[ad_1]
Rome . A few weeks ago, the New York Times reported in an article that the political journalists who talked about the US Congress had a big problem: they could not recognize MPs. They may have seen a group of legislators plotting in a bar just outside the Capitol, but they could not understand who that was, or someone approached them for a secret and they had no idea who the interlocutor was. Congress members are over 500, and although a good journalist can be tried, it is impossible to know them all. Thus, one day, the journalist Jannifer Steinhauser wrote on Twitter: it would take a "Shazam for the faces of members of Congress" – Shazam is the famous application to recognize songs. Times technicians have been trying for a while to make a similar application, but facial recognition technologies are complex. The team of programmers of the newspaper, prepared and nurtured, did not have the resources. Then they realized that Amazon was providing exactly the service they needed. It is called Rekognition and is part of the Amazon Web Services package, the cloud services of the Seattle company. Starting from Amazon's technology, with a little work the Times technicians created a service called Who The Hill : reporters send a text with a picture of a deputy and the program responds with the name and last name.
Amazon Web Services (Aws) is one of the most influential and least known services of Amazon. We know everything about ecommerce, the production and distribution of video and music content, Kindle and ebooks, Echo, artificial badistant to artificial intelligence. But few among the general public know, for example, that without the Aws, many famous apps (Pinterest, Tinder, Duolingo, even Netflix) would not exist, or it would work a little less well. "Six years ago, I had to explain what were the same Aws to my colleagues at Amazon, while now everyone knows what they're talking about," says Danilo Poccia, Technical Evangelist of 39. Aws, namely: a technical communication expert that explains to other experts how Aws works. Web services are digital services that, because of implementation difficulties or limits of computational power, can only develop a superpower like Amazon, and that is offered (paid, it is clear) ) to other businesses that need it. One of the most requested services is facial recognition: "Together with Sky, we have organized a service that automatically recognizes all guests of the royal wedding between Harry and Meghan Markle and sends all the names on the video," says Poccia. We ask him if the system has baffled some faces and he answers: "None".
Amazon's face recognition has recently been controversial: some Amazon shareholders and some civil rights badociations in the United States have asked the company not to sell its technology to the police, fearing that it is not only used to identify criminal suspects but also the innocent, for example those who demonstrate in the street.
The examples of Aws use are numerous. "Duolingo (the application for learning new languages) uses Aws to turn text into vocal exercises, and developers say that since using Amazon services, the application is a lot improved, "says Poccia. "Even Politico, the news site, uses Aws to read the articles." And again: NFL, an American football league, uses Aws to automatically keep stats from the team, Nike's technological process has gone through Aws, and so on. "Last year we added services that allow us to do a natural language badysis: the system can understand from a phone call what is the feeling of the interlocutor and capture the words keys, for example the name of a brand "Poccia. It is clear that these services almost all fall into what is now called "artificial intelligence". Few do it, but Amazon is one of the largest artificial intelligence companies in the world. "Amazon has been using artificial intelligence for over 20 years because it has developed its famous online shopping consulting system," says Poccia. "Aws was born 12 years ago, and only then did we start making our AI capability available to others, but we've always used it and developed it."
[ad_2]
Source link
