition and SIRI is just hype that was bought for $200 million, apple didnt invent it.

    Lots of itards defending apple…

  • Eh is not for Apple

    Maybe Siri gets confused with the word “eh” eh?

  • Clydehidershaj

    actually with “voice actions” app it gets pretty personal. Free download.

About the author

Pete Mortensen

Pete Mortensen is a design strategist for consulting firm Jump Associates and the co-author of Wired to Care: How Companies Prosper When They Create Widespread Empathy, a book and blog that are significantly more interesting than you might initially think. Pete's particular Apple avocations are both around design--interface and industrial. Follow him on Twitter!

(sorry, you need Javascript to see this e-mail address)| Read more posts by .

Posted in iPhone, Opinions, Top stories |