Several months ago, during an executive course at a Western university, I experienced a simulated board meeting discussing a set of files with an AI agent present as a member. The atmosphere was electric and strange at first, but over time, members began to get used to the AI agent and took the matter seriously. The key observations I and the other participants took away from this experiment were as follows:

The AI agent was very confident in its words and information, even when it was wrong.

There were hallucinations and fabrication of non-existent data and information that required verification and cross-checking.

Occasional lack of alignment with context and deviation from the discussion framework.

Tendency to agree and not object to opinions raised by other members. We decided to give direct instructions to the AI agent to be clear and frank and express its opinion without flattery. We also asked it to speed up its speech and shorten its remarks so that it wouldn't repeat every time: "As a member of this board with you..." Honestly, we ended up asking it to summarize the key points, compare opinions, make a final decision between two views where perspectives were divided, and share the meeting minutes via email. In other words, its role was sometimes closer to a coordinator and secretary of the board than a board member. Nevertheless, it performed well in presenting and organizing opinions neutrally and systematically.

This experience may lead to a profound question: Will the day come when AI systems manage companies and organizations and start issuing instructions to humans? In today's world, algorithms and their recommendations account for 80% of views on Netflix, and algorithms control the coordination and communication between customers, stores, and delivery drivers.

In fact, many participants in that executive course from various organizations and countries complained about strange reports they received from senior management containing facts based on hallucinations of AI systems prepared by consultants for managers who neither read, scrutinize, nor verify, and simply keep their teams busy responding to AI hallucinations.

AI systems may not have reached the stage of taking leadership yet, but it is clear that their indirect influence has already reached many organizations at the senior management level.

In short, the question is not: Will an AI agent manage you or not? Rather, the question is: How do you prepare for a world in which AI systems will interfere in many critical administrative decisions, directly or indirectly? Change is coming powerfully and quickly. So, will you be a victim of this change, or will you be in the vanguard that takes the initiative, innovates, leads change, and shapes the future with your own hands and will?