Apple explains how machine learning makes better products

Which technology has Apple been promoting the longest and most aggressively? If I were asked this question, I would call augmented reality without hesitation. Despite the fact that the company has not particularly shown itself in this area, the entire top management has been forcing AR for a long time, at each presentation talking about what vast opportunities this technology can offer us. But machine learning in Cupertino for some reason diligently does not mention, although in fact it underlies almost every Apple product, making it better.

Machine learning is a relatively new technology for Apple that it is actively developing now

Apple uses machine learning technology in a large number of its functions and is constantly increasing its presence. John Giannandrea, senior vice president of the company for the development of machine learning and artificial intelligence, told Ars Technica about this. He went to work for Apple from Google and actually brought the practice of using AI to a qualitatively new level. Despite the use of laboratory intelligence and machine learning in Cupertino and before him, it was Giannandrea who was able to standardize their application and direct them in the right direction.

Machine learning on iOS

This is the same John Giannandrea, who moved to Apple from Google and now leads the machine learning direction in Cupertino

When I came to Apple, I was already using an iPad, and I really liked the Pencil. So all the time I would catch someone from the software development team and say, "Okay, where is the development group that is developing machine learning that improves handwriting recognition?" It turned out that it was not I who could not find them, but they simply were not there. It was strange for me to realize that Apple has so many areas where it could apply machine learning, but for some reason it does not. Over the past 3-4 years, the situation has changed dramatically, and I think that in the coming years there will be no fragment in iOS that does not use machine learning, '' said Giannandrea.

Apple today uses machine learning for a wide variety of needs. With its help, language translation is performed, dictation is recognized, sleep quality is monitored and other mechanisms function that record the user's health. However, even more Apple projects that use machine learning affect users only indirectly. This is the formation of maps and panoramas based on lidars and training cars for autonomous driving, which does not require the presence of a live driver.

Why artificial intelligence in smartphones

IPhone 11 camera is deeply integrated with machine learning

Ppl admitted, Apple only actively took up the development of machine learning and artificial intelligence in 2015. Prior to that, the company considered this methodology for improving its systems not sufficiently confidential. The fact is that machine learning is based on the fact that it studies the behavior of many users and, based on the information received, trains itself. However, Giannandrea's experience has made it possible to take full advantage of machine learning without sacrificing client privacy.

Perhaps best of all, the development of machine learning and artificial intelligence is noticeable in the example of the cameras of the iPhone 11 smartphone line. Today, only they offer users the opportunity to create night photos in excellent quality, which special algorithms brighten, giving the impression that the shooting was carried out on professional equipment. Another example is Deep Fusion technology. It processes the frame after creation to improve all its aspects, based on the very technologies of machine learning. Perhaps machine learning is not entirely fair. Still, this technology is aimed at improving everything that it touches. But electricity was once considered a cheat.