Already today, instead of people, many important decisions are made by machines. Moreover, the result of such actions cannot be explained or predicted.
One of the former employees of Google, Apple and Microsoft, Kai-fu Lee, spoke about the main threats that artificial intelligence can pose to humans, reports Yahoo! Finance.
The main threat, according to the expert, can be called “smart” weapons of an autonomous nature, which can be easily taught to kill people on their own.
Technology like this will completely change wars. It will be unrealistic to establish the culprits of the hostilities, because weapons will become available.
It has a high programming speed and a high accuracy rate. Already today, you can spend as little as a thousand dollars and create a small machine that can kill.
Negative consequences arise when an AI begins to perform a specific task and at the same time does not pay any attention to everything else. Artificial intelligence will also collect information about people that must remain confidential.
On the one hand, this will make it possible to offer the Internet user suitable films, music or goods. On the other hand, this is a great opportunity to compromise this or that person.
The machine cannot explain why it made any particular decision in an important situation where human life is at stake. Initially, AI is programmed in such a way as to choose the most profitable solution, which will not always be easy.