Why is AI not a Panacea for Data Workers? An Interview Study on Human-AI Collaboration in Data Storytelling

TL;DR

This work addresses how data workers view AI collaboration in data storytelling by conducting 18 interviews across academia and industry. It reveals a holistic workflow—planning, implementation, and communication—with seven tasks and four AI roles (creator, optimizer, reviewer, assistant). Key insights show both enthusiasm for AI to reduce repetitive work and concerns about context understanding, communication, overhead, and ethics, guiding modular, audience-aware tool design. The study offers design implications for future AI-powered data storytelling tools, emphasizing flexibility, adaptive behavior, and guardrails to ensure trustworthy, effective human-AI collaboration in storytelling contexts.

Abstract

Data storytelling plays an important role in data workers' daily jobs since it boosts team collaboration and public communication. However, to make an appealing data story, data workers spend tremendous efforts on various tasks, including outlining and styling the story. Recently, a growing research trend has been exploring how to assist data storytelling with advanced artificial intelligence (AI). However, existing studies may focus on individual tasks in the workflow of data storytelling and do not reveal a complete picture of humans' preference for collaborating with AI. To better understand real-world needs, we interviewed eighteen data workers from both industry and academia to learn where and how they would like to collaborate with AI. Surprisingly, though the participants showed excitement about collaborating with AI, many of them also expressed reluctance and pointed out nuanced reasons. Based on their responses, we first characterize stages and tasks in the practical data storytelling workflows and the desired roles of AI. Then the preferred collaboration patterns in different tasks are identified. Next, we summarize the interviewees' reasons why and why not they would like to collaborate with AI. Finally, we provide suggestions for human-AI collaborative data storytelling to hopefully shed light on future related research.

Figures (3)

• Figure 1: This figure summarizes participants' opinions about where and how they would like to collaborate with AI. (a) shows tasks in the existing workflows of our interviewees and (b) illustrates the expected AI collaborators' roles and their tasks. Both y-axes in (a) and (b) are the tasks in the data storytelling workflow. In (b), the heatmap presents the breakdown frequency of task-role tuples. The bar chart on the top counts the frequency of roles, while the one on the right shows the frequency of tasks. Notably, since each participant can propose multiple roles of AI collaborators for one task, the count of tasks can be larger than the number of participants.
• Figure 2: This figure summarizes (1) the interviewees' opinions about why and why not AI collaborators are preferred and (2) our suggestions for future AI-powered data storytelling tools.
• Figure 3: This figure demonstrates AI collaborators' roles against AI automation and human agency. We use fully manual story creation and Heer's ideal "agency plus automation" collaboration mode heer2019agency as two anchor points and derive the relative rough positions of other roles.