Sorry!! The article you are trying to read is not available now.
Thank you very much;
you're only a step away from
downloading your reports.

Apple Has Big Plans for Siri After Falling Behind Google, Microsoft


Although Google, IBM, and Microsoft have invested in neural networks for speech recognition, Apple is a notable holdout. But that will soon change.

In the job market for virtual assistants, it's time for Apple's (NASDAQ:AAPL) Siri to pad her résumé.
Introduced in the iPhone 4S in 2011, Siri promised a bold new manner in which to interact with our mobile devices. If the promos were to be believed, Siri knew all, spoke all, did all. However, the end result left much to be desired as many vocal queries were misunderstood and were often relegated to a string of links from an old-fashioned Web search.
But Siri's shortcomings weren't most apparent until Google (NASDAQ:GOOG) updated its voice actions for Android into the powerful Google Search functionality which works in conjunction with Google Now. As a virtual assistant, Google Now seemingly knows what the user needs before he does, and verbal searches like "How old is William Shatner?" promptly retrieve a dedicated info card and a spoken answer.
And as Google Now and Google Search became more intelligent -- and even Microsoft (NASDAQ:MSFT) debuted its own capable assistant Cortana -- Siri remains the equivalent of Murphy Brown's string of bumbling secretaries (to use a dated reference). But at long last, it's been reported that Apple has big plans to finally send Siri into a training program to improve her prowess and clout amongst her virtual assistant brethren.
According to Wired's Robert McMillan, Apple is assembling a group of software engineers and researchers culled from Nuance -- the outfit in charge of Siri's voice-recognition -- to usher the service into a newer, faster, and far more accurate system.
Key to this update are computational models called "neural networks" that mimic the chemistry of the human brain. Put simply, the neural network algorithms learn with experience and begin to understand more spoken words the more they are used. Google, Microsoft, and IBM (NYSE:IBM) have put this form of artificial intelligence in place for each of their voice-recognition systems, with very impressive results.
And now it's time for Apple to enter the neural network fold. 
Along with engineers poached from Nuance, Apple nabbed Alex Acero from Microsoft's speech technology unit to work as the senior director of Siri's group, as well as Arnab Ghoshal, a researcher from the University of Edinburgh.
Also courted by Apple was Abdel-rahman Mohamed, a postdoctoral researcher at the University of Toronto. He told Wired, "Apple is not hiring only in the managerial level, but hiring also people on the team-leading level and the researcher level." Adding, "They're building a very strong team for speech recognition research."
Unfortunately, Siri's changeover to neural networking won't happen any time soon and most likely won't debut alongside iOS 8 later this year. Peter Lee, the head of Microsoft's research arm, has years of experience with the technology and believes Apple needs roughly six months to develop a system that can catch up with Google and Microsoft's. However, Lee asserts the results will drastically improve Siri from gimmick to all-star.
With the explosion of wearables and home automation under way, voice recognition will become more integral in our day-to-day usage with mobile devices. It's comforting to see Apple stepping up to the plate and ensuring Siri is up to the challenge.
< Previous
  • 1
Next >
No positions in stocks mentioned.
Featured Videos