General Examination - Varun Rao
Committee: Andrés Monroy-Hernández (advisor), Arvind Narayanan, Peter Henderson
General Exam Period: April 15 - May 17, 2024
Room and Timing: May 15, 3-5pm Sherrerd Hall 008
Zoom Link: https://princeton.zoom.us/j/97776899839
AI systems are increasingly mediating interactions on digital labor marketplaces, reshaping the nature of work and impacting billions of lives through opaque decisions. In this talk, I will explore the societal impact of AI on labor through the lens of rideshare and social media job ads. In the case of rideshare[1], through a unique mixed methods study design blending worker interviews with large scale LLM-assisted analysis of over 1 million driver comments on Reddit, we thickly characterize transparency-related harms, mitigation strategies, and worker needs. Motivated by our findings we propose an outline of the first rideshare transparency report. In the case of job ads on social media[2], our analysis of the Facebook Ad Library reveals a new form of discrimination through the selective use of demographic-specific images in job ads, affecting applicant demographics and reinforced by the ad delivery algorithm. Taken together, my research shows that AI-mediated platforms substantially shape job access for many through opaque, centralized decision-making which restricts user agency. Policy solutions mandating greater transparency may help with the decentralization of some platform power.
Slide Deck: Varun_Research_Overview
Summary Notes: Varun's Reading List Notes
Foundational Studies in CSCW, HCI and Social Computing:
Labor, Gig Work and Policy:
Social Media Advertising and Algorithmic Audits:
Textbooks:
Ch 2 When is automated decision making legitimate?
Ch 3 Classification
Ch 4 Relative notions of fairness
Ch 7 A broader view of discrimination
Ch 4 Regulating Behavior in Online Communities
Ch 6 Starting New Online Communities
Ch 4 The use of normative theories in computer ethics.
[1] Rao et. al., Navigating Rideshare Transparency: Worker Insights on AI Platform Design (in submission); Rao et. al., QuaLLM: An LLM-based Framework to Extract Qualitative Insights from Online Forums (in submission)
[2] Rao et. al., Discrimination through Image Selection by Job Advertisers on Facebook (FAccT 2023)