Talking About the Assumption in the Room

TL;DR

The paper reframes assumptions in the ML workflow as premises within an argument, using Informal Logic to explain persistent confusions. Through 22 semi-structured interviews, it identifies two core ontologies—independent (external axioms) and relative (embedded within workflow)—and several procedural uncertainties that hinder reflective handling and documentation. It proposes an articulation framework based on a premise-target lens with differentiators for understanding, action, and addressal, plus explicit/implicit status, and argues for assumptions inquiry as a lateral, parallel process to standard workflows. Collectively, the work contributes an empirical, theoretical, and practical lens to center, articulate, and improve how assumptions are identified, justified, and recorded in responsible AI practice.

Abstract

The reference to assumptions in how practitioners use or interact with machine learning (ML) systems is ubiquitous in HCI and responsible ML discourse. However, what remains unclear from prior works is the conceptualization of assumptions and how practitioners identify and handle assumptions throughout their workflows. This leads to confusion about what assumptions are and what needs to be done with them. We use the concept of an argument from Informal Logic, a branch of Philosophy, to offer a new perspective to understand and explicate the confusions surrounding assumptions. Through semi-structured interviews with 22 ML practitioners, we find what contributes most to these confusions is how independently assumptions are constructed, how reactively and reflectively they are handled, and how nebulously they are recorded. Our study brings the peripheral discussion of assumptions in ML to the center and presents recommendations for practitioners to better think about and work with assumptions.