Apple announced speech recognition for the next iPhone. Big deal. Android’s had it for more than a year. Apple is just playing “catch-up” and the feature’s not really earth-shattering anyway. Right?
Wrong. Everything in that opening paragraph is wrong, except the sentence that reads “big deal.” Siri is a very big deal, the biggest of deals.
In fact, Siri is the most important thing to happen to mobile in this decade so far.
Siri naysayers fall into two camps: 1) those who say it’s no big deal; and 2) those who say Android has had it since August. Both classes of naysayers are wrong.
Siri is a Very Big Deal
As I detailed in this Cult of Mac post, Siri traces its lineage directly back to the largest artificial intelligence project in history, the Pentagon’s CALO project. CALO stands for “Cognitive Assistant that Learns and Organizes,” and the project involved over 300 of the world’s top researchers in various aspects of A.I.
The entire Pentagon project was headed by Adam Cheyer, who is now director of engineering for Apple’s iPhone group.
Speaking to MIT Technology Review, Cheyer said that CALO sought to integrate “dialog and natural-language understanding, vision, speech, machine learning, planning, reasoning, service delegation and integrate them all into a… human-like assistant that can help you get things done.”
He described the Siri project as seeking to do the same thing in a consumer product. In fact, for the past four years, Cheyer and his team have been focused on optimizing the parts of CALO technology that can execute from a powerful cell phone and be usable by millions of everyday consumers. For the past year and a half, they’ve been working hard to integrate Siri technology into the iPhone OS and application set.
It’s not “voice recognition.” It’s artificial intelligence. And A.I. in your cell phone is a very big deal.
Siri Is Not Like Android Voice Actions
Android Voice Actions is great technology, and is widely used by many Android fans. But it’s not really in the same class as Siri.
Android Voice Actions offers a very solid and capable voice recognition engine that’s on the high-quality end of the spectrum among the wide range of similar products and services that have been around for awhile.
Like all existing voice-command and dictation products, it requires you to say a relatively narrow range of commands or it won’t understand you.
Siri, on the other hand, will be unlike anything the public has used before. You can say things that technically or literally have nothing to do with what you mean, but Siri will in many cases figure out what you mean based on context, history and and artificial intelligence designed to understand regular human speech.
For example, if you want to set an alarm for your nap, just say “wake me up in 20 minutes.” If you want to know what meetings you have scheduled for later, you can say, “how does the rest of my day look?”
These inputs specifically reference neither the application to be used nor the information desired. Yet Siri understands.
As humans, we take the understanding of such comments for granted. But getting machines to understand such tricky phrases is the Holy Grail of artificial intelligence.
Even more human-like is that once you’ve got a conversation started with Siri, it can understand requests that are even more cryptic. For example, you might ask: “Are there any top-rated Italian restaurants within walking distance?” If Siri replies, “no,” you can say, “how about Mexican?” Siri interprets your input in the context of a conversation about top-rated restaurants within walking distance.
Android Voice Actions can’t do anything like this because it’s voice command software, not artificial intelligence.
Siri sometimes gives you web search results, sometimes takes actions for you and sometimes controls the applications on your iPhone.
But Siri also answers questions, thanks to Wolfram-Alpha integration. You can ask random questions like “how many kilometers in 30 miles?,” “What time is it in Paris?,” “how many octaves on a piano,” or “why is the sky blue?” and Siri will just give you the answer. Not a web page. An answer to your question.
What’s the Greatest Great Thing About Siri
But the greatest thing about Siri from a historical and cultural perspective is not that it’s artificial intelligence. It’s that Apple via Siri will make A.I. a mainstream, everyday reality.
The reason is that Apple is baked Siri right into the core experience of using the iPhone. And also Siri is designed for mainstream, everyday use in a way that just about everyone will find compelling.
By mainstreaming, I mean the process of taking something that’s on the fringe of human culture, and making it an everyday part of life for a vast number of people. Right now, Google Voice Action is on the fringe of culture. The average personal on the street never heard of it.
Siri will become mainstream. Just about everyone will become familiar with it, even if they’re not iPhone users.
Edison didn’t invent the lightbulb. He mainstreamed it through product design and marketing.
Ford didn’t invent the automobile. He mainstreamed it through cost reductions and marketing.
We remember the mainstreamers because these are the people and companies that put technologies into every day use for everybody. And we can trace all current lightbulbs and cars back to Edison and Ford.
Google Voice Actions isn’t artificial intelligence. But it is an effective way for users to use voice to do things they would otherwise have to do with typing and touching and navigating through a visual interface.
However, the Android tool isn’t taking voice command mainstream. A lot of power users use it. But your mom will use Siri.
Samsung has also been really active over the course of its last two iterations of Galaxy line flagships, that is, Galaxy S3 and Galaxy S4, both of which have been acclaimed as the most human-friendly smartphones ever. Features like Smart Stay and Smart Dial make you feel that your device actually understands you, and reads your mind, even if the reality is just a matter of some cleverly placed sensors and making use of existing tech.
One of the features debuted in Samsung Galaxy S4 was the Air Call Accept, where you don’t even have to touch your device to answer an incoming call; rather, all you need to do is swipe your hand over the infrared sensor that’s next to the earpiece speaker, and the phone will read your gesture. Well, one of the developers for Android decided that such a feature should not remain restricted to one particular device, and hence, you have Air Call-Accept – a free app available in the Play Store that lets you wave over your phone to answer an incoming call, just like Galaxy S4, on any Android smartphone.
Since most Android phones do not carry that specialized infrared sensor, Air Call-Accept relies on the proximity sensor that has pretty much become the norm for all modern devices. This also means that unless your Android phone sports a proximity sensor, you won’t be able to use this app. Beyond that, there’s hardly anything more to it – all you have to do is install the app from Google Play Store, run it and switch the functionality on – at least, that’s the extent to which you can go in the free version. After that, any call that you receive, just wave your hand over the proximity sensor area and your phone should pick up the gesture.
Air Call-Accept comes in a paid variant as well that additionally lets you reject calls, turn the speaker on when you answer a call or send an automated SMS when you reject a call using the gesture. Should you not want these additional capabilities and feel fine with just the basic S4 feature being mimicked, the free variant would suffice.
0 comments:
Post a Comment