Samantha Reig, an assistant professor in UMass Lowell's Miner School of Computer and Information Sciences, is investigating how robots and AI can be designed for the best possible human interactions. Her work spans a $20 million National Science Foundation initiative for elder care and a new project examining the social impact of AI writing tools on students and workers. The research offers concrete insights into designing AI that supports daily life without overstepping.
Her work is part of a growing focus on AI for Science & Research, where understanding human behavior is as critical as the technology itself. Reig's projects probe how people actually use language with robots and how they perceive peers who rely on AI assistance.
Researching language and robots for elder care
Reig is a member of the AI-CARING institute, which brings together researchers from seven universities with industry support from Google and Amazon. The NSF-funded program's mission is to create AI technologies that support older adults, particularly those experiencing cognitive changes. In her role, Reig focuses on the nuances of how older adults use natural language to ask robots to complete tasks in their living spaces.
"Do they say, 'Bring me a cup of coffee with two milks and one sugar?' Or do they say, 'It's cold and I'm thirsty,' and expect the robot to infer that, because it's 9 a.m., it should bring them a coffee?" she said. Early findings suggest predictable individual and generational differences in how much detail people provide in their requests.
A grant to probe AI writing tools' social impact
In June, Reig received a $15,000 UMass Lowell Internal Seed Grant to kick-start research into the social impacts of AI writing tools. She is collaborating with Associate Professor Jose-Mauricio Galli Geleilate of the Manning School of Business. The project emerged from her own experience as an instructor watching the AI boom reshape education in real time.
"I had a lot of questions, just like everybody else, about how tools like ChatGPT and Claude can support learning by giving a little help without doing the work," she said. The team is exploring whether adding constraints-such as asking AI to act like a mentor-can preserve students' narrative voices while preventing the tool from doing the thinking. They are also examining how people perceive each other's use of AI when working together, and whether such constraints can make peers appear more competent and trustworthy.
Common threads across different domains
Though elder care and student writing seem distant, Reig sees a shared vision. "Across these projects, I'm trying to figure out what makes emerging technologies understandable and useful for everyday users, and what kinds of AI guardrails will help us live and work better rather than making things worse," she said. Both lines of work seek to define the boundaries where AI fills a gap without overstepping.
A path from psychology to human-computer interaction
Reig's route to computer science was unconventional. She started in psychology at Cornell University, then added a major in information science after taking an introductory computing course. A class project on team interactions with telepresence robots introduced her to the field of human-computer interaction.
A personal experience later shaped her focus. Her grandmother, who had lifelong cognitive disabilities and never used a computer or smartphone, was given a small robot dog at her senior center. "She loved the dog so much they let her take it home. Watching her interact with it and tell stories about it to her friends back at the center made me passionate about understanding the diverse ways that different people relate to robots," Reig said.
Next steps: AI as a mediator in tech support
Looking ahead, Reig and her student Haritha Malladi are examining interactions between older adults and younger relatives around technology troubleshooting. Early interviews suggest that AI agents could help translate jargon or mediate when emotions run high, without reducing user agency or intruding where they aren't wanted.
Why this matters for science and research professionals
Reig's work models a research approach that pairs large-scale, multi-institutional collaboration with targeted, user-centered studies. For those developing AI systems, her findings highlight the importance of capturing the real language and social norms people bring to interactions-before scaling solutions. The seed grant project underscores that even widely deployed tools like AI writing assistants still lack clear norms, and that studying those norms can directly shape design guardrails that improve team dynamics and trust.
Your membership also unlocks:
