I'm frustrated that there's no extensibility - fine Google, you haven't been able to negotiate with the Neanderthal Australian TV networks for TV guides [1], but don't just tell me that the "What's on TV" feature isn't available, let me plug in an XML feed from somewhere or write an app to do so!
Unfortunately Google probably can't ingest the data without the agreement in place. So unless Google Now actually ran on your phone (it doesn't), you're outta luck.
I don't see this. The partnerships page are just some "badges" a company gets when it starts selling other companies solutions/software/hardware.
In this case, google enterprise has nothing to do with android and the google now. The article is genuine, the author spends time explaining going from Siri to Google Now and his discoveries.
cheers! i am the author of this article and i work for CI&T, a Google partner for Cloud Platform. This wasn't an advert. My first article actually defended Siri (http://www.uxmindset.com/2011/11/between-apple%e2%80%99s-sir...) and as a technologist interested in the advances of machine learning, I decided to compare the two services again, this time pointing out what Google did to transform voice actions into Google Now and thus to outpace Siri.
Feel free to shoot me an email to cyrillo@ciandt.com if you have any questions.
OK, that's a bit creepy, I admit. But for a teenager, could be a great feature. Honestly, I'd also like to know when <insert some old rock band> concerts in my area.
I just tried it, for the bay area it said there were 1887 concerts, then gave me no way to facet the search results by genre, venue etc. I tried the search box, which, confusingly, is right next to the "sign up as an artist". The search results despite the bay area being set as my location showed me concerts in London as the first result.
That's not really the use case. The use case is more along the lines of "I like this band a lot and would love to see them live but I can't be bothered to track this stuff", so you go to songkick and it will send you an email when they announce dates near you. That part is awesome.
Your use case should be accommodated too, granted, but it's not the direction they're currently headed.
If they haven't really considered that use case, and their service doesn't handle it well, why on earth do they provide the means to make their service look bad?
Based on the GP's (and your) report, the search box should be replaced with a freeform text field or somesuch that lets you input artists you're interested in, and a button that subscribes you to notifications.
A whale jumping out of the sea next to a beautiful sunset is awesome.
I don't think a website is. But each to their own!
(I suspect this is the overuse of the word "awesome" that reduces the really "awesome" things to "super extra awesome".)
I was in Bangalore some weeks ago and I tried the query "OK google, directions to vidyarthi bhavan." And it worked perfectly. I was amazed!
The first thing that amazed me was that Google understood what I meant by vidyarthi bhavan! Note vidyarthi bhavan is a generic phrase in many Indian languages which just means 'student building.' In Bangalore, it refers to a specific restaurant in Gandhi Bazar famed for its masala dosas. And I bet there are plenty of people who live in Bangalore right now and don't know about this restaurant. The contextual knowledge here is amazing. And it wasn't just geolocation, I was about 15km away from Gandhi Bazar at that point. And Bangaloreans will know corresponds to an eternity given the traffic situation. And the software did all this by recognizing a phrase consisting of two words, both of which are not in English!
That moment was an epiphany for me, it the precise moment that I realized that personal digital assistants are here, and they actually work.
Sorry, but I think this is just Google Maps. Sitting here in the USA, typing "vidyarthi bhavan" into maps.google.com takes me to the one in Gandhi Bazar, too. Maybe because it has 155 reviews?
The author brings up an interesting example where Google Now provides him a tool to automatically identify and describe a TV show he is watching at night. When he later tries to demo this feature to his friend during the day, it simply does not appear.
The author takes this as an example of Google Now being very smart and learning from his behavior. It knows he watches TV at night so it provides him the identification feature during that time period. It doesn't provide it to him during the day, because it knows he isn't usually watching TV during the day (at least this is the author's assumption).
I take away a different message from this example and it is part of the reason I am very skeptical of software like Siri or Google Now.
Fundamentally I believe the one of the most important features of a software application is that it is predictable. When I do 'X', I expect the program to do 'Y' always. If it can't do 'Y' because of certain situations, I should get an explicit, informative error message.
The author's TV example is an illustration of deep unpredictability. It is quite possible that he is right and there are powerful learning algorithms driving the Google Now's behavior in this case. But all I see is unpredictability.
Yes, I want to customize and shape software to my own identity and patterns. But unless I control that process, I am likely to become very frustrated, very fast as a software application changes under my feet.
This is definitely a big issue with Google Now. It's not that it's unpredictable though, at least in my opinion. It's that it's unpredictable combined with no way to deal with the consequences of that.
It does what it thinks it should, and you have no knobs to turn. If it fails, it just fails. The first year of Google Now, it would often do exactly that. It popped up the card that it claims will appear at the end of the month with information about how far you've walked or biked maybe twice? The rest of the time, it just didn't appear, and there's no way to call it up manually.
I noticed in KitKat that it now offers to let you tell it whether you typically walk, bike, drive, etc. to particular destinations. It seems like in a number of areas, they've chosen to add some manual control to the automated process. I think that's going to be good enough. I'm fine with software that tries to intelligently guess my needs as long as it provides me with a way to override incorrect guesses.
I think the watching TV thing has to do with a TV connected to your WIFI network.
I've experienced what you're talking about though with the flight cards. I'm anxious so I want to know that my flight information is there and ready before I get to the airport, but Google only shows it to me 'when I need it.'
I just recently got a card that entitled 'Watch now' that has a list of shows to watch, and has a 'Listen to TV' button in the top corner. It appears that my phone can listen to the tv, and figure out what I am watching.
I agree! This kind of learning system will never be able to hit 100% accuracy, and even if it hits 95%, sometimes the 5% it misses are unusually important. It has to have a way to deal promptly with unexpected circumstances.
I don't think this requires manual user customization though. I think it's quite reasonable to expect the software itself to probe you for further info on the fly in case of failure.
Removing features unpredictably would be annoying. But merely suggesting a certain feature based on it's predictions would be helpful and useful I think. The feature should still be able to be found in the menus or through a search.
What the author is described is simply Google Now's contextual awareness that a feature could be useful in a given situation. Even if the Listen to TV card doesn't appear in Google Now, you can simply say "Ok Google, listen to my tv," and it will start listening.
You're looking at the glass half empty aren't you? Instead of never getting the right answer, (in this case) you'll now get it at night. How is that not an improvement?
If we suppose that it's not going to get /every/ thing wrong or right, but about an even mix between them, then I'd rather have the distribution of half always working and half always not working as opposed to all of them working half of the time.
Agreed - if I had asked for the weather in Sao Paulo, and it gave me the weather for Sao Paulo on Sunday night, I would be frustrated rather than impressed.
The app is making the implicit assumption that I didn't know what I really wanted, and correcting me. Rather than desirable, I find this creepy, incorrect behaviour and potentially offensive.
While I agree with you on all parts, a cynical side of me tells me that assuming that a majority of your users doesn't really know what they want (or less condescendingly: make assumptions they are not fully aware of) is a viable strategy.
I wish Google would open up an API for Google Now to developers. There are many custom notifications that I would love to add to my Now screen, but right now I am limited to the few that Google provides.
You can have your event, flight, hotel and restaurant reservations, orders and parcel deliveries, and prompt for reviews or custom actions in Google Now by adding meta-data to your notification emails:
From the other direction, you can have Google Now call back into your Android app, triggered by hot words, as demonstrated in the new version of Todoist:
My favorite feature is its ability to detect flight schedules based on e-tickets that have been emailed to me.
It's great when I on the phone and see a reminder about the flight and what time it leaves and how long it takes to go to the airport.
What's especially awesome, is if other people travelling with you forward on their confirmation emails as well, it'll aggregate them and show everyone with tickets + seat numbers.
I was visiting Hawaii and Google Now popped up saying that there has been an earth quake in Vancouver and there is an expected tsunami within 2 hours in Hawaii. I was just about falling sleep at the time. I got up, checked the TV and sure enough there was an expected tsunami. The expected water elevation was 3 feet and we were at 30 feet.
Another time it popped up to tell me that my flight was cancelled the day the Korean 777 crashed at SFO.
The tsunami thing would be less amazing if you lived there rather than being just a visitor. Tsunamis are a big deal in Hawaii. Growing up there, we had monthly tests of the tsunami warning system - sort of like air-raid sirens. They teach tsunami preparedness in the schools and spend a lot of time in the local history class (hawaiiana) talking about the devastation caused by tsunamis in the past. I would not be surprised to find out that the engineer responsible for the tsunami alerts in Google Now was a kama'aina.
Tsunami warnings in Hawaii are broadcast by the Pacific Tsunami Warning Center and through NOAA's weather broadcast. The PTWC is based in Ewa, so technically it's kama'aina. I live on Oahu right now and I've never received a tsunami warning with Google Now, only with the Wireless Emergency Alerts (WEA) that are not specific to any carrier or phone. Flood and severe storm warnings are also broadcast using the WEA.
Apple really dropped the ball with Siri. It feels like the upgrade to iOS 7 made Siri slightly worse, and I was hoping for much, much better.
The key use for me is responding to texts or making phone calls while driving. Being able to say "Call home" is really great, but the really annoying thing is that it I can't appear to force it to use speakerphone. Why is such a small feature unavailable to me? Also, the response time for Siri is extremely slow, and the responses are far too verbose. I wish you could do things like adjust the speaking speed, etc.
The other thing is Apple should have allowed people to integrate their apps with Siri, but it doesn't seem like it's possible. I would love to launch some apps verbally, but Siri isn't useful enough to do that.
Sure, it's clever enough to figure out "I'm locked out of my house" to mean call a locksmith, like the ad shows, but other than what I mentioned above, and setting timers and alarms verbally, it has become almost complete useless to me.
If you had a secretary, they would most likely know even more about your life, and you would have to pay them. I don't find it too scary - if it didn't know anything about me, what use would it be? It could place calls and set alarms for me, like Siri, but there's no way you can truly call that a "digital assistant."
That analogy is flawed. A human assistant probably wouldn't store every information about you forever. It usually is a single person, not a megacorp. You would pay the assistant as you already stated - the motivation of the assistant to provide services are clear and well understood.
I agree with the GP, this scares me. I am interested in this technology, but won't hand over the required information. A 'personal' life tracking/assistant service would be useful..
I want technology like this in my life. I mean what Google Now ultimately could become. A personal assistant powered by near-magical machine learning. But I don't want to sacrifice my privacy to companies that have proven untrustworthy.
More so than that, I don't want to invest a piece of my life into a service, when Google has a track-record for either A) abandoning the product, or B) mucking it up. Let's talk about the future, where Google Now feels like a human secretary in its capacity for intelligence. One day you ask it to schedule an appointment for a Friday lunch. Instead of the usual friendly acknowledgement, (s)he asks you if you'd like to share this appointment on Google Plus. Suddenly you can't make reservations at a restaurant, without having a Google Plus profile. They wanted to abate abuse of the system by linking reservations to real people. Now you have to "fire" your secretary, and spend weeks training up a new one. And (s)he'll never be as good as the old one.
Hopefully you understand the point I'm making. It wouldn't be so bad, if it weren't for the tendency of all these modern services to form their own little monopolies. Take YouTube for example. I want to switch away from YouTube. But the content creators I love don't have their content available elsewhere. We talk about all the freedom our modern internet enabled society brings ... but yet most of what we create these days gets linked inextricably with some walled garden.
I want a technology like Google Now, but I want the data it learns to be stored in an open format which I can port around as I see fit. I want to be married to my data, not to a service provider.
Having said all that, I know my dream won't come true for quite some time. Non-trivial technology almost always breeds walled gardens, due to the expense required to develop and maintain them. We likely won't see open standards here until the technology is old enough to be implemented well in an open source project.
His point isn't just abandonment. His point is that what if tmrw they introduce some new restrictions (like the g+ profile needed), that would mess things up. He doesn't want a situation where he has to comply with google's whims or be forced to leave the tech he loves. Not to mention all the effort put into training that tech. So he wants the data google collected on him to be portable. So that if tmrw he decides to use some other service in place of google now, he can just take out his data, and plug it into the new assistant. Now wouldn't that be cool?
PS: i understand google has absolutely ZERO incentive to do something like this. But hey, just explaining GP's point
Google Now and Siri are two different things. Now is a push system with minimal pull ability. It really sucks at commands, and it doesn't do well with things like "hotel in springfield with a pool".
Siri is a voice interface that doesn't really do anything preemptively.
Neither of them have the ability to "program" via voice the way Plexi does. http://plexivoice.com/
Reading this, I'm happy I never fell for the apple or google phone trends. These devices are intrusive snitches into your life constantly watching you, where you go and what you do and reporting to notoriously dubious giant corporations behind them with privacy and basic human rights respect issues.
How long before those "personal assistant" start influencing and controlling our decisions then our lives ?
The stasi or the eye of moscow would have love theses devices and probably made them mandatory.
What I would love is an app that does this locally, without sending my data to the cloud. I would readily pay a good deal for it. Better than my information being sold to advertising agencies and their cronies.
Here lies the benefits of handing your personal information over to a company like Google.
To me, it's always been a choice, no matter which service I use (paid or free) I am not going to naively believe that for one second my information won't be "abused." When someone says, "XYZ is collecting data on you" the obvious [and always correct] retort is "so is your XYZ provider; or at the very least the NSA is making them." The questions should rather be "who is going to do the least harm with your data?", "who will try to help you the most with the data that they have collected about you?" and "who will make a competent effort to fight for your privacy on your behalf?" From what I have been reading lately, Google gets the best scores - which is why they will continue to trusted with what I choose to share over the internet. Technology should get a job done and then get out of the way as quickly as possible - if "whoring" out my personal information to someone is the path to putting technology back into its rightful place, then so be it.
"Google Now is definitely a great response to Siri."
That's an interesting way to put it. One could also say that Google Now is a great continuation of Google Voice Search. Or a great extension to Google Search. But of course it's also ok to say it's a response to Apples response to these technologies. Probably makes a better headline as well.
Yeh, my HTC Desire had working voice search long before Siri came along.
Just goes to prove the power of Apple's marketing message, most people think Google Now is a copy of Siri rather than Siri being an enhancement of something Google was already doing.
I get annoyed with Siri every time I use it. Either there's a long delay and she doesn't quite understand me, or I get the response I got this morning when trying to look up directions: sorry, I'm not available right now, please try again later. It was amazing using my friend's Nexus phone and watching Google interpret what I said as I said it.
66 comments
[ 4.9 ms ] story [ 140 ms ] thread[1]: http://apcmag.com/Content.aspx?id=3786
Unfortunately Google probably can't ingest the data without the agreement in place. So unless Google Now actually ran on your phone (it doesn't), you're outta luck.
Marcio Cyrillo is the director of marketing at CI&T, who work for Google:
http://www.ciandt.com/us-en/about-cit/partnerships
In this case, google enterprise has nothing to do with android and the google now. The article is genuine, the author spends time explaining going from Siri to Google Now and his discoveries.
Check the link you posted for evidence they work with Google products.
Feel free to shoot me an email to cyrillo@ciandt.com if you have any questions.
I have disabled Google Now and Location Services (the service that tracks your location and also sucks a lot of battery life) on my phone.
TBH, my http://www.last.fm/events is more useful, I find.
I just tried it, for the bay area it said there were 1887 concerts, then gave me no way to facet the search results by genre, venue etc. I tried the search box, which, confusingly, is right next to the "sign up as an artist". The search results despite the bay area being set as my location showed me concerts in London as the first result.
I wouldn't say songkick is anything like awesome.
Your use case should be accommodated too, granted, but it's not the direction they're currently headed.
Based on the GP's (and your) report, the search box should be replaced with a freeform text field or somesuch that lets you input artists you're interested in, and a button that subscribes you to notifications.
I don't think a website is. But each to their own! (I suspect this is the overuse of the word "awesome" that reduces the really "awesome" things to "super extra awesome".)
The first thing that amazed me was that Google understood what I meant by vidyarthi bhavan! Note vidyarthi bhavan is a generic phrase in many Indian languages which just means 'student building.' In Bangalore, it refers to a specific restaurant in Gandhi Bazar famed for its masala dosas. And I bet there are plenty of people who live in Bangalore right now and don't know about this restaurant. The contextual knowledge here is amazing. And it wasn't just geolocation, I was about 15km away from Gandhi Bazar at that point. And Bangaloreans will know corresponds to an eternity given the traffic situation. And the software did all this by recognizing a phrase consisting of two words, both of which are not in English!
That moment was an epiphany for me, it the precise moment that I realized that personal digital assistants are here, and they actually work.
It's actually just plain old web search. Search this on Google (https://www.google.com/search?q=vidyarthi+bhavan) or Bing (http://www.bing.com/search?q=vidyarthi+bhavan) or DuckDuckGo (https://duckduckgo.com/?q=vidyarthi+bhavan) or Blekko (http://blekko.com/#?q=vidyarthi%20bhavan) or Yandex (http://www.yandex.com/yandsearch?text=vidyarthi%20bhavan) or Baidu (http://www.baidu.com/s?wd=vidyarthi+bhavan). Baidu is the only place where I saw it being the second result and not the first.
The author takes this as an example of Google Now being very smart and learning from his behavior. It knows he watches TV at night so it provides him the identification feature during that time period. It doesn't provide it to him during the day, because it knows he isn't usually watching TV during the day (at least this is the author's assumption).
I take away a different message from this example and it is part of the reason I am very skeptical of software like Siri or Google Now.
Fundamentally I believe the one of the most important features of a software application is that it is predictable. When I do 'X', I expect the program to do 'Y' always. If it can't do 'Y' because of certain situations, I should get an explicit, informative error message.
The author's TV example is an illustration of deep unpredictability. It is quite possible that he is right and there are powerful learning algorithms driving the Google Now's behavior in this case. But all I see is unpredictability.
Yes, I want to customize and shape software to my own identity and patterns. But unless I control that process, I am likely to become very frustrated, very fast as a software application changes under my feet.
It does what it thinks it should, and you have no knobs to turn. If it fails, it just fails. The first year of Google Now, it would often do exactly that. It popped up the card that it claims will appear at the end of the month with information about how far you've walked or biked maybe twice? The rest of the time, it just didn't appear, and there's no way to call it up manually.
I noticed in KitKat that it now offers to let you tell it whether you typically walk, bike, drive, etc. to particular destinations. It seems like in a number of areas, they've chosen to add some manual control to the automated process. I think that's going to be good enough. I'm fine with software that tries to intelligently guess my needs as long as it provides me with a way to override incorrect guesses.
I've experienced what you're talking about though with the flight cards. I'm anxious so I want to know that my flight information is there and ready before I get to the airport, but Google only shows it to me 'when I need it.'
http://www.youtube.com/watch?v=04dpHTxqI6U - apparently this has been around for a while.
I don't think this requires manual user customization though. I think it's quite reasonable to expect the software itself to probe you for further info on the fly in case of failure.
The app is making the implicit assumption that I didn't know what I really wanted, and correcting me. Rather than desirable, I find this creepy, incorrect behaviour and potentially offensive.
that would last about 5 minutes before Now turned into another ad platform.
https://developers.google.com/gmail/actions/google-now
From the other direction, you can have Google Now call back into your Android app, triggered by hot words, as demonstrated in the new version of Todoist:
https://todoist.com/blog/2013/11/todoist-for-android-adds-su...
I can't find any published API for the latter.
"OK Google, remove all semblance of my privacy"
> if other people travelling with you [forward] their confirmation emails
part?
I was visiting Hawaii and Google Now popped up saying that there has been an earth quake in Vancouver and there is an expected tsunami within 2 hours in Hawaii. I was just about falling sleep at the time. I got up, checked the TV and sure enough there was an expected tsunami. The expected water elevation was 3 feet and we were at 30 feet.
Another time it popped up to tell me that my flight was cancelled the day the Korean 777 crashed at SFO.
The key use for me is responding to texts or making phone calls while driving. Being able to say "Call home" is really great, but the really annoying thing is that it I can't appear to force it to use speakerphone. Why is such a small feature unavailable to me? Also, the response time for Siri is extremely slow, and the responses are far too verbose. I wish you could do things like adjust the speaking speed, etc.
The other thing is Apple should have allowed people to integrate their apps with Siri, but it doesn't seem like it's possible. I would love to launch some apps verbally, but Siri isn't useful enough to do that.
Sure, it's clever enough to figure out "I'm locked out of my house" to mean call a locksmith, like the ad shows, but other than what I mentioned above, and setting timers and alarms verbally, it has become almost complete useless to me.
I understand that he isn't personally concerned by this, but I feel that a lot of people would be scared by things like this.
I agree with the GP, this scares me. I am interested in this technology, but won't hand over the required information. A 'personal' life tracking/assistant service would be useful..
More so than that, I don't want to invest a piece of my life into a service, when Google has a track-record for either A) abandoning the product, or B) mucking it up. Let's talk about the future, where Google Now feels like a human secretary in its capacity for intelligence. One day you ask it to schedule an appointment for a Friday lunch. Instead of the usual friendly acknowledgement, (s)he asks you if you'd like to share this appointment on Google Plus. Suddenly you can't make reservations at a restaurant, without having a Google Plus profile. They wanted to abate abuse of the system by linking reservations to real people. Now you have to "fire" your secretary, and spend weeks training up a new one. And (s)he'll never be as good as the old one.
Hopefully you understand the point I'm making. It wouldn't be so bad, if it weren't for the tendency of all these modern services to form their own little monopolies. Take YouTube for example. I want to switch away from YouTube. But the content creators I love don't have their content available elsewhere. We talk about all the freedom our modern internet enabled society brings ... but yet most of what we create these days gets linked inextricably with some walled garden.
I want a technology like Google Now, but I want the data it learns to be stored in an open format which I can port around as I see fit. I want to be married to my data, not to a service provider.
Having said all that, I know my dream won't come true for quite some time. Non-trivial technology almost always breeds walled gardens, due to the expense required to develop and maintain them. We likely won't see open standards here until the technology is old enough to be implemented well in an open source project.
PS: i understand google has absolutely ZERO incentive to do something like this. But hey, just explaining GP's point
Siri is a voice interface that doesn't really do anything preemptively.
Neither of them have the ability to "program" via voice the way Plexi does. http://plexivoice.com/
...seriously?
How long before those "personal assistant" start influencing and controlling our decisions then our lives ? The stasi or the eye of moscow would have love theses devices and probably made them mandatory.
http://i.imgur.com/CoKDEVh.png
I would never trade these informations about me for some extra convenience.
To me, it's always been a choice, no matter which service I use (paid or free) I am not going to naively believe that for one second my information won't be "abused." When someone says, "XYZ is collecting data on you" the obvious [and always correct] retort is "so is your XYZ provider; or at the very least the NSA is making them." The questions should rather be "who is going to do the least harm with your data?", "who will try to help you the most with the data that they have collected about you?" and "who will make a competent effort to fight for your privacy on your behalf?" From what I have been reading lately, Google gets the best scores - which is why they will continue to trusted with what I choose to share over the internet. Technology should get a job done and then get out of the way as quickly as possible - if "whoring" out my personal information to someone is the path to putting technology back into its rightful place, then so be it.
That's an interesting way to put it. One could also say that Google Now is a great continuation of Google Voice Search. Or a great extension to Google Search. But of course it's also ok to say it's a response to Apples response to these technologies. Probably makes a better headline as well.
Just goes to prove the power of Apple's marketing message, most people think Google Now is a copy of Siri rather than Siri being an enhancement of something Google was already doing.
Are you sure about that?
Siri company founded 2007, iPhone app demonstrated at D7 conference May 2009.
Siri approached by Verizon to be a default Android app, Autumn 2009.
Siri on iOS first release date: Jan 2010 in the App Store.
HTC Desire first release date: Feb 2010.