Episteme 19 (4):534-560 (2022)
AbstractMany AI systems that make important decisions are black boxes: how they function is opaque even to their developers. This is due to their high complexity and to the fact that they are trained rather than programmed. Efforts to alleviate the opacity of black box systems are typically discussed in terms of transparency, interpretability, and explainability. However, there is little agreement about what these key concepts mean, which makes it difficult to adjudicate the success or promise of opacity alleviation methods. I argue for a unified account of these key concepts that treats the concept of understanding as fundamental. This allows resources from the philosophy of science and the epistemology of understanding to help guide opacity alleviation efforts. A first significant benefit of this understanding account is that it defuses one of the primary, in-principle objections to post hoc explainable AI (XAI) methods. This “rationalization objection” argues that XAI methods provide mere rationalizations rather than genuine explanations. This is because XAI methods involve using a separate “explanation” system to approximate the original black box system. These explanation systems function in a completely different way than the original system, yet XAI methods make inferences about the original system based on the behavior of the explanation system. I argue that, if we conceive of XAI methods as idealized scientific models, this rationalization worry is dissolved. Idealized scientific models misrepresent their target phenomena, yet are capable of providing significant and genuine understanding of their targets.
Similar books and articles
Scientific Exploration and Explainable Artificial Intelligence.Carlos Zednik & Hannes Boelsen - 2022 - Minds and Machines 32 (1):219-239.
Idealization and the Aims of Science.Angela Potochnik - 2017 - Chicago: University of Chicago Press.
Exemplification, idealization, and scientific understanding.Catherine Elgin - 2009 - In Mauricio Suárez (ed.), Fictions in Science: Philosophical Essays on Modeling and Idealization. Routledge. pp. 77-90.
The Pragmatic Turn in Explainable Artificial Intelligence (XAI).Andrés Páez - 2019 - Minds and Machines 29 (3):441-459.
Solving the Black Box Problem: A Normative Framework for Explainable Artificial Intelligence.Carlos Zednik - 2021 - Philosophy and Technology 34 (2):265-288.
The virtues of interpretable medical artificial intelligence.Joshua Hatherley, Robert Sparrow & Mark Howard - forthcoming - Cambridge Quarterly of Healthcare Ethics:1-10.
Solving the Black Box Problem: A Normative Framework for Explainable Artificial Intelligence.Carlos Zednik - 2019 - Philosophy and Technology 34 (2):265-288.
AI employment decision-making: integrating the equal opportunity merit principle and explainable AI.Gary K. Y. Chan - forthcoming - AI and Society:1-12.
Understanding Galileo’s Inquiries About the Law of Inertia.Yeon-A. Son, Byeong-Mee Min, Chun-Hwey Kim, YooShin Kim & Jun-Young Oh - 2015 - In Woosuk Park, Ping Li & Lorenzo Magnani (eds.), Philosophy and Cognitive Science Ii. Springer Verlag. pp. 193-207.
“When I Was Young” The Idealization of the Interchangeability of Phases of Life.Andreas Göttlich - 2013 - Human Studies 36 (2):217-233.
Added to PP
Historical graph of downloads
Citations of this work
No citations found.
References found in this work
Understanding from Machine Learning Models.Emily Sullivan - 2022 - British Journal for the Philosophy of Science 73 (1):109-133.
How the machine ‘thinks’: Understanding opacity in machine learning algorithms.Jenna Burrell - 2016 - Big Data and Society 3 (1):205395171562251.