"It Might be Technically Impressive, But It’s Practically Useless to us": Motivations, Practices, Challenges, and Opportunities for Cross-Functional Collaboration around AI within the News Industry

Research Background and Issues

• What problems or challenges did the authors identify?
  This study identifies several key issues in cross-role collaboration within the journalism industry due to the complexity and widespread application of artificial intelligence (AI) technologies. These issues include communication barriers, power imbalances, and the marginalization of data workers in collaborative processes. Additionally, the traditionally journalist-dominated power structure in the industry further sidelines the roles of engineers and data workers, hindering interdisciplinary innovation.

• Why is this issue important?
  AI technologies are profoundly transforming the processes of news production and distribution. However, there is a significant mismatch between the professional culture of journalism (which prioritizes editorial authority) and the practice logic of technical teams (which is technology-driven). If these cross-role collaboration issues are not addressed, they will not only hinder innovation in the journalism industry but may also undermine media independence and institutional autonomy.

• Research Motivation and Related Work
  Previous research has revealed the increasing adoption of AI technologies in the journalism industry but has not delved deeply into the specific processes and challenges of internal collaboration between journalists and AI technologists. This study fills that gap by proposing methods to improve interdisciplinary collaboration and exploring ways to integrate journalistic values with technical requirements.

Solutions

• What methods or solutions did the authors propose?
  Through interviews and workshops, the authors proposed a series of solutions, including fostering a shared language, introducing co-design and prototyping mechanisms, and implementing more inclusive collaboration processes. They also emphasized the importance of marginalized groups (e.g., data annotators) and advocated for a fairer distribution of participation rights and responsibilities within newsroom power structures.

• What is innovative about this solution?
  The core innovations include:

  1. Translating abstract journalistic values (e.g., "impartiality" or "newsworthiness") into technically actionable parameters.
  2. Enhancing understanding and communication across roles through co-design and rapid prototyping.
  3. Advocating for the transformation of traditional newsroom power structures to provide technologists and data workers with greater participation and contribution opportunities in news innovation.

• What are the implementation steps? What key technologies were used?
  The implementation involves the following steps:

  1. Creating shared language tools (e.g., glossaries): Standardizing journalistic concepts and technical terms within the team.
  2. Co-design and rapid prototyping: Conducting iterative prototype testing and feedback between journalists and technical teams.
  3. Data process transparency: Ensuring all participants understand the importance of data work and collaborate through shared workflows.
  4. Redefining newsroom power structures: Introducing mentorship programs and equitable career development pathways to provide AI workers with greater visibility and professional support.

Research Outcomes

• What specific outcomes were achieved?
  This study identified the main processes and challenges of cross-functional collaboration and proposed a series of actionable recommendations, such as using tools to support prototyping and shared workflows, and fostering a more inclusive team culture.

• What advantages does it have compared to existing solutions?
  Compared to traditional research that focuses solely on technical efficiency, this study emphasizes the integration of journalistic values with AI technological capabilities. It highlights the importance of fairness in collaboration and dynamic feedback mechanisms across teams, offering new perspectives to address the power imbalances in the journalism industry.

• What were the experimental or evaluation results?
  Through in-depth interviews and workshops, the authors found that co-design, building a shared language, and iterative feedback significantly improved collaboration efficiency and the practical applicability of tools introduced by AI in newsrooms.

• Limitations and Future Directions
  Limitations include:

  1. This study primarily focuses on the specific journalism environment in China, which may lack global applicability.
  2. The investigation sample is limited to large news organizations, excluding small and medium-sized news outlets.
  3. The diversity of roles within data workers (e.g., content moderators) has not been explored.

  Future research could examine collaboration models in different geopolitical contexts, expand to smaller news organizations and other professional groups, and incorporate partnerships with external technology companies. Additionally, integrating organizational behavior theories and mixed research methods could provide more comprehensive insights.