Last month, the world witnessed its first-ever instance of an AI-powered store manager terminating a human employee at Andon Market, an experimental retail outlet in San Francisco. The AI store manager, which operates on Anthropic's Claude model deployed by Andon Labs, dismissed an employee who had been late for 17 out of 23 shifts. Initially, the AI manager failed to recognize the issue and overlooked the employee handbook's guidelines due to its constrained working memory capacity. It was only after being prompted by human staff that the AI initially recommended a written warning and eventually decided to terminate the employee. This experiment has underscored that AI still falls short of human capabilities in terms of commercial judgment and personnel management, and the project is still in progress.