الهلوسة المعرفية

Artificial intelligence has become a generative system of knowledge, capable of producing texts, analyzing data, and answering inquiries. However, its tools have generated misleading information, incorrect or nonexistent interpretations, and data that lacks a basis in reality, yet they present it with high confidence as if it were accurate and documented information.

From here, the urgent necessity arises to be aware of the risks and limitations of these tools, especially since technology, no matter how advanced it becomes, remains an assisting tool, while the responsibility for verification, judgment, and decision-making remains an indispensable human responsibility.

These technological disasters have led to what is known as "cognitive hallucination," which has eroded trust in artificial intelligence products, especially those that people rely on heavily, such as religious, legal, and medical consultations.

There are reasons for these cognitive hallucinations, the most prominent of which are:

First: A lack or limitation of the data used in training.

Second: The presence of conflicting or inaccurate information within learning sources.

Third: The attempt to provide an answer even in cases where sufficient information is lacking.

Fourth: Incorrect generalization and inaccurate inference.