Do not make the mistake to think that Google is the only company on the internet who gathers information concerned with artificial intelligence. As reported a week ago by MIT Technology Report, Facebook is working on an initiative to deep learning, which means that it is very concerned about Artificial Intelligence (AI).
Company formed a small research group, called internal AI team. Details of their work are still secret, but involves deep learning experiments with technology that is practically taught machine on steroids that can make computers think, to understand, to learn by themselves beyond what they programmed people to do .
Learning deep layers of neurons in the neocortex copies – the rational part of the human brain. It is the concept on which the robot Siri, semantic search, IBM Watson, new medical breakthroughs, natural language recognition and, in the future, machines and robots autopilot field work. At the moment, it's the best chance to make the dream of artificial intelligence science fiction into reality.
In the past two years, Google has dominated the deep learning – for example, by employing the most known and appreciated minds in the field, Geoffrey Hinton, the first to use this technology to recognize images. But nothing that has to compete with Facebook's rival in this game. The social network has a very precious resource: the data they gather from a billion users (of which 300 million are active) hyper-personal sharing information about their lives on the platform every day.
Large amounts of information and learning go hand-in-hand with cars, because the neural network can learn only by making connections between information they access. Can you remember when Google has revealed that his brain could identify virtual cats videos on YouTube, though no one had told him somethin about that cat.
What's even more interesting about this discovery is that the computer could realize more obscure things – for example, to distinguish the differences between several varieties of the same species of fish, what a man can do. In other words, maybe our brains are too complex to be imitated by a machine, but databases Google and Facebook collect amounts of information that a human being could ever retain.
So thought Russian capital firm when he invested hundreds of millions of dollars in Facebook two years ago because he was sure that the social network giant was the future of artificial intelligence. Just a thought and Facebook when it invested in the project's co-founder Vicarious Dustin Moskovitz, which aims to mimic the human brain to create smart machines, with rational spirit of observation.
What is disturbing in this? Facebook's virtual brain not only is built with users' personal information, but it could help the company to know more about you and your life. Technology Review reported that learning technology can detect deep emotions statuses or events, even if they are not explicit photos that could recognize objects and could predict how people will behave in the future. The implications of this level of knowledge are many and frightening.
The phenomenon is already happening to some extent, yet with traditional learning machines that Facebook uses to improve the user experience and develop new technologies. It was recently reported that Facebook knows the details and features about yourself that you did not tell anyone – for example, the fact that you are gay or democrat. This revelation has led to silly games like this, the site that allows access to all your information on Facebook, and I say that's your personality. (They fit me pretty good results, except for the shyness).
The question is: do people care that their personal data are used to build an intelligent artificial brain that can learn even more things about them? Do these powers will be used for good – to improve Facebook's news feed and develop new useful features of the network – or for evil ?
Srinivas Narayanan, a Facebook engineer in charge of the new investigation team, told Technology Review that in addition to new technology company, researching technology for deep learning "a subject of general interest will be published." Ultimate goal the Facebook site is still unknown, but users, be careful: the information you give them everyday social giant would be able to get out of control much louder than you think.
5 reasons why you can become an Add-Energy Renewable Romania PARTNER
To register your Company as ADD ENERGY PARTNER , please access 
For other informations click to 