|
Hi, I'm a first-year Computer Science PhD student at Columbia University, focusing on human-computer interaction (HCI). I'm fortunate to be advised by Dr. Lydia Chilton and part of the Computational Design Lab. My current work focuses on designing, building, and evaluating AI-powered productivity and collaboration supports. Specifically, I am researching the design of human-AI interactions with large language models (LLMs) and with voice user interfaces (VUIs). Before Columbia, I earned my two Bachelor's with Summa Cum Laude from Cornell University, where I had the privilege of conducting research with Dr. Andrea Stevenson Won, Dr. Natalie Bazarova, and Dr. Susan Fussell. I also worked closely in teaching with Dr. Gilly Leshed and Dr. Kyle Harms. Email / GitHub / LinkedIn / Twitter / Google Scholar |
📨 long [at] cs.columbia.edu |
See More
|
|
My Info/Sci and Communication background made me understand that technology has both communicative capacity and responsibility. Therefore, as designers and developers, we should speak, design, and build for those who can't. As a PhD newbie, I am also more than eager to learn and grow by joining exciting HCI/AI projects. Don't hesitate to reach out for collaboration ;) |
|
Katy Ilonka Gero, Tao Long, and Lydia B. Chilton. ACM Conference on Human Factors in Computing Systems. CHI 2023 (To Appear) Large language models (LLMs) have proven able to follow complicated instructions and generate coherent text, while less work considers the social-technical gap that this technology poses. We discover three elements that govern a writer’s interaction with support actors: 1) what writers desire help with, 2) how writers perceive potential support actors, and 3) the values writers hold. We align our results with existing frameworks of writing cognition and creativity support, uncovering the social dynamics which modulate user responses to generative AI. |
|
Tao Long and Lydia B. Chilton. Under Review Voice User Interfaces (VUIs) have opened up many new types of interaction, while they are still underused. We synthesize existing VUI design guidelines and distill four design challenges
that contribute the most to users' non-use and technology abandonment.
Then, we identify opportunity spaces for
designers and developers to explore such as focusing
on 1) information support while multitasking (cooking, driving,
childcare, etc), 2) incorporating users’ mental models,
and 3) integrating calm design principles.
|
|
Tao Long, Dorothy Zhang, Grace Li, Batool Taraif, Samia Menon, Kynnedy Simone Smith, Sitong Wang, Katy Ilonka Gero, and Lydia B. Chilton. Under Review Effective communication of STEM is crucial in a rapidly changing world. Tweetorials are popular, but experts struggle to make complex topics engaging. We propose using large language models (LLMs) to help writers create relatable hooks and spark curiosity. Our evaluation shows this reduces cognitive load and motivates readers. LLMs can support contextualizing technical information for public engagement. |
|
Yingjie Wang, Sadia Sharif, Vivian Liu, Sitong Wang, Tao Long, and Lydia B. Chilton. Under Review Describing text-to-image generations is challenging due to the difficulty of summarizing an image's aesthetics. Adding composition keywords can help by providing specificity on color, texture, angle, and lighting. In an annotation study of 9216 images, we found composition keywords helped, but efficacy varied by category. We discuss how composition keywords can be used and integrated with existing practices. |
|
Tao Long, Swati Pandita, and Andrea Stevenson Won. Under Review There is a lack of research on the application of user-centered design methodologies in non-HCI fields, particularly for designing social science paradigms. Thus, we identify shared considerations as well as disagreements between the naive participants and experienced researchers for future iterations of our prototype and for designing virtual paradigms for social science research in general. Themes include: 1) Design for intuitive use, 2) Design for inclusivity, and 3) Realism vs minimalism in design. [award 1] [award 2] |
|
Aparajita Bhandari, Diana Freed, Tara Pilato, Faten Taki, Gunisha Kaur, Stephen Yale-Loehr, Jane Powers, Tao Long, and Natalie Bazarova. Proceedings of ACM Human-Computer Interaction. CSCW 2022 A lack of clear public benefits information for asylum applicants negatively impacts their healthcare engagement, but digital tools offer a promising solution. We identify four challenges for immigrants using current digital tools: lack of information, inaccessibility, emotional barriers, and lack of cultural sensitivity. The study recommends considering multiple perspectives when designing tools for immigration information, and suggests specific design recommendations to improve digital tools for asylum seekers and those who support them. [pdf] [video ] [press 1] [press 2] [press 3] [project site] |
|
My passion for teaching and socially impactful research significantly contributes to my decision to pursue a PhD and (possibly) an academic career in Computer Science. For the sake of intersectionality, I always remind myself that I am indeed a privileged person along my journey. It helps me rethink myself, my role, and my responsibilities in better helping others with the resources I have and embracing intersectionality. Always happy to contribute to new initiatives ;)
|
Last Updated: Feb 28, 2023. Template.