AI: risks and control, said the rector

Vadim Bogush, rector of the Belarusian State University of Informatics and Radioelectronics (BSUIR), noted in his speech that technologies should be used for the benefit of the country, but must be viewed through the lens of national security. According to him, it is important to study and promote AI, while simultaneously controlling its development and assessing risks.

Bogush emphasized that the "first" risk is unpredictable results: "there is already ample evidence that artificial intelligence produces results that are not entirely valid, colored by so-called hallucinations, and sometimes even tells users that something does not exist."

The second risk, he noted, relates to evaluating end products created by AI and understanding all components of these systems in order to forecast their performance in a timely manner. The head of the academy called for careful control and a systematic approach to implementing new technologies.

Context

The state's strong interest in AI development in the context of ensuring security and participating in BSUIR's research projects.