Maintaining "Balanced" Conflict: Proactive Intervention Strategies of AI Voice Agents in Online Collaboration of Temporary Design Teams

Research Background and Issues

• What problems or challenges did the authors identify?

  • Temporary design teams often face communication barriers, coordination difficulties, and low efficiency in online collaboration due to low familiarity and lack of trust among members.
  • Conflict is an inevitable factor in team collaboration. While moderate conflict can promote team innovation and decision quality, excessive or insufficient conflict can negatively impact the team.
  • Current research primarily relies on human leaders or coordinators for conflict management, which is difficult to scale in geographically dispersed online environments.
  • Existing studies overlook the bidirectional regulation of conflict escalation and mitigation in team dynamics, as well as the potential of using artificial intelligence to manage conflicts.

• Why is this issue important?

  • Temporary design teams often need to quickly establish connections and collaborate effectively. Failure to manage conflicts appropriately can lead to strained relationships and hinder goal achievement.
  • Successfully managing conflict dynamics can not only enhance team productivity but also significantly improve relationships and collaboration experiences among members.

• Research Motivation and Related Work

  • While previous studies have explored the role of AI in facilitating team collaboration, there is a lack of research on how AI can manage conflicts, particularly in regulating conflict escalation and mitigation mechanisms.
  • This study addresses the gap by investigating how AI can manage team conflicts through proactive intervention strategies and explores the potential of AI voice assistants in improving collaboration within temporary design teams.

Solution

• What methods or solutions did the authors propose?

  • Developed a set of proactive intervention strategies for AI voice agents to manage conflicts in online collaboration within temporary design teams.
  • Identified five types of conflict management (ineffective conflict, obstructive conflict, constructive conflict, unidirectional events, non-directional events) and six conflict management strategies (resolve, mitigate, transform, promote, guide, ignore).
  • Conducted experiments to explore the practical effectiveness of AI voice agents in managing conflicts within short-term design teams.

• What are the innovative aspects of this solution?

  • Introduced a new model for AI voice agents to actively participate in human interactions, moving beyond passive service.
  • Provided a systematic theoretical framework for bidirectional dynamic regulation of conflicts.
  • Incorporated "role positioning" and "intervention styles" to refine the adaptability of AI voice agents in various conflict scenarios.

• What are the implementation steps and key technologies used?

  1. Exploratory Study: Observed and conducted interviews to summarize conflict types and needs in team collaboration.
  2. Expert Workshops: Collaborated with domain experts to design intervention strategies for AI voice agents, including roles, styles, and dialogue content.
  3. Strategy Validation Experiment: Simulated AI voice agent functionality using the Wizard of Oz method and tested its effectiveness in real tasks.
  4. Data Analysis: Verified intervention effectiveness through multidimensional analysis of team video recordings, self-reported questionnaires, and scoring.

Research Outcomes

• What specific outcomes were achieved?

  • Identified five main types of conflict management and appropriate intervention strategies.
  • Validated that AI voice agents can effectively enhance team familiarity and cohesion, balance task conflicts, and reduce relational conflicts.
  • Found that AI voice agents significantly improved communication quality and overall satisfaction within teams.

• What advantages does it have compared to existing solutions?

  • Achieved bidirectional regulation of team conflict dynamics rather than simply avoiding conflicts.
  • Provided a more natural, immediate, and efficient voice intervention method, which is better suited for discussion-intensive team collaboration compared to traditional text prompts.
  • Emphasized a balance between user autonomy and intervention timing, reducing the intrusive nature of proactive interventions.

• What are the experimental or evaluation results?

  • Comparative experiments showed that AI voice agents significantly reduced unidirectional events and increased the frequency of constructive conflicts.
  • Teams with AI voice agent interventions demonstrated significantly better familiarity, cohesion, and collaboration experiences compared to teams without AI assistance.
  • Facilitated higher consensus among team members and optimized the quality of dialogue management and information comprehension.
  • Participants expressed high satisfaction with the use of AI voice agents and showed strong intentions for future use.

• Limitations and Future Directions

  • Limitations:
    • Experiments focused on short-term, low-risk tasks; applicability to long-term, high-risk tasks requires further validation.
    • Research centered on initial experiences with AI voice agents; behavioral changes over prolonged use remain to be observed.
    • Participants predominantly consisted of young design teams, lacking diversity across industries and team compositions.
    • Specific effects of voice attributes (e.g., tone) and multimodal signals (e.g., visual and auditory cues) have not been thoroughly explored.
  • Future Directions:
    • Test the robustness and adaptability of strategies in more complex task scenarios.
    • Explore ways to enable AI voice agents to better balance user autonomy and clarity.
    • Further optimize conflict management strategies by integrating predictive models with real-time intervention technologies to enhance AI efficiency.

Conclusion

This study successfully explored the potential of AI voice agents in managing and facilitating dynamic conflicts within temporary design teams during online collaboration. The research not only proposed and validated systematic intervention strategies but also paved the way for further applications of artificial intelligence in facilitating complex human collaboration.