Faculty Research Spotlight | Apostolos Filippas, Ph.D., Assistant Professor, Information, Technology, and Operations; Christopher Blake Distinguished Research Scholar in Business
Faculty | Dec 22, 2025 | Gabelli School of Business
A Recently Published Study on Social Media Engagement, & a New Study on AI as a Computational Model for Humans
User engagement is a cornerstone of social media. But what truly drives likes and shares? According to Apostolos Filippas, Ph.D., an assistant professor at Fordham University’s Gabelli School of Business, it’s often a case of attention bartering.
Filippas co-authored an article published in Management Science titled, “The Production and Consumption of Social Media,” which offers a novel model of attention bartering to explain how social media users exchange engagement. Drawing upon empirical evidence from #EconTwitter, the article explores how social incentives shape content production, consumption, and network formation online, providing key insights into platform design and user behavior.
“On social media, you might get into a weird little relationship that we call attention bartering, where you give somebody attention in order to get some attention back,” Filippas explained. “You follow them, or you’re clicking ‘like’ on their tweets, not because you actually like them, but because you want to attract a follower yourself. So, you’re consuming their content, just so they consume your content back.”
The study’s findings, Filippas notes, can help social media platforms and advertisers to increase profits. “Everything is designed toward getting users to spend more time on social media and see more advertisements,” he shared. “Some social media platforms have made the conscious decision to encourage attention bartering because, as the research indicates, the more attention you receive, the more active you are on social media.”
This line of inquiry was a natural fit for Filippas, who is an economist with interests in market design and the economics of technology, primarily in the context of online platforms.
His current research is focused in the area of AI. He and his co-authors posit that newly developed large language models—because of how they are trained and designed—are implicit computational models of humans, making it possible to utilize AI simulations to apply simple models of micro-decision-making to larger phenomena. “Large language models have allowed us to take an agent-based models approach and put it on steroids,” he said.
Filippas believes that these kinds of AI-powered simulations have the potential to transform the way researchers approach their work. “In real life, we just see the road taken. We just see what happened. But in a simulated world, we can change things and run the same thing at any speed we want, wind it back, study it a million times, and see a million different roads, and make better decisions,” he asserted. “I think this has tremendous promise for decision-making, and it will actually be transformative for how we conduct research and make decisions in the next decade.”
Written by: Kimberly Volpe-Casalino
