The Epistemology of Machine Learning

Bibliographic Data

ID19370459
AuthorsHuiren Bai (Zhejiang University, corresponding author)
Year2022
Volume33
Issue1
Publication date2022-02-18
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueFilosofija. Sociologija (JOURNAL)
Journal identifiersISSN: 0235-7186 • E-ISSN: 2424-4546
PublisherLithuanian Academy of Sciences (PUBLISHER • LT)
DOI10.6001/fil-soc.v33i1.4668
OpenAlexW4212842256
LanguageEN
Citations received1
References cited8

This paper argues that machine learning is a knowledge-producing enterprise, since we are increasingly relying on artificial intelligence. But the knowledge discovered by machine is completely beyond human experience and human reason, becoming almost incomprehensible to humans. I argue that standard calls for interpretability that focus on the epistemic inscrutability of black-box machine learning may be misplaced. The problems of transparency and interpretability of machine learning stem from how we perceive the possibility of ‘machine knowledge’. In other words, the justification for machine knowledge does not need to include transparency and interpretability. Therefore, I am going to examine some sort of machine learning epistemology and provide three possible justifications for machine knowledge, which are formal justification, model justification and practical justification

Cognitive science · Epistemology · Information retrieval · Interpretability · Machine learning · sort · Computer Science · Epistemology, Ethics, and Metaphysics · Ethics and Social Impacts of AI · Explainable Artificial Intelligence (XAI · Philosophy · Psychology · Artificial Intelligence

Tools

Open DOIOpen Access