Artificial intelligence cannot be held responsible for its actions because a personality is not defined by legal systems. In other words, artificial intelligence is not responsible for its actions because it is not the subject of contemporary legal systems. In order to eliminate this deficiency in legal systems, a personality definition for artificial intelligence has been made in this study. The concept of electronic natural person has been created and detailed.
entry
In today's world where technological developments continue rapidly, legal systems cannot keep up with this speed. Our legal systems have to be able to create new institutions in line with the needs of the age. Otherwise, it is possible to encounter acts that are not provided for in the law and cannot be sanctioned. The legal institution needed in the age of technology is the determination of the legal status of artificial intelligence. Many artificial intelligences have carried out actions that will constitute the elements of the crime. However, artificial intelligence has not yet been sanctioned because it is not the subject of the legal system. So much so that, as a result of these actions, there are court decisions that the responsibility usually belongs to the user or the manufacturer. Therefore, it is necessary to define a personality specific to artificial intelligence, taking into account the level of development that artificial intelligence can achieve in the future. After this definition, the responsibility of artificial intelligence will become editable. However, before a personality definition for artificial intelligence can be made, we need to understand the technology of artificial intelligence.
The concept of artificial intelligence emerges as a sub-branch of computer science. This concept was first mentioned by American scientist John McCrathy in 1956 (1). Artificial intelligence is generally used for technological products that can learn and process information. The field of use of artificial intelligence is quite wide. Technological products using artificial intelligence can be in a rocket launched into space, as in an iron in your home. The person who made the first research on artificial intelligence is Alan Turing(2). Turing developed a test that he named himself and tried to prove that machines can learn information like humans. However, due to the conditions of the period and insufficient technology, the test could not be successful (3). The general definition of artificial intelligence covers many different types of artificial intelligence. However, according to the distinction made by artificial intelligence scientists, it can be examined under four different sub-titles.
The Reactive AI type specializes in only one area. In other words, this type of artificial intelligence lacks the ability to produce new memories. In addition, this type of artificial intelligence does not have memories, it can only make some evaluations about the future. Artificial intelligences called AlphaGo developed by Google and Deep Blue developed by IBM can be given as examples of this type.
In the type of Artificial Intelligence with Limited Memory Capacity, there is the ability to know and remember the past. This type of artificial intelligence has a memory created with software made by the manufacturer. Examples of this type are seen in autonomous driving assistance used in everyday life. Self-driving vehicles make short-term history records. For example, while changing lanes, the vehicle moves in a way that does not cause an accident according to the previously detected objects in the environment. However, these past records are short-lived.