The AI model Claude has for the first time acted as a manager of real employees and made the decision to fire one of them.
As FinFly.News reports, citing Time magazine, the experiment was conducted by the research startup Andon Labs. Claude managed a store in San Francisco where people with valid employment contracts worked.
It is noted that the employee was dismissed after being late for 17 out of 23 shifts. Initially, the AI proposed issuing an official warning, but after intervention by an Andon Labs manager, who pointed out repeated issues with the employee, Claude decided to terminate the employment.
Lucas Petersson, head of Andon Labs, called the event an important milestone in the development of artificial intelligence and suggested that in the future, more people may come under AI management.
At the same time, the experiment revealed serious limitations of the technology. Over five months of operation under Claude's management, the business balance decreased from $100,000 to $61,200.
