"Artificial intelligence" out of human control .. risks and threats

FastCo Design said that Facebook has closed an artificial intelligence program because it has developed a language of communication other than English. The robots "Bob and Alice" began to communicate with each other and reach an agreement between them to accomplish a certain task using a new language. Unknown programmers could not identify.

The robots have been able to develop this new language because of human error from the programmers. Robot engineers are supposed to limit or limit the ability of the robot to learn until it can be controlled. Engineers have failed to program the robot to be the only communication language. , Making the robot develop its own new language.

Which is related to the words of "Elon Musk," founder of the rocket company SpaceX and CEO of Tesla, which leads the development of the automotive industry self-driving, the need to slow in the development of artificial intelligence in order to avoid the risks of increasing reliance on it, Which was rejected by "Mark Zuckerberg," founder and president of Facebook, calling for the slow pace in the development of artificial intelligence "disappointed", for his part, "Mosk" response that "Mark" understanding of the subject is limited.

Mosk is not the only one who is concerned about the growing development of artificial intelligence. Bill Gates, the founder of Microsoft, and Stephen Hawking, the great physicist, have also warned about the ability of artificial intelligence to destroy human life. . Is it exaggerated, or is Mark right?

Intelligent Machinery Engineering:

John McCarthy, nicknamed the father of artificial intelligence, defines the concept as "the science of intelligent machine engineering, especially computer software." It is based on the creation of computer hardware and software capable of thinking in the same way as the human brain, simulating human behavior.

In this sense, artificial intelligence is the process of simulating human intelligence through computer systems. It is an attempt to imitate human behavior, style of thinking, and decision making. Human behavior is studied by experimenting with their actions and putting them in specific situations, monitoring their reactions and patterns of thinking and dealing with these situations, Then attempt to simulate human thinking through complex computer systems.

However, this concept can not be invoked on any electronic piece that operates through a specific algorithm and performs specific tasks. To put this term into an electronic system that must be able to learn, collect and analyze data and make decisions based on analysis, Simulates the way people think, which means that there are three main characteristics:

1. The ability to learn, ie to acquire information and to establish rules for the use of such information.

2 - The possibility of collecting and analyzing these data and information and create relationships between them, and helps in the growing proliferation of data giant (Big Data).

1. Making decisions based on information analysis, not just an algorithm that achieves a particular goal.

Types of artificial intelligence:

Types of artificial intelligence can be divided into four main types, ranging from simple reaction to cognition and self-interaction, as follows:

Narrow AI or Weak AI: The simplest form of artificial intelligence, where artificial intelligence is programmed to perform certain functions within a specific environment.

Its behavior is considered to be a reaction to a given position and can only function in environmental conditions Its own example, the Android Deep Blue, BM. (IBM) and defeated world chess champion Gary Kasparov.

1. Strong or general artificial intelligence (General AI or Strong AI): It has the ability to collect and analyze information, and to benefit from the accumulation of experience, which enables it to make independent and autonomous decisions. Examples include self-driving cars, instant chat robots, Personal self.

2. Super AI: Models that are still experimenting and trying to emulate humans. Here, two basic patterns can be distinguished: first, trying to understand human thoughts, emotions that affect human behavior, and have limited ability to interact socially; A model of the theory of the mind, where these models can express their internal state, predict the feelings and attitudes of others and are able to interact with them, and is expected to be the next generation of machines of the most intelligent.

Smart Machine Threats:

Artificial intelligence is the next locomotive of human development. The advantages it offers to serve people at all levels, personal, medical, industrial and commercial, can not be ignored. In many areas, its development is aimed at protecting people and preserving their lives, such as the use of robots in hard and dangerous work. Military battlefields, and is able to monitor patients' health, provide assistance to persons with disabilities, and monitor homes and institutions from theft, assault and other essential uses.

On the other hand, there are many negative consequences of the increasing reliance on artificial intelligence techniques, and are divided into threats: security, social, economic, humanitarian, and legal, which can be explained as follows:

1 - Economic threats: These technologies will affect the size and quality of jobs and jobs available, as it is expected that the robot will negatively affect jobs in the manufacturing industry, the automotive industry and electrical tools, in addition to customer service.

2 - Security Threats: One of the dangerous consequences posed by artificial intelligence techniques is the threat of these technologies human right to life, as evidenced in the case of independent combat systems (Lethal Autonomous Weapons)