Flexible and Context-Specific AI Explainability: A Multidisciplinary Approach

Abstract

The recent enthusiasm for artificial intelligence (AI) is due principally to advances in deep learning. Deep learning methods are remarkably accurate, but also opaque, which limits their potential use in safety-critical applications. To achieve trust and accountability, designers and operators of machine learning algorithms must be able to explain the inner workings, the results and the causes of failures of algorithms to users, regulators, and citizens. The originality of this paper is to combine technical, legal and economic aspects of explainability to develop a framework for defining the "right" level of explainability in a given context. We propose three logical steps: First, define the main contextual factors, such as who the audience of the explanation is, the operational context, the level of harm that the system could cause, and the legal/regulatory framework. This step will help characterize the operational and legal needs for explanation, and the corresponding social benefits. Second, examine the technical tools available, including post hoc approaches (input perturbation, saliency maps...) and hybrid AI approaches. Third, as function of the first two steps, choose the right levels of global and local explanation outputs, taking into the account the costs involved. We identify seven kinds of costs and emphasize that explanations are socially useful only when total social benefits exceed costs. arXiv:2003.07703v1 [cs.CY]

Identifying the 'Right' Level of Explanation in a Given Situation

“…This paper summarizes the conclusions of a longer paper [1] on context-specific explanations using a multidisciplinary approach. Explainability is both an operational and ethical requirement.…”

“…Although several different definitions exist in the literature [1], we have treated explainability and interpretability as synonyms [16], focusing instead on the key difference between "global" and "local" explainability/interpretability. Global explainability means the ability to explain the functioning of the algorithm in its entirety, whereas local explainability means the ability to explain a particular algorithmic decision [7].…”

“…• Input approaches, which pre-process the inputs to the machine learning model, making the inputs more meaningful and/or better structured [1]; • Genetic fuzzy logic.…”

Towards a maturity model of human-centered AI – A reference for AI implementation at the workplace

“…intuitiveness, usability; Bauer et al 2021). Explainable AI is thus considered a multidimensional challenge, as it involves not only technical trade-offs between prediction accuracy and transparency of results but also political and societal efforts (Beaudouin et al 2020).…”

Hybrid Cognitive-Affective Strategies for AI Safety

“…In the last years, the general requirement for a multidisciplinary approach to AI has as well been stressed e.g. in the domains of AI ethics [146], AI governance [127], responsible AI [336] and explainable AI [63].…”