It seems that smartphones' technology will make a giant step forward: Google and Movidius, world leader in the development of mobile vision processor technology, announced that they will work together to create a system which will allow mobile devices to recognize faces and objects. The goal is to integrate the new technology into the hardware rather than through a software. In this way, there will be no need of a specific application neither a network connection. The system has already been tested in Google Photos, which allows you to filter them by using a keyword. However, it requires an Internet connection and it shows only uploaded images. Facebook uses the same type of technology, though its European debut is still in discussion due to privacy issues. Movidius wants to enhance the deep learning and, to achieve his goal it intends to develop an artificial intelligence which will work like the human brain, allowing machines to learn. Devices will get data from their own hardware, and so they won't need a network connection to acquire them from other sources. Google will work on Movidius' latest chip: according to the manufacturer, MA2450 is the most powerful chip of its processor family, and also the only commercial solution able to perform complex neural network computations in an ultra-compact form factor currently available on the market.