Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

BLIP: Facilitating the Exploration of Undesirable Consequences of Digital Technologies

Rock Yuren Pang, Sebastin Santy, René Just, Katharina Reinecke

TL;DR

BLIP addresses the problem of anticipating undesirable consequences of digital technologies by automatically extracting, summarizing, and categorizing real-world incidents from online articles into life-aspects, then presenting them in an interactive web interface. The approach combines an NLP-driven content curation pipeline (including title/content filtering, abstractive summarization via GPT-3.5, and 10-aspect categorization) with open-source extensibility and weekly updates. Two user studies with CS researchers demonstrate that BLIP increases the number and diversity of consequences considered and can yield immediately actionable insights for specific projects, while participants report enhanced reflection and idea generation. The work contributes an open-source, end-to-end system that bridges resource gaps in anticipating societal impacts and offers a practical tool to support responsible research, ethics articulation, and broader stakeholder awareness, with clear directions for expanding sources and integrating into academic and policy workflows.

Abstract

Digital technologies have positively transformed society, but they have also led to undesirable consequences not anticipated at the time of design or development. We posit that insights into past undesirable consequences can help researchers and practitioners gain awareness and anticipate potential adverse effects. To test this assumption, we introduce BLIP, a system that extracts real-world undesirable consequences of technology from online articles, summarizes and categorizes them, and presents them in an interactive, web-based interface. In two user studies with 15 researchers in various computer science disciplines, we found that BLIP substantially increased the number and diversity of undesirable consequences they could list in comparison to relying on prior knowledge or searching online. Moreover, BLIP helped them identify undesirable consequences relevant to their ongoing projects, made them aware of undesirable consequences they "had never considered," and inspired them to reflect on their own experiences with technology.

BLIP: Facilitating the Exploration of Undesirable Consequences of Digital Technologies

TL;DR

BLIP addresses the problem of anticipating undesirable consequences of digital technologies by automatically extracting, summarizing, and categorizing real-world incidents from online articles into life-aspects, then presenting them in an interactive web interface. The approach combines an NLP-driven content curation pipeline (including title/content filtering, abstractive summarization via GPT-3.5, and 10-aspect categorization) with open-source extensibility and weekly updates. Two user studies with CS researchers demonstrate that BLIP increases the number and diversity of consequences considered and can yield immediately actionable insights for specific projects, while participants report enhanced reflection and idea generation. The work contributes an open-source, end-to-end system that bridges resource gaps in anticipating societal impacts and offers a practical tool to support responsible research, ethics articulation, and broader stakeholder awareness, with clear directions for expanding sources and integrating into academic and policy workflows.

Abstract

Digital technologies have positively transformed society, but they have also led to undesirable consequences not anticipated at the time of design or development. We posit that insights into past undesirable consequences can help researchers and practitioners gain awareness and anticipate potential adverse effects. To test this assumption, we introduce BLIP, a system that extracts real-world undesirable consequences of technology from online articles, summarizes and categorizes them, and presents them in an interactive, web-based interface. In two user studies with 15 researchers in various computer science disciplines, we found that BLIP substantially increased the number and diversity of undesirable consequences they could list in comparison to relying on prior knowledge or searching online. Moreover, BLIP helped them identify undesirable consequences relevant to their ongoing projects, made them aware of undesirable consequences they "had never considered," and inspired them to reflect on their own experiences with technology.
Paper Structure (40 sections, 6 figures, 1 table)

This paper contains 40 sections, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. Discussions and interventions for addressing undesirable consequences in research
  4. Methods for forecasting and anticipating undesirable consequences
  5. Supporting ideation of potential undesirable consequences
  6. Blip
  7. Design Choices and Rationale
  8. User Interface Usage Scenario
  9. Content Curation Pipeline
  10. Article Filtering
  11. Filtering by Title
  12. Filtering by Content
  13. Content Summarizing
  14. Aspect Categorizing
  15. Implementation Details and Costs
  16. ...and 25 more sections

Figures (6)

  • Figure 1: Blip's main user interface. Users can view summaries of undesirable consequences and filter them by technology domains, aspects of life they affect, or search keywords.
  • Figure 2: An overview of Blip's content curation pipeline. Given online article sources, Blip filters out those that discuss undesirable consequences, extracts and summarizes the consequence, and categorizes it into different aspects of life that it affects, such as the environment, equality, or politics. The undesirable consequences are then displayed in an interactive, web-based user interface.
  • Figure 3: Cumulative unique number of consequences in each condition. The line represents means of unique consequences between the three conditions.
  • Figure 4: Total time taken for each of the three conditions separately (in minutes).
  • Figure 6: Quotes from P5's reporting of of voice assistants and the associated aspects of life in the three conditions of our user study.
  • ...and 1 more figures