More Than "Means to an End": Supporting Reasoning with Transparently Designed AI Data Science Processes

2026-03-25

AI Summary

Existing end-to-end generative AI systems struggle to support users in evaluating alternatives or reframing problems in high-stakes domains, thereby limiting effective reasoning in open-ended data science tasks. To address this, this work proposes a transparent AI workflow centered on human-readable intermediate artifacts—such as query languages, structured concept definitions, and input-output examples—which serve as “thinking tools” to enable user reflection and knowledge injection within otherwise opaque AI pipelines. Empirical evaluation in a healthcare setting demonstrates that this approach significantly enhances non-expert users’ understanding of analytical decisions, empowering them to refine their initial queries and effectively integrate domain knowledge. Consequently, the method improves both the collaborativeness and reliability of data science workflows in high-risk contexts.

Technology Category

• Humans and AI: Explainable AI (XAI) for Human Understanding
• Machine Learning: Transparent, Interpretable, Explainable ML
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Security and Privacy: Data transparency and provenance

Abstract

Generative artificial intelligence (AI) tools can now help people perform complex data science tasks regardless of their expertise. While these tools have great potential to help more people work with data, their end-to-end approach does not support users in evaluating alternative approaches and reformulating problems, both critical to solving open-ended tasks in high-stakes domains. In this paper, we reflect on two AI data science systems designed for the medical setting and how they function as tools for thought. We find that success in these systems was driven by constructing AI workflows around intentionally-designed intermediate artifacts, such as readable query languages, concept definitions, or input-output examples. Despite opaqueness in other parts of the AI process, these intermediates helped users reason about important analytical choices, refine their initial questions, and contribute their unique knowledge. We invite the HCI community to consider when and how intermediate artifacts should be designed to promote effective data science thinking.

Problem

• generative AI
• data science
• reasoning
• intermediate artifacts
• open-ended tasks

Innovation

• intermediate artifacts
• transparent AI design
• AI-assisted data science
• reasoning support
• human-AI collaboration