Understanding Why Siri Understands You
V Introduction
* Siri is artificial intelligence (AI), not simply voice recognition
* There is a huge difference
* Interprets what is said, uses history to learn from it, uses context to derive meaning
* I also suspect that Apple is using the aggregate knowledge from Siri input to their servers to help make Siri smarter. She can learn.
V Siri has powerful ancestry
V SRI International, founded as Stanford Research Institute
* One of the world's largest contract research institutes.
* Established by the trustees of Stanford University in 1946 and made private in the '70s
V Siri has its roots in the Cognitive Assistant that Learns and Organizes (CALO) project which was funded out of DARPA (Defense Advanced Research Projects Agency )
* CALO was inspired by Radar O'Reilly in MASH. An assistant who always knew what the captain wanted sometimes before he did.
* Project involved over 300 of the top minds in fields relating to A.I.
* Adam Cheyer was head of the project and is now director of engineering for Apple’s iPhone group.
* CALO was supposed to allow a machine to have 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
* The Siri project was designed to take the AI technology from CALO and turn it into a consumer product.
* For the speech part Apple also licenses Nuance which is also based on technology spun out from SRI
V SRI began development of Siri in 2007 and spun out the project in 2008.
* They released an App for iOS
* Apple acquired Siri in 2010
V Why Siri is different
V It has voice recognition and more
* There are four elements to a machine that can function like a person
* 1) speech
* 2) Decision making algorithms
* 3) Data - Siri gets this from the integrated apps and the web (Wikipedia, wolfram alpha, Yelp, etc)
* 4) Agency, the ability to execute tasks on your behalf.
* Most existing voice systems rely on specific commands. Key word triggers that need to know what you want to do.
* Siri is designed to not require those (although in current form is does to a certain degree)
V It applies "reasoning" to try and figure out what you mean.
* That reasoning is the application of those 4 elements of machine AI
* Also can rely on past responses and context to derive the meaning of your requests.
V Why is Siri such a big deal
* It's the first wide spread consumer application of this type of technology
* Millions of people now have Siri. It's giving the first AI technology to a mainstream audience, asking them to use it, and they are.
* It's arguably comparable to what Apple did in the 80s by taking computer technology and making it mainstream.
V It represents a big shift it he ways we can interact with computers
* Like the move from text commands to the mouse and GUI
* Now we can move from mouse to natural speech
* Remember this is still BETA, but it's pretty incredible beta.
V Apple now owns this technology and the years of research that went into it
* The Siri team is reportedly one of the largest teams at Apple
* They also have the talent and have hired on more in the past few years
V No doubt Google and Microsoft will try and catch up
* Google's Android VP Andy Rubin said at the All Things Digital conference, "I don't believe that your phone should be an assistant. Your phone is a tool for communicating. You shouldn’t be communicating with the phone; you should be communicating with somebody on the other side of the phone."
* Google is concerned enough about Siri's potential that is has shifted a key speech recognition engineer, Dave Burke, from the U.K. to join the Android team at Google's Mountain View. Calif. headquarters,
* Apple has a significant lead
V Use response and adoption may go slower than Apple wants
* Speaking to a device in public can be a very public and often socially awkward way to interact.
V iPhone 4S does have Bluetooth 4.0 which allows low power short range operation.
* Talks of things like watches that could act as mic for Siri input
V It's only the beginning
* As TidBits points out, disruptive innovation doesn't happen overnight, but often seems that way in retrospect
* Siri is the seed of disruptive change.
* A paradigm shift to a Star Trek future where voice and speech may be the primary interface to data
* It's not going to happen overnight. Just like the mouse wasn't understood by many when it was first introduced.