Siri was impressive when it first launched on the iPhone 4s, but almost three years later Apple is lagging behind Google and Microsoft when it comes to voice recognition and the “neural networks” those companies use to understand natural human speech. That may change soon, however, according to a report from Wired revealing Apple’s plans to improve Siri’s voice recognition software.
The voice recognition technology currently used by Siri actually comes from outside Apple’s campus. The company opted to license software from Nuance, but it is now building up a team of its own. Wired wasn’t able to get any official confirmation from Apple, but cites the close-knit artificial intelligence community to confirm the iPhone-maker’s recent hiring spree.
One of Apple’s biggest hires appears to be Alex Acero, who spent 20 years at Microsoft working on voice recognition before joining the Siri group. Apple also poached researchers from Nuance, including Gunnar Evermann, who previously managed Siri. According to Abdel-rahman Mohamed, a researcher at the University of Toronto who says he was approached by Apple, the company is hiring researchers, team-leaders and managers. “They’re building a very strong team for speech recognition research,” he says.
It’s unclear whether we’ll see these improvements included in the upcoming iPhone 6. Apple didn’t put much emphasis on Siri during WWDC 2014 last month when iOS 8 was announced, though it did note we’ll soon be able to call up the assistant by saying “Hey Siri.” How she responds may improve drastically in the coming months and years.