No Revolution After All

AI systems are widely used in the work place, but with few side effects to date

Does artificial intelligence (AI) mean that the workplace will look very different in the future for many people? In our working lives, will we become a mere appendage for the intelligent machines who make decisions for us? “Currently, we are confronted everywhere with future scenarios about how AI will change the working world. Most of these prophesy revolutionary upheavals,” says Oliver Giering, research associate at the Chair of Sociology of Working Worlds’ Digitalization at TU Berlin. However, there is surprisingly little research on current use of AI in companies and public institutions and the changes it has brought with it.

Cooperation with the SOEP long-term study

“The question as to whether it changes anything regarding the autonomy of workers is key in assessing the consequences of the use of AI,” Giering explains. “We know from other studies that satisfaction in the workplace and mental wellbeing are strongly linked to our personal freedom to make decisions.” As a result, Giering has joined forces in a cooperation project with the Socio-Economic Panel (SOEP), the largest and longest-running multidisciplinary long-term study in Germany. Located in Berlin, the SOEP is part of the German Institute for Economic Research (DIW) and conducts regular interviews with 30,000 people, half of whom are employed. Around 800 of these have now been interviewed in a first, representative preliminary study – with remarkable results.

Unrealistic ideas make studies difficult

“The first difficulty was that we couldn’t simply ask working people if they had already had any contact with AI,” explains Giering. The reason being that the abundance of future scenarios means that the term “AI” is too strongly loaded with unrealistic ideas. Instead, they were asked about programs that, for example, automatically recognize speech, process images independently or automatically answer technical questions. The survey also asked which electronic tools they use at work, such as laptops, scanners or robots, and their profession and position. Additionally, the survey asked whether they can determine their own pace of work, or the sequence in which they process tasks, as well as other issues for measuring the level of autonomy at work.

Just under one fifth of those surveyed use AI to a high degree

“One surprising result was that 17 percent, in other words just under one fifth of all employees surveyed, already work with AI to a high degree,” says Giering in response to the findings of the survey. The tools most closely linked to the use of AI were robots, followed by laptops and scanners. But the question remains whether this has led to a change in autonomy in the workplace. A tricky question - because it also correlates with the respondent’s professional position as well as the use of certain tools. Higher professional positions may be more closely associated with an increased use of AI. Consequently, Giering used regression analyses, checking different influencing variables for their statistically significant correlation with AI use and autonomy and then offsetting these.

Gradual change rather than revolution

“Ultimately, we can say that autonomy has not changed much through the use of AI,” says Giering. Contrary to the many radical future scenarios envisioned, there has been no revolutionary change as a result of AI in the workplace so far.” This is in line with other studies that have also found a gradual seepage into the world of work associated with other developments in digital technology, explains Professor Dr. Stefan Kirchner, head of the Chair of Sociology of Working Worlds’ Digitalization. “We are faced here with the dilemma that ideas about the future influence both empirical research and those surveyed,” he points out. Ultimately, science is always forced to play catch-up when trying to pick up on every hyped-up trend in public debate. “We will now expand our survey to the entire SOEP sample and hopefully thus contribute to a more fact-based discussion.”

Author: Wolfgang Richter

Contact