|Future of Search? Siri vs. Google|
|Yes, Google, Siri is a serious threat |
...Google saw search as an entry point into artificial intelligence. Apple saw it the other way around -- Siri's AI was its way of helping people navigate the web. The question isn't whether Siri is a search engine that can replace Google's search box. It's a different kind of search -- that is, it's the future of search.
Read the...Full Article [tech.fortune.cnn.com]
I'd say that's a relatively well balanced article. No doubt that Siri already matters and will matter more. She got lots of mentions at Pubcon this week.
Slightly earlier thread: [webmasterworld.com...]
Moreover, refer to this interesting article: Siri Gives Apple a Two-Year Advantage Over Google -
|Google has Voice Actions, a voice search application for Android. So what’s the big difference? It comes down to semantics, Morgenthaler says: “Siri understands what you mean.” She has a far more precise understanding of what you’re saying and the context you’re saying it in, in other words. |
I do not have an iPhone and have not used or even seen Siri, but this quote in the Mashable article got my attention:
|"Google has made a huge contribution to all of our lives … they’ve made search comprehensive and instantaneous … but the whole paradigm is wrong," he says. "People don't want a million blue links, they want one correct answer. All the rest is noise that you'd rather have go away. |
We all should take notice of the word "one". I don't know whether we should take that literally, but if so, then the current Google organics, limited though they may be, looks like a vast smorgasbord. As I understand the word "one", it means only one and none other, which means if you are not THE answer, then you are nowhere.
(And we thought that dropping from #2 to #9 was bad ~ if to get visitor traffic you have to be the one and only "answer", it could be a disaster!)
We shouldn't take it that far.
People need different angels/opinions/ideas/covers for any specific query. Wouldn't they?
A one and only answer regardless of its quality won't work. It's obvious.
On the other hand, :-) I don't trust SE engineers to understand that readers want diversity. Refer to the Panda.
>>People need different angels/opinions/ideas/covers for any specific query. Wouldn't they?
but life has never been about what people need or want...
it is always what big power (currently big business) wants to serve up ... the tv industry is a great parallel, the opportunity that arose from cable was the chance for vast numbers of tv channels - and what happened, they all transmit the same rubbish!
|They all transmit the same rubbish! |
Not the same rubbish, different rubbish, many types of rubbish, but mostly rubbish.
You have to work hard to find what you want.
It is true in TV broadcast as well as in the Internet with search engines.
I know about a new startup here in Israel, that develop a kind of search engine per niche. Instead of using Google search, they research the niche with their experts and give you digested resultS per query - and it works.
|Not the same rubbish, different rubbish, many types of rubbish, but mostly rubbish. |
You have to work hard to find what you want.
It is true in TV broadcast as well as in the Internet with search engines.
The beauty of the six sensed human being is not every one has the same preferences and likes and one answer or one store or one form of entertainment will never suffice.
People aren't just looking for definitions on the web and many times there are multiple definitions for any given word.
People do look for the good, the balanced and the negative opinions, though "she might have a far more precise understanding of what you’re saying and the context you’re saying it".
|People need different angels/opinions/ideas/covers for any specific query. |
When I was telling my wife last night about the Siri article and specifically the line about "one correct answer", her immediate response was "that will never work because people want choices".
As I said I don't have an iPhone so I'm hoping that someone here who has used Siri can give us some firsthand feedback. It is hard for me to believe that any service that gives ONE answer can possibly be the so-called "future of search"!
We are on the same page.
I don't know Siri but have been working on software speech recognition 12 years ago, conducted many tests and went to dozens of conferences/seminars and never thought (still don't think) that this technology could replace human.
To my best knowledge, there are too many aspects in human voice/speech that no machine is capable to recognize and understand.
So far I haven't done anything but local type Siri searches, but *in that space* I do get choices from it.
If I ask it to find me a nearby Chinese restaurant, it will answer back "I've found 15 Chinese restaurants that look like they're nearby" or whatever. Or it will tell me it's found seven possibilities for whatever else I was looking for, and then present them to me.
But I haven't used it for any straight up non-localization related search queries yet. Maybe I'll try that.
|But I haven't used it for any straight up non-localization related search queries yet. |
I'm wondering what will happen if you ask for an online product along the lines of "small blue round widget", as many of us here are selling via the web rather than local.
With Google's Eric Schmidt and various tech writers seeing Siri as a very big deal, the way it delivers its results may have a huge impact on our incomes, if in fact it becomes the "future of search".
|To my best knowledge, there are too many aspects in human voice/speech that no machine is capable to recognize and understand. |
Doing voice to text translation in a hand-held device was impossible 12 years ago yet it's available in most smart phones today. We're almost to the point of having a universal translator as seen on Star Trek when I was a kid, getting real close. Obviously it's not perfect, yet, may never be, but the approach Google took with sampling voice data for years on GOOG411 will probably be the same approach needed to finish the project - lots of samples.
All the technology is quickly coming together way beyond search, it'll be completely disruptive to what we know today. Combining all the technologies for search, speech recognition, visual recognition, etc. and then putting some basic intelligence like Siri behind it to make it all work autonomously to pull those technologies together to serve your needs quicker and faster, it's going to rock our world.
Back to Siri vs Google, I think the changes seen in what we consider 'search' and Google's role in mobile search is going to be much bigger than most expect. Look at mobile search, Google Maps and Navigation is used far more often as the dominant interface to Google Search as one would rightly expect it to be.
Using just what we have today, imagine just combining Google Search and Navigation, Google Goggles, Word Lens, Shazam, Yelp, Open Table, and a few other cool apps all under on personal assistant like Siri and it could be a real game changer but not a Google killer whatsoever, just a Google integrator. Google has wisely positioned itself as the dominant player with Google Maps/Navigation/Places so everything Siri needs to know Google already has and replicating that data is a next to nearly impossible task, nothing Apple is ever going to do with a few thousand servers and a copy of Nutch. The only real challenger to Google's position here is probably Bing at this point, not Siri.
Another game changer Apple probably hasn't anticipated will be that the ad revenue from visitors for the individual services Siri integrates will disappear as those services become transparent. This will either cause those services (Google, Bing, etc.) to charge a service like Siri to access their data or they'll simply block access by Siri and make the personal assistant essentially brain dead. If Apple were to implement it's own search back-end, it would be to protect Siri from this eventuality of being held hostage for excessive fees or being hobbled IMO.
A little forward thinking and rambling...
Another potential model will be that the end user will pick the 3rd party services Siri uses and pay a monthly fee, such as new TVs offering direct access to Hulu, Netflix, etc. via the internet on top of the existing cable company VOD offerings.
The mashups will easily go beyond basic search, thinking Google is the only type of search is a bit limiting in what it means for mobile users.
Having something like Siri make a reservation at the nearest highly rated steak house with any open seating from 8-9pm and getting driving directions using a mashup of GPS, Google search, yelp, open table and navigation should be a snap, doing things like that in seconds that it might take you 10-30 minutes to manually do today. Now imagine the reservation mashup ties into movie times and Siri picks matching movie times and matches it with available Open Table dinner reservations and gives you a list of possibilities, and so on and so forth.
In the future, your phone may be doing search all day long even when you don't ask it.
Big Brother won't be watching you but your HTC EVO 4G just might! Existing bleeding edge technology will allow phones to charge from ambient light and when power isn't a real issue, visual and audio sampling will probably be continuous and your phone might just start searching for things before you ask. Example, your phone might chime in "Do you like that new INXS song playing in the background? Would you like me to add it to your playlist?" or while driving "We're within 2 blocks of a Target, didn't you say you needed some socks and a new pair of pants the other day?".
Anything is possible as the hardware and mashups rapidly evolve and I think the smart money is that someone thinking along these lines will go into acquisition mode and snap up the best of breed and start working on the next gen apps ASAP and lock out the competition from using those services.
It's possible today actually ...
In the end Google, and Bing, etc., will just be one plug-in component of that mashup future unless they own it.
Oh wait, can you say Android?
There are some Siri-like apps for Android already.
Maybe Google already is working on one itself? :)
You have had some great posts but I believe your vision is as clear as the all seeing eye on the one dollar bill today. Damn bro, great insight.
Siri technology is about 10-20 yrs away from being major, by that stage search/google would have changed so much then it makes this whole conversation moot.
|Siri technology is about 10-20 yrs away from being major |
Siri technology is as close as how much money you're willing to throw at integration today.
It's really a simple matter of building APIs into all the services Siri needs, which could be accomplished in months, not years, and then expose those APIs to the internal Siri engine.
It's more a matter of time and resources than anything else, if the backend integration picks up steam, Siri could be a SIRIous force in no time at all.
The fact that Apple acquired them could mean either massive resources are now being thrown on the project to make it happen as quickly as possible... OR ... it's about to be muddled down in so much corporate quagmire that it will suffer a major falter and allow competitors to quickly catch up and bypass it's current status.
However, it won't take 10-20 years to make it major, if it ever becomes major, it'll happen much sooner than that, the technology is all here if there's enough money to make it happen and Apple has boatloads of cash so if anyone can do it, they can.
Here's a good article telling why speech recognition, no matter how good it might become, will never work for all aspects of human-computer-interaction.
I ran some plain old searches from products, services and information on Siri last night, and pretty much all I got were the same Google results that I'd have gotten had I typed them in.
I expect that will change eventually.