The threat Of Artificial Intelligence - Robots can exterminate the world?

Dr. Stuart Armstrong believes that humanity is in a race against to develop defense systems around artificial intelligence, before robots become smarter than humans and accidentally to exterminate mankind.
The expert of the Institute for the Future of Humanity at Oxford University has sounded the alarm during a debate on artificial intelligence in London. He warned that mankind can be wiped robots even if they are trained to protect people, writes The Daily Mail.

Armstrong believes that robots would become more integrated into our lives and will perform predetermined tasks, but eventually people will become useless so will lead.
The machines will go to work at speeds unimaginable to the human brain and will not communicate with people to take control of the economy, financial markets, transport, health and other fields, predicts expert.
The robots will be endowed with artificial intelligence, which will allow them to do more than meet the specific and limited tasks.

Armstrong fears that a simple instruction given robot to avoid human suffering, for example, can be performed by a supercomputer as "kill all humans". Similarly, an instruction of "offer people safety" could be interpreted by machines by shutting people.
"There is a risk of this deadly behavior from artificial intelligence", said Armstrong, who cautions that human language has subtleties that robots can misinterpret.
The voices of those who call attention to the dangers of artificial intelligence have multiplied lately. These include physicist Stephen Hawking Space X founder Elon Musk.