Major changes are planned for Siri with the help of neural networks

With the imminent arrival of the final version of iOS 10, probably in September, the voice assistant, «Crab"It should sound a bit more natural thanks to the machine learning technology that Apple is implementing."

The company is developing some kind of licensed technology for a deep neural network (DNN), as hinted at in a conference today attended by several top Apple executives . Tom Gruber, head of advanced development for Siri, indicated that while Siri's responses are still based on a central database of recordings, the voice assistant's machine learning will "smooth" and make Siri's responses sound more "human."

The assistant's robotic voice has often been a point of discussion and criticism. Most recently, singer Barbra Streisand called Apple CEO Tim Cook to complain about how Siri pronounces her last name, and Cook responded by promising to fix the problem in an upcoming update. This seems more like a marketing tactic to disguise the announcement of a future update to the voice assistant.

In an interview featuring high-level executives such as Eddy Cue and Craig Federighi, it was revealed that Apple actually migrated Siri's voice recognition to a neural network system in July 2014 , but was unaware of this until today. The implementation of this new technology is said to have dramatically improved Siri's ability to understand commands.

On the other hand, it's known that Apple has a large team working on machine learning technologies, which includes not only Siri but also other company products. Even so, we'll have to wait a while to see the effects of this technology more directly—perhaps until September?