Fairness in crowdwork: Making the human AI supply chain more humane

7 citations

Abstract

The vast quantities of data required to build artificial intelligence (AI) technologies are often annotated and processed manually, making human labor a critical component of the AI supply chain. The workers who input this data are sourced through digital labor ("crowdwork") platforms that often are unregulated and offer low wages, raising concerns about labor standards in AI development. Using the results of a survey, this article aims to shed light on the experiences and perceptions of fair treatment among workers in the AI supply chain. The study reveals significant variability in workers' experiences, identifies potential drivers of fairness, and highlights how design choices by labor platforms can significantly affect worker welfare. Drawing on lessons from physical supply chains, this article offers practical guidance to managers on how to enhance worker welfare within the AI supply chain and how to ensure that AI technologies are responsibly sourced.

7
Citations
Research
Paper Only

Study specs

Year
2024
Human Data Platform
Prolific

Peer Review & Critical Discussion

3 threads

Potential Selection Bias in 2023 Cohort

DSJDr. Sarah J.
Verified PhD Candidate
12 replies

The participant pool shows a concerning overrepresentation of users from high-income demographics. Looking at Table 3, we can see that 78% of respondents had annual incomes above $75k, which significantly limits the generalizability of these findings to broader populations.

2 hours ago

Non-naive Participants Issue

MCM. Chen (OpenAI)
Data Scientist
8 replies

I've noticed a methodological concern regarding participant naivety. Given that Prolific users often complete multiple studies, there's a real risk that participants had prior exposure to similar experimental paradigms, which could confound the results.

5 hours ago

RLHF Applicability to This Study Design

PRWProf. R. Williams
Verified Researcher
15 replies

The implications for RLHF training pipelines are understated. If we accept the authors' conclusions about preference stability, this has direct consequences for how we should structure reward model training. The temporal decay effect described in Section 4.2 is particularly relevant.

1 day ago

Verify your expertise to join discussion

Create an account and verify your credentials to participate in peer discussions.