Human agency in AI-mediated creation and interaction.

I design generative AI systems that strengthen creative, representational, and relational agency—helping people create, express their experiences, and shape meaningful interactions.

Keywords: Human-Computer Interaction, Human-AI Interaction, Human Agency, Generative AI

Profile photo placeholder
Dr. Shuo Niu
Computer Science
Clark University
Office: CMACD 323

Introduction

I’m an associate professor in the Computer Science Department at Clark University.

Generative AI can expand people’s capacity to create, interpret experiences, and sustain interactions. Yet these capabilities can also displace human reasoning, misrepresent personal meaning, and weaken interpersonal boundaries. My research examines how people retain meaningful authority to direct, evaluate, revise, and reject AI contributions through three interconnected research pillars:

  • Creative Agency — Assistance ↔ Authorship: How can AI support creation while preserving human reasoning, judgment, and ownership? I study how learners, teachers, and social media creators make consequential decisions as they develop artifacts, instructional designs, and content.
  • Representational Agency — Interpretation ↔ Self-Definition: How can people retain authority over how AI interprets and represents their experiences, identities, intentions, and values? I design ways to inspect, contest, and revise AI interpretations so that people determine their own meaning.
  • Relational Agency — Engagement ↔ Relational Autonomy: How can people establish and revise the roles, boundaries, and authority of AI-mediated interactions? I study responsible AI companions and therapist-authored conversational support that preserve user autonomy, client authority over personal meaning, and therapist clinical authority.

My interdisciplinary work spans HCI, social media, mental health, and education. I use mixed methods, including large-scale analyses of online content and conversations, interviews and surveys, and the design and evaluation of interactive systems.

Before joining Clark, I obtained my Ph.D. degree in Computer Science from Virginia Tech. My advisor was Dr. Scott McCrickard.

Quick Info

  • Research: HCI, CSCW, GenAI
  • Recent venues: CHI, CSCW, CUI
  • Lab: AI4UGC Lab
  • Student research: There is no PhD program in my department; research only with undergraduate and master’s students.

Recent News

  • Aug 2026
    I received tenure from Clark University!
  • Jan 2026
    Three Papers Accepted at CHI 2026
    When Generative AI Is Intimate, Sexy, and Violent: Examining Not-Safe-For-Work (NSFW) Chatbots on FlowGPT
    Negotiating Digital Identities with AI Companions: Motivations, Strategies, and Emotional Outcomes
    Creating Disability Story Videos with Generative AI: Motivation, Expression, and Sharing
  • Dec 2025
    One paper accepted at CSCW 2026
    Watch Me Watch: Reaction Videos as a Social Form of Online Video Engagement
  • Dec 2025
    Awarded $8,500 AI Innovation Grant by Clark University
    Topic: Developing and Teaching Meta-Prompting for Media Production, to the AI Innovation Fund.
  • Oct 2025
    One journal paper accepted by ACM Transactions on Recommender Systems
    A Literature Review of Ethical Considerations in Recommender Systems for User-Generated Content in Human-Computer Interaction
  • July 2025
    One paper received Best Paper Award at CUI 2025
    Chat with the 'For You' Algorithm: An LLM-Enhanced Chatbot for Controlling Video Recommendation Flow
  • July 2025
    Nominated to serve as a committee member at the Natural Sciences and Engineering Research Council of Canada (NSERC).
  • April 2025
    Serve as a panelist for the NSF Directorate of STEM Education.

Project Highlights

All Publications
GenAI for Constructivist Learning

Creative Agency: Learning and Design

Designing AI support that expands creative capability while preserving the thinking that makes creation meaningful. MASON scaffolds learner sensemaking through artifact creation; CLARA supports teachers’ pedagogical reasoning as they design Creative Learning lessons.

  • Surveyed 360 users about their perceptions of AI-generated learning videos.
  • Studying how MASON and CLARA can preserve human reasoning, judgment, and authorship.
GenAI for Constructivist Learning

Representational Agency: Personal Stories

Helping people retain authority over how AI represents their identities and lived experiences. Through disability storytelling and Narreflect, I study how people direct AI narratives, evaluate alternative interpretations, and revise stories to reflect their intended meaning.

  • Analyzed disability creators’ videos to understand identity expression and advocacy.
  • Interviewed disability advocacy groups about the expression and representation of lived experiences in GenAI-created stories.
  • Designing and evaluating tools for narrative reflection, self-expression, and ownership.
GenAI for Constructivist Learning

Relational Agency: AI Companions

Studying how people negotiate the roles, boundaries, and continuation of AI relationships. BOND develops creator-facing boundary controls for shared companions, examining how engagement can coexist with user autonomy and meaningful ways to revise or leave interactions.

  • Examined the risks and governance of Not-Safe-for-Work (NSFW) chatbots on FlowGPT.
  • Examined identity negotiation and emotional experiences among Character.AI users.
  • Developing BOND to support disclosure, identity, action, and exit boundaries.
GenAI for Constructivist Learning

Representational Agency: User Interests

Helping users inspect and revise how recommender systems interpret their interests. I design conversational interfaces that make those interpretations more understandable and give users ways to steer the content they encounter.

  • Designed and evaluated TKGPT, a conversational interface for controlling video recommendations.
  • Developing visualizations of the topics used to shape video flows so users can inspect recommendation logic.
GenAI for Constructivist Learning

Creative Agency: Social Media

Studying how creators direct GenAI-supported production while navigating audience engagement and perceived platform preferences. My Algorithm-Aware Strategy Work research examines how creators evaluate, adapt, or resist strategies in relation to their own goals and identity.

  • Examined creators’ practices, motivations, and risks in using GenAI.
  • Mapped GenAI use cases across the video production pipeline.
  • Developing tools that make algorithm-aware creation strategies explicit and negotiable.
GenAI for Constructivist Learning

Relational Agency: Support and Care

Examining how mediated interactions support connection and well-being. My social video studies provide foundations for relational agency research. ACTTA extends this agenda through therapist-authored conversational homework that preserves client authority over personal meaning and therapist clinical authority.

  • Studied how YouTube videos support connection and coping with loneliness.
  • Examined ASMR experiences and videos about addiction, personal experience, and advocacy.
  • Developing ACTTA to guide therapist-assigned ACT exercises between sessions within agreed roles and boundaries.

Teaching

Teaching Focus

I emphasize constructivist learning, active learning, project-based development, and responsible AI literacy. Students learn through hands-on experiences, building real systems, and critiquing design trade-offs from human-centered perspectives.

Computer Science Data Science

Courses

The course introduces foundational web-development concepts and skills for building modern full-stack applications. This course is designed for computer science majors who already have basic programming and software engineering knowledge. The goal is to let students experience front-end and back-end development by learning essential web-programming languages, having hands-on tutorials, and building real-world applications. The course focuses on the front-end but covers basic knowledge in the back-end. The course covers internet basics, HTML, CSS, JavaScript, React, RESTful API, NodeJS, and SQL/NoSQL database. Through the course, students are expected to be able to design, develop, and deploy full-stack web applications for different use cases.

The primary objective of this course is to teach how to provide software-based mobile solutions to complex problems for mobile devices. The course focuses on twelve main modules that are unique to mobile computing: Intro to Mobile Programming, Mobile GUI, Activity and Fragment, Navigation, Architecture Components, Internet and Database, Cloud Computing, Background Processing, Media and Animation, Sensors and Location, and Touch and Camera. Other advanced topics such as mobile VR and smartwatch will also be introduced. Through this course, students are expected to be able to design and develop mobile applications for different use cases, with chances to practice solving real-world problems with mobile solutions. This semester's course focuses on the Android development platform, based on the Android development language - Kotlin. The course will focus primarily on the mobile phone platform, with development opportunities for tablets, Android TV, and wearables.

This course aims to equip students with foundational knowledge in HCI and provide practical skills in analyzing user needs, designing interfaces, developing prototypes, and assessing their effectiveness. A significant component of the course is a team-based project where students will apply their skills to design innovative applications utilizing emerging Generative AI technologies, such as ChatGPT and Midjourney.

The course introduces foundational statistical and computational concepts and skills in data-centered computing and applications. It provides a toolkit of data processing and analysis methods and techniques, with hands-on opportunities for students to handle real-world datasets and extract information and knowledge from the data. The course covers data representations in Python, visualizing data, statistics and probability, data gathering and processing, intro to machine learning, regression, big data, and data ethics. Social issues surrounding data science, such as data privacy, bias, fairness, and social impacts, will also be discussed.

Develops computational problem-solving skills through programming, and exposes students to a variety of other topics from computer science and its applications. The focus of the course is to learn fundamental computational concepts (information, algorithms, abstraction, and programming) that are central to computer science, and that also happen to be instrumental for the computational investigation of science. Design, analysis, and testing of problem-solving techniques are applied to a variety of domains across the sciences and liberal arts. This is the first course for computer science majors and anyone seeking a rigorous introduction. No prior knowledge of programming is required, but good analytical skills are helpful.

Contact

If you are a Clark student and have questions about my class, you are welcome to attend my office hours as listed on Canvas or use the Schedule button on the right.
If you are interested in doing research with me, please email me with a brief introduction and a description of your research interests.
For students applying to PhD programs, please note that my department does not offer PhD programs.