As a result, the company spent a lot of time training new employees who were hired to replace those who had quit. Many of the skills needed were what the researchers called “tacit knowledge,” experiential knowledge that cannot be easily codified but that large language models can pick up from chat logs and then mimic. The company’s bot assisted with both technical and soft skills, directing agents to relevant technical documents and suggesting witty phrases to calm angry customers, such as:
After the bot started helping, the number of issues the team solved per hour increased by 14 percent. Additionally, the likelihood that an employee would quit in any given month decreased by 9 percent, and customer attitudes towards employees also improved. The company also saw a 25 percent drop in customers asking to speak to a manager.
But when the researchers broke down the results by skill level, they found that most of the chatbot’s benefits accrued to the least skilled workers, who saw a 35 percent productivity boost. The top-skilled reps saw no gain and even saw a slight drop in their customer satisfaction scores, suggesting the bot might have been a distraction.
Meanwhile, the value of this high-skilled work multiplied as the AI assistant trained less-skilled workers to use the same techniques.
It is doubtful whether employers honor this value of their own accord. Aaron Benanav, historian at Syracuse University and author of the book Automation and the future of worksees a historical parallel in Taylorism, a productivity system developed by a mechanical engineer named Frederick Taylor in the late 19th century and later adopted in Henry Ford’s auto factories.
Using a stopwatch, Taylor broke down physical processes into their component parts to determine the most efficient way to complete them. He paid particular attention to the most skilled workers in a trade, Benanav says, “to get less-skilled workers to work the same way.” Instead of a sophisticated engineer with a stopwatch, machine learning tools can now identify industry best practices Collect and disseminate workers.
That didn’t run so hot for some employees in Taylor’s time. His methods have been linked to falling wages for higher-skilled workers because companies could pay less-skilled workers to do the same work, Benanav says. While some high performers remained necessary, companies needed fewer of them and competition between them increased.
“By some accounts, that played a pretty big part in organizing all these lower- and middle-skilled workers in the 1930s,” says Benanav. However, some less punitive systems emerged. One of Taylor’s followers, mechanical engineer Henry Gantt – yes, that one chart type– created a system that paid all workers a minimum wage but offered bonuses to those who also achieved additional targets.
Even when employers see an incentive to pay top performers a premium for teaching AI systems, or employees win them over, it can be difficult to share the spoils fairly. First, data from multiple workstations could be pooled together and sent to an AI company that builds a model and sells it back to individual firms.