We propose and explore the concept of Partial Participation, facilitating remote collaborators to contribute to meetings in which they are not able to fully participate via an AI agent acting as a proxy. During the meeting, users can monitor LLM-generated real-time meeting updates and respond to questions posed by other attendees. Through a mixed-methods user study with 24 participants using our prototype, ProxyMe, we investigated how the frequency of updates (high vs. low) and the type of response style (multiple choice vs. text input) impact perceived presence and mental workload. Our findings reveal that no single setup is universally optimal, and the partial participation fosters a moderate level of social presence and attentional mental workload. Our contributions introduce partial participation as a new paradigm for remote collaboration and highlight how AI can mediate participation when full presence is not feasible.
Code edit histories can offer instructors valuable insight into students’ problem-solving processes, revealing unproductive behaviors that final code alone cannot capture. For example, a correct solution may contain large copy-and-pasted segments (suggesting the code originated elsewhere) or unguided trial-and-error (suggesting a lack of clear strategy). Timelines are a common way to visualize code histories, but existing timeline visualizations of code or document histories show only when and where edits occurred, not what changed. Without this context, it is difficult to answer key questions about how students invested effort or to infer their intentions. We present CodeStream, a visualization system that augments timelines with situational code annotations, whose granularity and visibility dynamically adapt to scale and interaction state. A comparison study shows that CodeStream enables context-aware navigation of coding histories, supporting fast and accurate pattern identification, and helping instructors reason about students’ coding behaviors and identify who may need intervention.
Developers now have access to a growing array of increasingly autonomous AI tools for software development. While many studies examine copilots that provide chat assistance or code completions, evaluations of coding agents—which can automatically write files and run code—still rely on static benchmarks. We present the first controlled study of developer interactions with coding agents, characterizing how more autonomous AI tools affect productivity and experience. We evaluate two leading copilot and agentic coding assistants, recruiting participants who regularly use the former. Our results show agents can assist developers in ways that surpass copilots (e.g., completing tasks humans may not have accomplished) and reduce the effort required to finish tasks. Yet challenges remain for broader adoption, including ensuring users adequately understand agent behaviors. Our findings reveal how workflows shift with coding agents and how interactions differ from copilots, motivating recommendations for researchers and highlighting challenges in adopting agentic systems.
Artificial intelligence is increasingly used in hiring, raising concerns about how applicants perceive these systems. While prior work on algorithmic fairness has emphasized technical bias mitigation, little is known about how avatar identity cues influence applicants’ justice attributions in an interview context. We conducted a crowdsourcing study with 215 participants who completed an interview with photorealistic AI avatars varied in phenotypic traits (race and sex), followed by a standardized rejection. Using self-reports, sentiment analysis, and eye tracking, we measured perceptions of trust, fairness, and bias. Results show that racial mismatch heightened perceptions of ethnic bias, while partial match (sharing only one identity) reduced fairness judgments compared to both full and no match. This work extends the Computers-Are-Social-Actors paradigm by demonstrating that avatar appearances shape justice-related evaluations of AI. We contribute to HCI by revealing how identity cues influence fairness attributions and offer actionable insights for designing equitable AI interview systems.
LLMs are widespread across educational and professional environments, often used to tackle tasks beyond users' prior expertise. {However, there is limited work on task-agnostic, turn-level frameworks to characterize human communication styles with LLMs that are linked to better task outcomes.} We introduce TurnStyle, a framework that provides a turn-level taxonomy of human contributions in human–AI conversations, grounded in collaborative learning theory, with a reliability protocol and open-source tooling for public datasets. We apply TurnStyle to two public multi-turn corpora with objective outcomes – StudyChat (college-level assignments) and DevGPT (software engineering issues and pull requests (PRs)) – and to a workplace reskilling randomized trial in which management consultants used ChatGPT for coding, statistics, and machine learning. Across 3,365 conversations with 26,335 human turns spanning the three datasets, mixed-effects and sequence analyses converge on task-agnostic, trainable behaviors that predict task success; this has implications for training, evaluation, and the design of collaborative AI systems.
Prior work has mapped which workplace tasks are exposed to AI, but less is known about whether workers perceive these tasks as meaningful or as busywork. We examined: (1) which dimensions of meaningful work do workers associate with tasks exposed to AI; and (2) how do the traits of existing AI systems compare to the traits workers want. We surveyed workers and developers on a representative sample of 171 tasks and use language models (LMs) to scale ratings to 10,131 computer-assisted tasks across all U.S. occupations. Worryingly, we find that tasks that workers associate with a sense of agency or happiness may be disproportionately exposed to AI. We also document design gaps: developers report emphasizing politeness, strictness, and imagination in system design; by contrast, workers prefer systems that are straightforward, tolerant, and practical. To address these gaps, we call for AI whose design explicitly focuses on meaningful work and worker needs, proposing a five-part research agenda.
Robot malfunctions are unavoidable in human–robot collaboration and oftentimes detrimental. Yet humans are rarely instructed on how to respond in such moments, leaving ample room for spontaneity and unpredictability. We studied 65 participants working alongside a collaborative robot under both normal operation and deliberate malfunction conditions. We analyzed unscripted vocal and action responses regarding situational awareness (SA)—whether malfunctions were noticed—and task-oriented response appropriateness—whether responses advanced or undermined the collaboration. During malfunctions, SA was universal, as was frustration and confusion, yet appropriateness diverged sharply: unscripted responses ranged from clarifying questions and corrective actions to sarcasm, comedic gestures, and intentional mismarkings. Efficiency malfunctions elicited far more productive responses than effectiveness malfunctions did, underscoring how actionability fundamentally shapes human intervention. Our findings reveal a fragile link between SA and task-aligned action, highlighting the need for robot transparency, explainability and adaptability, so collaborators are actively supported when things fail.