Artificial intelligence has taken a first step into personnel management: Claude, an AI model appointed as manager of a retail store in San Francisco, made the decision to fire a human employee. This case could mark a turning point in AI's impact on the economy, reports infohub.kz.
According to Time magazine, citing research firm Andon Labs, the system called Aversion of Claude, which was tasked with managing a store including a team of real employees, dismissed its first employee last month. Researchers called the move a watershed moment in AI's influence on the economy.
In March, Andon Labs launched an experiment to determine whether an AI agent could independently run a business and what impact such systems might have on the economy. To that end, Claude, an AI model by Anthropic, was given the role of store manager at Andon Market in San Francisco.
According to Andon Labs data cited by Time, the fired employee was late for 17 out of 23 shifts. However, the AI did not notice the pattern for a long time: the employee handbook it had created "disappeared" from the model's limited working memory.
The decision to fire was not fully autonomous, admitted Andon Labs CEO Lucas Petersson. An Andon Labs employee first asked Claude to "find" the forgotten handbook and analyze the situation. Initially, the model only suggested issuing a formal warning, and only after a prompting question from an Andon representative about whether the employee was truly suitable for the job did it decide to terminate them.
It is noted that overall, Claude proved to be a rather lenient manager and had previously told employees not to worry about being late.
So far, the experiment shows that AI is not more effective than humans in business management, the publication notes. In March, the store received a starting balance of $100,000, but over five months it shrank to approximately $61,000. Researchers cite Claude's overly lenient management and questionable commercial decisions as possible reasons, the portal writes.