Society. Spaces. Screens
Abstract
In today’s rapidly evolving society, the linguistic divide between generations complicates family communication, making it challenging to achieve effective and intimate interactions. This paper introduces “GenSync,” a translation tool specifically designed to bridge these generational language gaps using OpenAI’s GPT-4.0 models. GenSync focuses on transforming idiomatic expressions, slang, and colloquialisms to facilitate better understanding among family members of different ages. Our experimental study, involving 32 participants, revealed significant improvements in the quality of conversations for groups utilizing GenSync, especially those using the transparent version that provides both translated and original dialogues. These findings demonstrate the tool’s ability to enhance communication quality, intimate connections, and overall family dynamics. However, the blackbox version, which only offers translations without showing the original dialogue, scored lowest in terms of conversational quality and usability. Participants from post-experiment interviews expressed that transparent GenSync brought about newfound familial closeness and understanding. The study highlights not only the opportunities but also the challenges of implementing such technology in family settings. Drawing from these insights, we propose design implications for future intergenerational translation technologies: (1) enhancing transparency, (2) improving context sensitivity, (3) increasing adaptability, and (4) focusing on user-centric design. Ultimately, GenSync exemplifies how technology can be utilized to mend generational divides, fostering more meaningful and empathetic family connections.
Sora Kang is a Ph.D. candidate in the Human-Computer Interaction (HCI) at the Human-Computer Interaction+Design Lab at Seoul National University. She holds a Master of Design(MDes) in HCI from the University of California, Berkeley. Her research spans human-computer interaction, large language models, and social computing, with a particular focus on AI-driven applications in new media and interactive design.
Youjin Hwang is an HCI researcher who graduated with a Ph.D. from Seoul National University and is currently working at Samsung Electronics. Her research interests focus on human-AI interaction and conversational AI.
Jongwon Lee is an AI researcher at Samsung Electronics. He holds a Master’s degree in Computer Engineering (2014) from Columbia University. His research includes Machine Learning, Natural Language Processing, Large Language Models, and LLM-based agents, including Planning and Reasoning.
Rae Noh is an AI engineer at Samsung. She started her career as a mobile software developer at Motorola. She participated in the development of payment systems and back-end systems of app marketplaces at Samsung Electronics since 2012. She was also responsible for the CTO in the Samsung In-house Startup Project (C-lab) in 2020. Currently, she is involved in the development of Samsung’s AI platform.
Joonhwan Lee is a Professor in the Department of Communication at Seoul National University. He holds a Ph.D. in Human–Computer Interaction (2008) from Carnegie Mellon University. His research includes Human-Computer Interaction, human-ai interaction, ai journalism, social computing, situationally appropriate user interaction, and information visualization. He directs the Human-Computer Interaction+Design Lab (hcid.snu.ac.kr).