Artificial Moral Agents?


Event Details

Friday February 1, 2019 from 12:00 PM to 5:00 PM
Luna Building - Corona hall
Data Science

About the event

Increasingly, Artificial Intelligence (AI) systems or artificial agents begin to replace human decision makers in real life. For example, software systems currently assess someone’s risk profile and financial situation, and make decisions about mortgages. An algorithm called Compas (Correctional Offender Management Profiling for Alternative Sanctions) is used throughout the US to assess the extent to which defendants awaiting trial or sentencing are at risk of reoffending, and whether they can be released on bail. The use of artificial agents as decision makers has become a significant concern, not only for society and governmental organizations, but also for technological corporations such as Google, Microsoft, Facebook and Amazon. There is a clear backlash if these systems make morally questionable decisions.

The challenge, then, is to build ethics into AI systems. The ultimate goal and ‘man on the moon’ project is the development of an artificial moral agent which is able to make complex moral decisions and act upon them. This requires the ability for moral reasoning and ‘all things considered’ judgements, taking into account relevant knowledge about the social and physical context, relevant knowledge about the humans involved, as well as the moral values and principles relevant for the situation.

The research meet will address some of the many challenges in integrating ethics within AI, and developing artificial moral agents


12.00 - 12.30
Arrival and small lunch

12.30 - 12.40
Anthonie Meijers, university professor, Philosophy & Ethics, TU/e

12.40 - 13.25
Title to be announced
Vincent C Müller, Inter-Disciplinary Ethics Applied Centre, University of Leeds, United Kingdom

13.25 - 14.10
The Ethics of Artificial Ethics Advisors
Elizabeth O'Neill, Philosophy & Ethics, TU/e and Cornell university, USA

14.10 - 14.30
Small break

14.30 - 15.15
Fairness and Transparency of Machine Learning behind AI
Mykola Pechenizkiy, scientific leader Responsible Data Science, TU/e

15.15 - 16.00
Hybrid intelligence for a healthy Human AI Society
Jurriaan van Diggelen, senior research scientist Perceptual and Cognitive Systems, TNO

16.00 - 16.30
Debate chaired by Anthonie Meijers

16.30 - 17.30
Networking and drinks
This event is organized by the TU/e Center for Humans and Technology, in collaboration with the Data Science Center Eindhoven (DSCE) and the High Tech Systems Center (HTSC).


High Tech Systems

In next-generation high-tech and mechatronic systems, extreme functionalities and performance requirements will be realized by using a model-based multi-physics systems approach. New sensing technologies and actuator designs for multi-physics processes (forces, flows, temperatures, acoustics, optics) will need to be integrated with and used by distributed on-line model-based control and optimization tools. The control systems are adaptive, auto-tuned, are implemented in optimized hardware and software architectures, and use effective (wireless) communication.