Study prompts new theory of human-machine communication

Hed: Study prompts new theory of human-machine communication

LAWRENCE – In a new paper, two University of Kansas scholars propose a novel theory of communication analysis that takes into better account how people interact with ubiquitous technology in the 21^st-century workplace.

In “Socio-Technical Exchange with Machines: Worker Experiences with Complex Work Technologies,” ----- link to https://doi.org/10.30658/hmc.10.3 ----------- in the Human-Machine Communications Journal, Cameron Piercy, associate professor in the Department of Communication Studies, and Reaia Turner-Leatherman, a former McNair Scholar at KU currently pursuing a graduate degree in couple and family therapy at Kansas State University, polled nearly two dozen people from all walks of life about how they relate to the “complex technology” they use to do their work.

Based on the responses they got, the scholars propose “Socio-Technical Exchange” as an improvement upon the well-worn Social Exchange Theory (SET) when it comes to contemporary workplace communication.

“Social exchange theory was formed in the 1960s,” Piercy said. “Probably half a million papers cite some form of it ... because it's a logical argument about how people engage with others: They engage with people who are rewarding, and they avoid people who are costly. Your co-worker might know a lot about a certain subject, but they talk your ear off. And with technologies, it's the same.”

The authors say that people tend to form “machine heuristics,” which are basically beliefs about the relative worth of various forms of tech.

“People like machines because they're objective and consistent,” Piercy said, “but they also see machines as ineffective at subjective judgment. So we don't want a machine hiring us.

“But this paper says it's not just that straightforward. It's not always that we see machines as objective, but not good at subjective judgment. It's also when we use a machine over and over, then we start to form a specific belief about it -- that it's good at certain things and bad at others; that it's suitable for certain tasks and not for others.”

Although there still appears to be a role for what the authors call “human uniqueness,” machines are preferred for certain tasks, they found, writing:

“When they felt expertise was important, people often preferred a human coworker, finding coworkers more efficient and knowledgeable. However, with simple or embarrassing questions, a machine was deemed a superior collaborator.”

For now, anyway, machines don’t make judgments about their human users, which means SET doesn’t truly apply to such relationships. Thus, the proposition of Socio-Technical Exchange, or STE.

Bear in mind, Piercy said, that the study was done in 2022, before the widespread adoption of artificial intelligence tools such as the Large Language Model ChatGPT. Even then, the authors wrote, “our findings suggest that humans still feel interdependence with machines and that they develop stable cognitive representations for both human and machine work partners.”

Piercy has formed and directs KU’s Human-Machine Communication Lab, ----- link to: https://hmc.ku.edu/ ----------- where he is currently using grant funds to expand upon this new theory with a quantitative study.