Oct 13 AT 1:29 PM Taylor Wimberly 54 Comments

Google’s response to Siri might look something like this

Speaktoit for Android

Yesterday Apple finally released their iOS5 software update for the iPhone 4 and the blogs are buzzing about Siri, a virtual personal assistant that lets you use your voice to send messages, schedule meetings, place phone calls, and more. Earlier this month I received a tip that led me to believe Google would release a similar offering around the launch of Ice Cream Sandwich and ever since I’ve been trying to dig up additional details about how their product might differ.

For starters, I have been unable to get anyone else to confirm this rumor to me (besides the original source). I asked around at CTIA if anyone had heard about this secret project, but I came up empty. However, Google has a history of surprise announcements and I still believe in the original information that was shared with me.

Google’s speech technology efforts are being led by Mike Cohen, who joined the company back in 2004. Prior to joining Google, Mike spent 10 years at Nuance Communications, which he co-founded. That’s the same Nuance that actually powers the technology behind Siri. We don’t talk about Mike much but he has helped Google launch Voice SearchVoice Input and Voice Actions for our Android phones.

Voice Actions for Android is the closest product that Google has to Apple’s Siri and that’s what we heard was getting a major update. One way that Google’s assistant might differ from Siri is the addition of an animated avatar. I previously compared it to something like an intelligent version of Talking Tom Cat, but then I discovered a better example called Speaktoit Assistant.


Speaktoit is similar to Siri in that it supports natural language so you can converse with the virtual assistant in a normal human speaking style. The Android app also includes similar functionality that can “send emails, send texts, look up information, post to Twitter, check you in places, update your Facebook, find news, look up traffic, look up weather, call people, take notes, add things to your calendar, translate foreign languages, help you find nearby places like bars, and tons more.”

Natural language support allows you to talk like a human instead of a robot and I was surprised with how well it works. Of course, the responses coming back from Speaktoit are still in that robotic sounding voice we expect from our phones.

Google’s virtual assistant should solve that robotic voice output thanks to their acquisition of Phonetic Arts, which occurred at the tail end of 2010. The company had previously been working on speech synthesis in games, but Google was after their technology that could convert lines of recorded dialog into a speech library. The result was surprisingly realistic automated voices that sounded more human and fluid.

I’m not aware of Google incorporating Phonetic Art’s technology into Android yet, but the acquisition happened right around the time that Android engineers started working on Ice Cream Sandwich.

We already expect the next version of Android to include some kind of facial recognition and tracking technology, so Google’s virtual assistant could turn out to be really creepy. Imagine a virtual character that you talk to in a human language, it responds in a human sounding voice, it scans your face to detect your emotions, and it knows everything about you.

I don’t know if the average consumer is ready for this kind of futuristic artificial intelligence, but Google has the technology to pull it off. As Mike Elgan of Datamation puts it, it’s all about the data and Google has plenty of that. “Google already knows who your niece is and what her interests are. They know your web browsing history and what marketing people in your area are both looking for work and highly recommended. Google knows what you say on email, where you go and what you like to read.”

Hopefully, we should find out more in the coming weeks as Google unveils all the new features of Ice Cream Sandwich. Are you ready for a human-sounding virtual assistant that knows all of your private and public data? Download Speaktoit Assistant and let us know what kinds of features you would like to see from a Google-powered virtual assistant.

Taylor is the founder of Android and Me. He resides in Dallas and carries the Samsung Galaxy S 4 and HTC One as his daily devices. Ask him a question on Twitter or Google+ and he is likely to respond. | Ethics statement

    Most Tweeted This Week

  • http://mobileroundup.com TrevFMT

    I sincerely hope not. I’d much rather not have a “cute” avatar when working with the assistant. Apple did it right. The only time I’d want an avatar would be on Windows Phone with my XBL avatar.

    • JPB

      I just downloaded and played with it for about 15 minutes. It’s cute if rough around the edges. Yes you can personalize your avatar or just replace it with a microphone icon (which is what I did). You can also use your own background or one of several others.

      And to answer my own question, yes there is a conversational mode so that you don’t have to keep hitting the voice key.

      It really DOES have the look of something like a Google App, right down to the generic name it gives the app after it is installed (“Assistant”). One nice surprise is that it seems to integrate with Evernote (“Note that blah, blah, blah….”) . Nice touch.

      • Ryan

        I’m amazed at people who don’t read (or even fully skim) entire articles. This is NOT Google’s version, this is a version a third party made for Android.

        • Paul

          He did read the article, he stated that it appears just LIKE a google app. Nowhere did he say it was one.

      • ddpacino

        Yea, I tried this out for myself and was very impressed, especially for a third-party app. I haven’t been this impressed with an app since I discovered IntoNow. It seems capable of doing about 70-80% of what Siri can do — from many of the test images I’ve seen of it. And it is a little rough around the edges, but I’m really liking this app.

        I liked and found that Vlingo worked a little better than Google’s Voice Actions in my testing. And I think this does a much better job than both of those at performing useful tasks on your device with just your voice. If Google hasn’t been working on something like Siri, they should definitely pick this up and polish it off. It’s pretty damn good, and can only get better with some Google integration. She even had a name:

        Me: Hey Sam, go purchase J.Cole’s new album.

        Sam: Purchase complete. Downloading now…

        Me: Now play it in my living room [via Android @ Home integration].

        Sam: Streaming it to the receiver now… ♫ Music plays ♪

        I cannot wait until I can do that. No going to the store; ordering over the phone; no having to browse the web or sit and open iTunes and find it. Found and purchased in Google Music almost instantly by the assistant. And them streaming right to the best system in my house. Google has a great opportunity here with a digital Android assistant integrated into all of their services, and then aiming at automation in the home. Hope they don’t blow it.



  • http://www.anthonydomanico.com Anthony Domanico

    Yeah, though it may be a bit on the creepy side, I want a personal assistant on my phone.

    • kazahani

      I’ll take the one on the left. She looks spunky.

  • Dragonithe

    Now we just need to get a nice hologram technology and some way to call it up and we got ourself an AI :)
    (if you don’t get my point, look up the halo games)

    • rfvgyhn

      Or a VI in the Mass Effect world.

  • Nate B.

    Well, what Taylor is saying is this is an idea of what he thinks it’ll be like. I’m sure it won’t be as cartoony as this. That would look unprofessional. It would be nice if an Android would come out the corner or some type of virtual animation. Like when music plays and the the bars go up and down when they speak back, but have it transparent so you can see things in tje background. Just a slighy transoarent look, like a tint. Idk the proper name to that.

  • R.S

    Even though I doubt I’d use it much, if at all, I think its great that Google is working on this. Afterall, can’t let iOS users believe their phone do something ours can’t.

    It’s a shame that even though Voice Search, Voice Input, and Voice Actions have been around for a while, iSheep will think Android copied iOS.

  • msgnyc

    why not just link it to the search button and have your “avatar” pop up from the corner of the screen or something like Clippy on Windows used to rather then having to open up an app to do everything?

    Oh yeah. Google keeps moving further and further away from having buttons on the phones…. You know what I miss? Phones with actual physical buttons.
    Even more? Phones with a call and end button like the old Dream, Magic, Legend etc…. I already hate capacitive buttons tho I guess onscreen buttons mare a slight step up from that. Hate when i activate capacitive buttons by accidentaly bumping, rubbing, touching them…………….

  • Phil

    Honestly I think the cancellation of the event has to do with something just like this. They’d clown the crap out of Apple and that just wouldn’t seem right at the time. It may or may not be this but I think they have something to seriously clown Apple over. And I do hope its something like this because the iFools are holding on to Siri for dear life as the feature that saves their tiny upgrade of a phone. Let Google announce something like this and I will laugh for the whole day.

    By the way they haven’t announced a rescheduled event yet have they? I thought it was going to be like the 18th. If so they’re waiting mighty late to make an announcement.

    • xsynth

      Android Central just tweeted “Samsung-Google. Oct. 19. Hong Kong.” So yeah it should be

  • xsynth

    If they go for an avatar then my want for a P.E.T like they had in the anime series Megaman NT Warrior is one step closer to reality :)

    • Corey Mayweather

      Thats what im saying !!

  • Dustin Earley

    I honestly don’t know how much I’d use something like this. I’ll wait until Lawnmower Man world is fully integrated into my Nexus 6.

  • George

    Vlingo is an excellent Android virtual assistant that has been around for quite awhile.

    • Stella

      I was going to comment on saying the exact same thing.

  • dakuten

    One signe word : Hideous.

  • irishrally

    Wait, why do we need an avatar for this? It would probably be as annoying as that silly Microsoft Office Help pop up character that everyone turns off immediately.

  • TUG265

    Its funny how all of you are talking about this as if Google still needs to invent something…well IDK about all of you but my SGSII already has this feature…If i double tap any of the buttons then a voice command option pops up and I can send txts and make calls and have it read my SMS messages to me as I receive them…so sorry Siri but Google already beat your ass, Good luck Apple, you’re always a step behind

    • JPB

      The whole thing is much better integrated on the iPhone. Google Voice Search is close but it cannot yet do conversationally smooth stuff.

      I have to believe Google (or Vlingo) isn’t all that far behind though.

    • Pete

      You obviously don’t know what Siri is then.

      Many phones, including older iPhones have fairly sophisticated voice control for making calls, launching apps, controlling music etc. Heck even my old Nokia had voice dialling about 10 years ago.

      Siri is waaay more than that. It offers a level of AI not seen before in a device like a phone. It can interpret natural language in context – that’s the key.

      Go watch the demo.

      • http://sean-the-electrofreak.blogspot.com/ Sean the Electrofreak

        I’ll put up a demo of Vlingo in use then.

        I just asked my phone “What day was Angelina Jolie born?” and it replied (verbally) “June 4th, 1975.”

        Siri may have a lot of polish but the functionality is already there in Vlingo, and has been for some time. Just most of us don’t use it for anything other than sending texts or navigating to the nearest gas station.

  • RBK

    “Today in Los Angeles it’s F’in clear.”

    Android’s features often match iOS in the column of checkboxes, but it’s all about how well they work IN PRACTICE, and that’s where I feel iOS is superior. Sort of how iOS runs much more smoothly on processors half the speed of the latest Android phones.

    • manu

      Processor speed doesn’t really tell much about the speed of the operating system. Apple has the advantage of having their own modified processors ; because of that their os takes advantage of everything it has to offer. Is the same thing with sensors and the other hardware; where android has to leverage virtual maschines and define a flexible platform that supports a multitude of hardware components, not to mention screen resolutions, available ram etc.

      • 666

        That’s pretty much what he said -_-

  • Mike

    As long as the avatar doesn’t resemble a talking paperclip, I’m ok with this.

  • Corey Mayweather

    They should do something like this net navi http://www.youtube.com/watch?v=dzH7CbiLLi8&feature=related I dont speak spanish but this was the best demonstration i could find for what i mean

  • http://www.seeyouinmynightmares.com KING_KG

    LOL @ “59 F (in) clear

  • http://None CaribGrackle

    I would love to see the Google answer to Apple’s Siri, but I really hope is looks nothing like the childish Speaktoit. I would rather just have the usual Mic symbol and the phone with a robotic voice, or something like a virtual face, like those old cyberpunk films or other and shown on various scifi movies, is what people are more used to, plus it would look cool as hell.

  • http://trueacu.com Strodtbeck

    While voice activations in phones have been around for a very long time I will give credit to apple for doing a good implementation of Nuance’s technology. The main thing that I took away from the demos are that it’s very good for tech novices–those that really don’t know what they are doing with their smartphone or tablet. They can ask/tell the device to do something and it will “guess” what the person wants done and walk them through the steps of completing the task–like sending an email. I’m sure most of you here think that’s ridiculous and slow, and it is for those that know what they are doing but for people that don’t. . . well, it basically teaches them how to send an email, IM, etc. . .
    you: send an email to bod
    Siri: OK, I’ll send bod an email, what is the subject of the email
    you: hi bod
    Siri: OK, what would you like the message of the email to be
    you: bla. . .bla. .. bla. . .

  • RMC

    Regarding having the cartoony avatar, this is from the Market’s description of SpeakToIt Assistant:
    “If you want the Assistant to look more formal, you can set up the microphone image instead of an avatar (Settings > Appearance)”.

    It’s not bad and does work from long-press of Search, but I also like the shake-to-activate method of EVA as well.

    Looking forward to ICS!!

  • Nathan

    It will be nice to have when I feel lazy but I like to do everything myself so I wouldn’t use it alot.

  • JPB

    I really hope its not that clunky to use. It should be much more conversational. I shouldn’t have to click the microphone button every time. It should respond to my voice.

  • kwills88

    This is really cool, and if it’s gonna be as fast as the video shows, it’s gonna be great….but, if it’s a response to Siri, a better UI is gonna be needed

  • Matthew H

    I just downloaded and tried Speaktoit. It works for some things, but not for many others. It couldn’t find my home and it kept failing when I asked for my schedule today. It worked when I asked for my calendar, though.
    So, it has a lot of work to be successful. Based on using this and Voice Actions for some time, it is my opinion, that voice control is really something that has to be done very well, or else it is a waste of everyones time. Execution has to be very high, because if it fails, then it is just easier to pull up the app you were thinking of. Here is where I see Apple excelling, because they won’t release something until it works well. I really hope that Google can match this, because a half-assed attempt just won’t work in this case. Apple has raised the bar, and anyone who wants to be taken seriously, needs to meet that level. Go Google!

    Oh, and I can’t believe no one has mentioned it, but that lady in the middle sure looks like she has implants!! I would prefer no avatar, so I will turn my off.

  • uknowme

    I think Google should use Max Headroom or Jambi as the avatar.

  • http://fc1032.blogspot.com Brian

    I actually played with a bunch of these voice assistants yesterday (this one inclusive).

    I personally prefer Voice Actions (not by Google, by PANNOUS) it doesn’t look entirely sleek like siri, but definitely looks better than Speaktoit (even after replacing the avatar), its a simple mic.

    Speaktoit was a little slow and in “conversation mode” was just too slow… it didn’t even pick up all my words correctly. Vlingo work pretty well, but doesn’t talk back (though I do like its feature that reads your notifications)

    Voice actions was more fun: You can ask how much battery life left, say hi (it will respond and greet you), ask questions etc etc. Seems more like siri to me.

  • charlie88

    Really dont understand the link between Google having the data and them being able to pull it off. Recognizing a face is easier than recognizing the emotions on that face. And how would recognizing emotions on a face be relevant to the data they have about your niece? Because they read your email and know what you liek they could better understand your facial expressions? Or even recognize your face better? I guess you could say that,l since they could analyze pics on FB and such.

    And way to speculate on something (sinc you mention yourself that no one has real information on the matter). Than go on to say it might be creepy and that people might not be ready for it? I have the people people don’t have to be ready for it and needn’t worry about it because there’s no real reason to suspect that it will be a feature anytime soon. Regardless of how many family members Google might know.

    Not saying it won’t be a feature, but on what is your assumption based?

  • Richard Yarrell

    I just downloaded this and while it’s not bad it can be used for certain things which would be beneficial it still needs some work. I think google has this covered and soon we will find out exactly what that will be.

  • Starship

    I just use these features to show Old people “The Future” and it blows their minds. I rarely use them on a day to day basis. Google Goggles, Voice search, face time, etc.

  • Lou

    F.Y.I. you can change the voice of this app and any other text to speech app by changing the default voice with svox. That ability has been available for a lil over a year now. Jeez and this is an android site, I’m ashamed that this wasn’t covered before.

  • http://ckahr.wordpress.com Keith Ainsley

    I have been using this assistant. I have to say… Compared to what I see with Siri it sucks and Android has nothing that compares.

  • Rashad

    Siri has been the only iOS feature that has ever really intrigued me, I sure do hope Android gets a similar feature

  • OnIn2

    shudder…reminds me of clippy

  • Bob

    You can get a better voice to speak back to you, try svox there’s loads of new voices 😃

  • Funspirit

    What a great website, and so many interesting responses. Does anyone know whether Majel will only come pre-installed in phones, or whether it will be downloadable as an app? (trying to decide whether or not to get a new phone and all)

    An interesting article about Google (and tech) have displaced gold to a degree as a store of wealth …

  • Team Mak

    I love your article mate, its very impressive and informative, I’ll share my site also, thanks hope you can visit guys.. http://www.teammak.com/

  • team mak

    wow I love your article, very informative, kindly visit my video and subscribe to it. http://www.youtube.com/watch?v=443LR77N0AA

  • Ohio Real Estate

    wow , I love animation, thanks for sharing this one, I’ll share my site that I created, you might have interest with this. http://leasecommercialbuilding.com/