From Model Performance to Claim: How a Change of Focus in Machine Learning Replicability Can Help Bridge the Responsibility Gap

2024-04-19

Conference on Fairness, Accountability and Transparency

AI Summary

This paper addresses the “responsibility gap”—a critical ethical challenge in machine learning research arising when researchers operate detached from real-world application contexts. To bridge this gap, we propose “claim replicability” as a normative alternative to conventional “model performance replicability.” Drawing on conceptual analysis and interdisciplinary frameworks—including Latour’s notion of circular reference, AI ethics, and sociology of science—we systematically distinguish these two forms of replicability and argue that claim replicability better grounds ethical accountability for societal harms stemming from non-replicable claims. We reconceptualize replicability as a socially embedded practice involving explanatory labor and iterative referential alignment, rather than a purely technical property. Our work yields an actionable responsibility mechanism for ML research, fostering coordinated evolution of scholarly communication norms, interpretive practices, and ethical accountability structures. (149 words)

Technology Category

• Machine Learning: Ethics, Bias, and Fairness
• Philosophy and Ethics of AI: Accountability, Interpretability & Explainability
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems

Abstract

Two goals – improving replicability and accountability of Machine Learning research respectively, have accrued much attention from the AI ethics and the Machine Learning community. Despite sharing the measures of improving transparency, the two goals are discussed in different registers - replicability registers with scientific reasoning whereas accountability registers with ethical reasoning. Given the existing challenge of the responsibility gap – holding Machine Learning scientists accountable for Machine Learning harms due to them being far from sites of application, this paper posits that reconceptualizing replicability can help bridge the gap. Through a shift from model performance replicability to claim replicability, Machine Learning scientists can be held accountable for producing non-replicable claims that are prone to eliciting harm due to misuse and misinterpretation. In this paper, I make the following contributions. First, I define and distinguish two forms of replicability for ML research that can aid constructive conversations around replicability. Second, I formulate an argument for claim-replicability’s advantage over model performance replicability in justifying assigning accountability to Machine Learning scientists for producing non-replicable claims and show how it enacts a sense of responsibility that is actionable. In addition, I characterize the implementation of claim replicability as more of a social project than a technical one by discussing its competing epistemological principles, practical implications on Circulating Reference, Interpretative Labor, and research communication.

Problem

• Bridging responsibility gap in ML via replicability shift
• Shifting focus from model to claim replicability
• Enhancing accountability for non-replicable ML claims

Innovation

• Shift from model performance to claim replicability
• Accountability via non-replicable claims justification
• Claim replicability as a social project