TY - JOUR
T1 - On Explainable AI and Abductive Inference
AU - Medianovskyi, Kyrylo
AU - Pietarinen, Ahti Veikko
N1 - Funding Information:
The research was supported by the Estonian Research Council’s Personal Research Grant PUT 1305 (“Abduction in the Age of Fundamental Uncertainty”, PI A.-V. Pietarinen); Chinese National Funding of Social Sciences “The Historical Evolution of Logical Vocabulary and Research on Philosophical Issues” (Grant no. 20& ZD046).
Publisher Copyright:
© 2022 by the authors. Licensee MDPI, Basel, Switzerland.
PY - 2022/4
Y1 - 2022/4
N2 - Modern explainable AI (XAI) methods remain far from providing human-like answers to ‘why’ questions, let alone those that satisfactorily agree with human-level understanding. Instead, the results that such methods provide boil down to sets of causal attributions. Currently, the choice of accepted attributions rests largely, if not solely, on the explainee’s understanding of the quality of explanations. The paper argues that such decisions may be transferred from a human to an XAI agent, provided that its machine-learning (ML) algorithms perform genuinely abductive inferences. The paper outlines the key predicament in the current inductive paradigm of ML and the associated XAI techniques, and sketches the desiderata for a truly participatory, second-generation XAI, which is endowed with abduction.
AB - Modern explainable AI (XAI) methods remain far from providing human-like answers to ‘why’ questions, let alone those that satisfactorily agree with human-level understanding. Instead, the results that such methods provide boil down to sets of causal attributions. Currently, the choice of accepted attributions rests largely, if not solely, on the explainee’s understanding of the quality of explanations. The paper argues that such decisions may be transferred from a human to an XAI agent, provided that its machine-learning (ML) algorithms perform genuinely abductive inferences. The paper outlines the key predicament in the current inductive paradigm of ML and the associated XAI techniques, and sketches the desiderata for a truly participatory, second-generation XAI, which is endowed with abduction.
KW - abduction
KW - causal attributions
KW - counterfactuals
KW - explainable AI (XAI)
KW - explanation
KW - induction
KW - machine learning
KW - understanding
UR - http://www.scopus.com/inward/record.url?scp=85127609711&partnerID=8YFLogxK
U2 - 10.3390/philosophies7020035
DO - 10.3390/philosophies7020035
M3 - Article
AN - SCOPUS:85127609711
SN - 2409-9287
VL - 7
JO - Philosophies
JF - Philosophies
IS - 2
M1 - 35
ER -