Overview
The course is a postgraduate series of seminars addressing selected topics of artificial intelligence. The course is structured into two branches. The AI Robolab is responsible for Branch II: AI Ethics and Explainability.
AI Ethics and Explainability
With the widespread use of sophisticated AI applications,researchers started to study the underlying ethical principles guiding these systems. Furthermore, this tendency has been encouraged by recent calls, white papers, and guidance documents stressing the importance of AI Ethics and AI Transparency. In this first part, this course gives an overview of recent research works addressing AI Ethics, their applications, and their implementation in robots and multi-agent systems. In the second part, the course focuses on introducing eXplainable AI (XAI) and presenting its applications in machine learning (interpretable ML and DNNs, LIME, etc.) and explainable agents and robots.
Form
The course supervisor defines a list of research articles. Each lecture, a research article is presented by a student followed by a discussion animated by the course supervisor and questions asked by other student participants.
Aims
In addition to introducing the students to this recent topic, the course aims to make students familiar with research methodology and help them get used to reading scientific articles, identifying their contributions, limitations, positioning them vis-à-vis other related research works, and highlight the future research perspectives
List of lectures
Introductory lecture
Overview of the course content and structure, presentation of further topics.
Reading task for students in the branch "AI Ethics":
- An Introduction to Ethics in Robotics and AI (Bartneck, Lütge, Wagner, Welsh) Chapter 3 (p. 17-26)
- The Nature, Importance, and Difficulty of Machine Ethics (Moor) (p. 1-4)
Teachers : Sana Nouzri, Amro Najjar
Lecture 1
Topic.1 Machine and AI Ethics
Title : Why Machine Ethics? (Allen, Wallace, Smit)
Name of student : KUMARI Archana
Teacher : Amro Najjar
Date : November 5, 2020
Papers
+ Why Machine Ethics? (Allen, Wallace, Smit) (p. 2-7)
+ Incorporating Ethics into Artificial Intelligence (Etzioni, Etzioni) (p. 1-16)
Presentation
Lecture 2
Topic.2 Ethics of Autonomous Vehicles and other applications
Title : Psychological roadblocks to the adoption of self-driving vehicles
Name of student : RAHMAN Md Mafizur
Teacher : Sana Nouzri
Date : November 10,2020
Papers
+Psychological roadblocks to the adoption of self-driving vehicles (Shariff, Bonnefon, Rahwan) - (p.2-9)
+The Ethical Governance of the Digital During and After the COVID‑19 Pandemic(Mariarosaria Taddeo1) - (p.1-6)
Presentation
Psychological roadblocks to the adoption of self-driving vehicles
Lecture 3
Topic.3 Ethics and Intelligent/Autonomous Agent
Title : Building Jiminy Cricket: An Architecture for Moral Agreements Among Stakeholder
Name of student : CAETANO Nélson
Teacher : Amro Najjar
Date : November 12,2020
Papers
+ Building Jiminy Cricket: An Architecture for Moral Agreements Among Stakeholders (Liao, Slavkovik, van der Torre) (p.1-15)
+ On The Morality of Artificial Agents (Floridi, Sanders) (p.1-29)
Presentation
Building Jiminy Cricket: An Architecture for Moral Agreements Among Stakeholders
Lecture 4
Topic.4 Ethics in Multi-Agent Systems
Title : Ethical Judgment of Agents' Behaviors in Multi-Agent Systems
Name of student : BORCHULUUN Maitsetseg
Teacher : Amro Najjar
Date : November 17,2020
Papers
+ Ethical Judgment of Agents' Behaviors in Multi-Agent Systems (Cointe, Bonnet, Boissier) - (p.1-10)
+ A Voting-Based System for Ethical Decision Making (Noothigattu et al.) - 1-2 (p.1-25)
Presentation
Ethical Judgment of Agents' Behaviors in Multi-Agent Systems
Lecture 5
Topic.5 Ethics and Robots
Title : An architecture for ethical robots inspired by the simulation theory of cognition
Name of student : MOAD Hani
Teacher : Amro Najjar
Date : November 19,2020
Papers
+ Towards the Ethical Robot (Gips)(p. 2-13)
+ An architecture for ethical robots inspired by the simulation theory of cognition (Vanderelst, Winfield) (p.1-11)
+ Ethical Decision Making in Robots: Autonomy, Trust and Responsibility (Alaieri, Vellino) (P1-10)
Presentation
An architecture for ethical robots inspired by the simulation theory of cognition
Lecture 6
Topic. 6 Data Driven and Goal Driven XAI
Title : Explainable agents and robots: Results from a systematic literature review (Anjomshoae, Najjar, Calvaresi, Framling) AND Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) (Adadi & Berrada)"
Name of student : HO HaoCheng
Teacher : Amro Najjar
Date : November 24,2020
Papers
+ Explainable agents and robots: Results from a systematic literature review (Anjomshoae, Najjar, Calvaresi, Framling) AND Peeking Inside the Black-Box: A Survey on
+Peeking Inside the Black-Box- A Survey on Explainable Artificial Intelligence (XAI) (Adadi & Berrada)
Presentation
Lecture 7
Topic 7: XAI: LIME
Title : Why should I trust you?" Explaining the predictions of any classifier (Ribeiro, Singh, Guestrin)"
Name of student : SHOJAEE Nooshin
Teacher : Sana Nouzri
Date : November 26,2020
Papers
+"Why should I trust you?" Explaining the predictions of any classifier (Ribeiro, Singh, Guestrin)
Presentation
Why should I trust you?" Explaining the predictions of any classifier
Lecture 8
Topic 8: XAI:Gradient Based DNN Explanation
Title : Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization (Selvaraju1 et al.)
Name of student : HOSSAIN Saddam
Teacher : Sana Nouzri
Date : December 1,2020
Papers
+Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization (Selvaraju1 et al.)
Presentation
Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization
Lecture 9
Topic 9: XAI: Context importance and utility
Title : Explanations of black-box model predictions by contextual importance and utility (Anjomshae, Framling, Najjar)
Name of student : CAUSHI Ada
Teacher : Amro Najjar
Date : December 3,2020
Papers
+Explanations of black-box model predictions by contextual importance and utility (Anjomshae, Framling, Najjar)
Presentation
Explanations of black-box model predictions by contextual importance and utility
Lecture 10
Topic 10: Understandable Robots
Title : Understandable robots-what, why, and how (Hellström, Bensch)
Name of student : JAHIC Alen
Teacher : Sana Nouzri
Date : December 8,2020
Papers
+Understandable robots-what, why, and how (Hellström, Bensch)
Presentation
Lecture 11
Topic 11: Explainable Planning
Title : Plan Explanations as Model Reconciliation: Moving Beyond Explanation as Soliloquy (Chakraborti et al.)
Name of student : SHABANI ABETARE
Teacher : Amro Najjar
Date : December 10,2020
Papers
+Plan Explanations as Model Reconciliation: Moving Beyond Explanation as Soliloquy (Chakraborti et al.)
Presentation
Plan Explanations as Model Reconciliation: Moving Beyond Explanation as Soliloquy
Lecture 12
Topic 12: Explainability in robots: case study
Title : Personalised self-explanation by robots: The role of goals versus beliefs in robot-action explanation for children and adults (Kaptein et al.)
Name of student : ZOHAIRI Fateme
Teacher : Sana Nouzri
Date : December 15,2020
Papers
+Personalised self-explanation by robots: The role of goals versus beliefs in robot-action explanation for children and adults (Kaptein et al.)
Presentation
Wrap up lecture
- Overview, conclusions and discussions of the AI Ethics and XAI.
- Q&A