[Paper Review] Explainable Artificial Intelligence (XAI) 2.0: A Manifesto of Open Challenges and Interdisciplinary Research Directions

Luca Longo, Mario Brčić+17 more|arXiv (Cornell University)|Oct 30, 2023

Artificial Intelligence in Healthcare and EducationMedicine240 references14 citations

TL;DR

A collaborative manifesto outlining 27 open XAI challenges across nine categories, proposing interdisciplinary directions to advance explainable AI and its real-world deployment.

ABSTRACT

As systems based on opaque Artificial Intelligence (AI) continue to flourish in diverse real-world applications, understanding these black box models has become paramount. In response, Explainable AI (XAI) has emerged as a field of research with practical and ethical benefits across various domains. This paper not only highlights the advancements in XAI and its application in real-world scenarios but also addresses the ongoing challenges within XAI, emphasizing the need for broader perspectives and collaborative efforts. We bring together experts from diverse fields to identify open problems, striving to synchronize research agendas and accelerate XAI in practical applications. By fostering collaborative discussion and interdisciplinary cooperation, we aim to propel XAI forward, contributing to its continued success. Our goal is to put forward a comprehensive proposal for advancing XAI. To achieve this goal, we present a manifesto of 27 open problems categorized into nine categories. These challenges encapsulate the complexities and nuances of XAI and offer a road map for future research. For each problem, we provide promising research directions in the hope of harnessing the collective intelligence of interested stakeholders.

Motivation & Objective

• Synthesize diverse expert perspectives to identify open problems in XAI.
• Propose a coordinated, cross-disciplinary research agenda to advance XAI 2.0.
• Highlight practical applications and real-world relevance of explainability.
• Offer directions to synchronize research across domains and stakeholders.

Proposed method

• Synthesize expert input from philosophy, psychology, HCI, and computer science to extract open problems.
• Organize problems into nine categories comprising 27 specific challenges.
• Provide proposed research directions and potential solution approaches for each problem.

Figure 1: A manifesto for eXplainable Artificial Intelligence (XAI): High-level challenges

Experimental results

Research questions

• RQ1 What are the key open problems in XAI identified by a multidisciplinary panel?
• RQ2 How can XAI research be synchronized across disciplines to accelerate real-world adoption?
• RQ3 What directions and approaches are proposed to advance XAI 2.0 for generative, concept-based, and robust explanations?

Key findings

• Identification of 27 open problems framed as a nine-category manifesto for XAI 2.0.
• Emphasis on challenges in explaining new AI types (generative models, concept-based learning) and improving current XAI methods.
• Discussion of evaluation, robustness, and human-centered assessment gaps in XAI explanations.
• Illustration of application domains including medicine, finance, environment, and education to motivate interdisciplinary work.