Inside OS X 10.8 Mountain Lion GM: Dictation & speech - Page 2

Rather than being in the dark, perhaps Google is breaking with tradition and releasing a product when it was pretty well ready rather than "rushing it to market".

 

It's not completely unreasonable to think that Apple was well aware of the significant Google Voice Search improvements and the new Google Now feature, realized that if they didn't release Siri last cycle that Google would already have theirs out before the next iOS update, and released Siri a bit early themselves to avoid looking reactionary.

 

Hard to say for certain yet, but early indications are that Google's voice features are more polished and complete than those offered by Apple's Siri (which you have to admit would be unusual if Google was the one who rushed it), which might be evidence that you have things backwards.

Edited by Gatorguy - 7/14/12 at 1:44pm

 

Yeah i agree, another reason to why apple might have release siri while still in beta, just because the iPhone 4S didn't have a major selling point over the 4, dual core & faster GPU & better camera wasn't enough i guess.
siri was the buzz on the iPhone 4S release, even my mum knows what siri is.

It was good enough for every other iPhone launch. And better, even, than some.

You're taking that view only because Google Now came after Siri. There is absolutely no evidence to support that and plenty to support that Google Now is taking a significant number of cues from Siri's UI design, just as Samsung is with S Voice.

The difference is that Google had this IP already. Most of is the data sourcing and cloud service infrastructure that is still new to Apple.

You can't say Apple rushed Siri because it's Beta because it would have always been Beta the way Google spent years with their voice service grabbing requests before they canceled that project. If Apple wanted to release a non-Beta Siri from the get go they only have to options: create a different service for years to grab and process voice request before then releasing virtually the same service under a new name, Siri, or they could license the data from another company, such as Google, assuming that was even possible.

Bottom line: Google played catch up to Siri's interface but Apple is playing catchup to Google's backend. For a change Apple is holding the short stick in that this is Google's wheelhouse. Apple can't just look at the results on an Android-based phone and create a Siri backend for it but Google can look at an iPhone and easily copy the UI.

Soli, I'm sure even you would admit that Siri is rougher than a typical Apple Beta and what you would expect from an Apple product. Does it feel rushed or unpolished to you? Apparently not even adequate server support was in place, another indication of a possible rush to release.

 

In any case what about Google Now copies Apple's Siri UI? Apple doesn't even offer results based on anticipated search terms (no need for user input), much less the swipeable "cards" displaying the found information. 

Edited by Gatorguy - 7/15/12 at 12:55pm

Not in the least! I've stated from the start that any such service that requires extensive data sourcing and processing of millions of users over a long period of time, especially one based on their language, culture, speech patterns and other demographics can not be released to the world as finished product. It's simply not possible.

It's just insane or consciously biased to think that because it's Beta it must have rushed. And the argument that their data centers can't handle the load of a new service is just as bad as saying Apple didn't plan properly when they have 4 million units sold in a weekend and run out of supply. What makes this worse is that I'm sure you said there is no technical reason that Siri can't be on all other iDevices when it launched on the iPhone 4S.

Nearly everything about Google Now is a rip from Siri. Google had all these elements sitting around for years, some of the shelved and forgotten, but less than a year after Apple unveils Siri they have a system in place that mirrors what it does in nearly the same way. The only key difference, and a nod to Google, is their showing of web results in the Google Now UI instead of requiring a launch in the browser app itself. But that's expected since that is Google's home turf.

What specific UI features are copied then if nearly everything is a "rip from Siri"?

 

Google Now goes much further than Siri, anticipating search requests and displaying them without user input. It didn't take a year to do so either with at last one beta-tester commenting on the anticipated search and cards display on his tablet  back in December of last year as well as it's inclusion in the next Android OS update, well before public knowledge of the feature. That doesn''t sound like Google seeing Siri two months prior and then wondering how to answer it. Even Google's engineers aren't that good.

 

I still think it's just as possible that Apple's Siri was the one pushed out sooner than they may have preferred, knowing that Google's much improved voice search and Google Now would be introduced before Apple got another OS version out.

Edited by Gatorguy - 7/15/12 at 1:40pm

The entire focus of the layout and fundamental way you access the service didn't exist before Siri. I don't that is something Apple can protect or ave rights to but that doesn't it wasn't copied.

Google has some additional features but the one you mention is the one I have no seen in action. Just a propaganda video from Google, much like their Google Glasses. I have seen a demo from some users that are impressive but we also need to consider the number of devices that can access the service.

How does it anticipate requests? Just like Apple's Passport that uses location services to know where you are? I know it's suppose to learn patterns in your behavior, like Nest does, but you they are trying to achieve does not seem like something that can easily work. Passport is very focused to certain kind of task and Nest is very focused to a specific location and particular chore, and yet they aren't perfect.

So how does Google have this all seeing all knowing service that works with absolutely no user sourced desirability for a given location and/or direction? I guess they can use your location when you made a Google search for something; butt do they really record the exact location of Android devices every time you use Google? I guess they can use their Maps service to get an idea of things around you or in the direction you're going, but unless you live on a one road town how does it know where I'm going if there are a dozen places I frequent in the next block? Is it going to serve up a message every time it thinks it knows what I want? Is it going to constantly send my location and time/date stamp back to Google so they can build a reasonable database? If they already have on, who in their right mind would authorize that?

According to what I see stated there is nothing it uses that would have that level of "second sight."

• http://en.wikipedia.org/wiki/Knowledge_Graph

I don't know what you're trying to say there.

Siri was well polished. They bought the original Siri from the devs well before that then integrated it into their OS with plenty of time to spare. They also worked to get the data center going a long time before that. In fact, we kept waiting and waiting for that data center to go live long after we got work that it was completed.

Again, the data you think is so easily determined by a developers sitting in a room typing in responses to questions they conceive isn't realistic. There is no service that can understand the context of human speech without an incredible amount of data sourcing. Google did this years ago and then shelved it. Good on them, but Siri is phenomenal for coming out at complete as is.

As an aside, to show you I'm not being subjective I do think Apple's Maps for iOS 6is rushed. It's not nearly as complete as it should be to compete with what Google has spent many years perfecting. If it doesn't get better by many factors in many areas by the time iOS 6 goes live then I say they should have just let it sit until iOS 7 or until it's actually a good replacement.

Uh, this is a rumor site.

Is there a possibility to get medical dictionary to Dictation in OS X?

Does it learn as it does the same mistake over and over again in my dictation?

It uses the same back end as Siri for speech-to-text. It'll learn but I doubt it'll learn the way you want it to if you are using it for medical transcription. I don't know if there is any good programs for this as it seems like a tough pill to swallow for developers do to all the terminology but if there is it will be something you would install on your PC or Mac and run locally, not a generic speech-to-text service that ships with an OS.

Would this help?

https://drchrono.com/ipad_ehr/

 

and demoed here:

http://www.youtube.com/watch?feature=player_embedded&v=jwSeWqMsP8M

 

EDIT: I found where Nuance announced an SDK for medical transcription apps two years ago. I would think someone would have made an app that at the minimum allows you to make a one-time manual correction and have it stick. Perhaps not tho.

http://www.macworld.com/article/1146772/nuance_medical.html

Edited by Gatorguy - 7/29/12 at 3:18pm

Hi i got a Macbook Pro late 2011 Core i7 model i just updated it to OSX Mountain Lion but when i go to system settings i dont have dictation & speech does it only work on the new Retina Pro Models ???

It's a feature with every Mac running ML.

Dude i dont have that in my menu ?? i have attached screen shots of my Mac plz help ????

 

Any one got any tips ??? i dont have dictation & speech  after i upgraded to Mountain Lion OSX ????

 

Check System Report under USB. Not sure what you should be looking for, as I'm on my Mac Pro, which only reports the Cinema Display, iSight (probably that), and Display Audio as separate entities.

USB? im running a Thunderbolt Cinema Display via my Macbook Pro ill let you know what i can dig up... Thanks  

 

Oh! That could be it, then. Go to System Preferences/Sound/Input and make sure you've selected at least one of the appropriate options there. 

 

It might be confusing your microphones.

No idea what the issue could be but this site isn't the best for getting help. I'd go to MacOSXHints.com forums. There they solve problems, here we just talk about them. :D

Because the voice recognition software IS in Apple servers, not in iOS devices. iOS only task is to capture your voice, digitalize it to a given format and send it to Apple servers where voice recognition occurs with huge language databases, dialects, words, idioms and a huge computational power that permits the system to LEARN your particular mood and makes the whole process appear as if it were magic.