Querying Multimodal Scientific Papers with AI: Practices and Preferences Across Blind, Low-Vision, and Sighted Scientists
2026-07-22
Summary
The article explores how blind, low-vision (BLV), and sighted scientists use AI tools, specifically ChatGPT and Gemini, to interact with multimodal scientific papers. The study found that BLV scientists primarily use AI to interpret visual content like figures and tables, while sighted scientists use it more for synthesizing methods and findings. Both groups start with broad queries to get an overview before diving into specifics, but they face challenges due to vague AI-generated descriptions of complex visuals.
Why This Matters
Understanding how different groups of scientists use AI to access scientific papers is crucial for making scientific research more accessible and inclusive. The study highlights the need for AI systems to provide more precise and interpretive descriptions of visual content, which can impact the effectiveness and trustworthiness of AI in scientific workflows. This is particularly important for BLV scientists, who rely more heavily on AI to access visual information.
How You Can Use This Info
Professionals can leverage these insights to advocate for or develop AI tools that offer richer, more context-aware descriptions of visual content in scientific papers. This could enhance accessibility and efficiency in research workflows. Additionally, understanding the different ways BLV and sighted scientists use AI can help in designing more inclusive tools and systems that cater to diverse needs without requiring users to disclose their disabilities.