When Conor Sprouls, the representative of customer service department of MetLife insurance call center (Call Center), he always had to keep an eye in the lower right corner of the screen. There, in a small blue box, the AI is giving instructions and assessing how he works.
Speak too fast? The program will flash a speedometer icon, indicating that he should reduce the speed of speaking.
Sounding sleepy sound? The software displays a hint about the energy of employees, pictures of a cup of coffee.
Not enough sympathy? A heart icon pops up.
Conor Sprouls works with AI software at MetLife's Call Center.
For decades, people have been terrifyingly imagined about super-efficient robot armies that will invade offices and factories, competing for every human work. But of all the worries about the potential of artificial intelligence that can replace employees, it seems that people have ignored the possibility that they will also replace them to become bosses.
Sprouls and other call center staff at his office in Warwick, Rhode Island, still have many supervisors who are human. But software called Cogito on their computer screens – produced by Cogito, a Boston-based AI company – has become an auxiliary manager, keeping track of them at work. . At the end of each call to the customer, the software will make comments about the Sprouls session to send to the supervisor. If Sprouls tries to hide the software display window by minimizing it, the program will notify his supervisor.
Cogito is one of several AI programs used in call centers and many other workplaces. Its goal, according to Cogito's CEO, Joshua Feast, is to make employees more productive by providing them with real-time feedback.
"There is a change in human activity," Mr. Feast said. "We can deduce from the way people talk to each other whether things go smoothly or not."
Communication display board of Cogito software.
The goal of automation is always effective, but in this new type of work, AI is watching people themselves as things that need to be optimized. Amazon also uses complex algorithms to track workers' productivity in their logistics centers, thereby automatically making judgments that dismiss those who do not meet their goals, according to The Verge. recently. IBM has also used Watson, an AI platform in the staff recruitment assessment process to predict future performance and claims it has a 96% accuracy rate.
And yet, Cogito has partnered with major insurance companies like MetLife and Humana as well as financial and retail companies. Accordingly, this AI has more than 20,000 users. Percolata, a Silicon Valley company, has also persuaded Uniqlo and 7-Eleven to become its customers. The company uses sensors in the store to calculate actual productivity scores for each employee and rank them from low to high.
Algorithm management is not a new concept. At the beginning of the 20th century, Frederick Winslow Taylor revolutionized production worldwide with his scientific management theory. He tried to eliminate factories' inefficiencies by defining time and measuring each aspect of the job. Recently, Uber and Lyft and other platforms have saved billions of dollars by outsourcing computer companies to handle common tasks in human resource management such as planning, salary calculation. , performance evaluation…
But the use of AI to manage workers in common jobs has caused much controversy. Critics accused companies of using algorithms for management tasks, saying that automated systems could work inhumanely and offer unfair punishments. Because AIs can track everything workers do, it won't be clear why they do so. While a manager can see and analyze the underlying cause clearly.
"Surprisingly, any company can lay off their workers without human involvement," said Marc Perrone, chairman of the International Federation of Food and Trade Industries. in a statement about Amazon in April.
In a free economy, algorithmic management also creates tension between workers and the platform that connects them with customers. Recently, drivers for delivery companies like Postmate and DoorDash objected to an algorithmic payroll method because they thought they were unfair.
Aaron Osei, another employee in the center.
There are no demonstrations at MetLife's call center. Instead, the staff seemed to see Cogito as a "minor nuisance". Some people say they like to receive notifications that pop up to remind Cogito. However, there are people who try to find a way to fight it out from their computer. MetLife now uses software with 1,500 employees, saying that since applying the technology, customer satisfaction has increased to 13%.
"This really changes the behavior of people they don't know," said Christopher Smith, head of MetLife's global operations.
In fact, many companies have used psychological tricks borrowed from video games, such as using badges or rankings to try to push workers to work better. However, not everyone feels comfortable being monitored by the machines.
Phil Libin, executive director of All Turtles, an AI startup company in San Francisco, was horrified after visiting MetLife's call center.
"It's a dark hell," he said. "Why would anyone want a world building like this, where you can be judged by a computer?"
However, guardians who use AI are told that the systems are not "overbearing". Instead, it makes employees better by reminding them to thank their customers, improve their fickle voices to avoid depressed moods.
Office work at MetLife
Pymetrics, a New York startup, has applied AI to the process of screening and recruiting personnel in some companies through games. The algorithms are then analyzed to ensure that they do not produce biased recruitment results, or favor one group over another.
"We can adjust data and algorithms until we can eliminate bias," said Frida Polli, executive director of Pymetrics. "We can't do it with a human being."
It can be said that the use of AI to correct human shortcomings is a good thing. But as more and more AI enter the office, executives will have to resist the temptation to use it to tighten grip on workers, subjecting them to constant monitoring and analysis. . If that happens, a rebellion against machines is what seems to be a reality.