Project launch: Ethical Dilemmas In AI – Engineering The Way Out

This is project is a small collaboration with the Association of Nordic Engineers, the IEEE and the Data Ethics ThinkDoTank. Our goal is to organize an online Hackathon event on September 3rd and 4th for Nordic engineers with the outcomes recorded in a published report. Please contact lvr@ida.dk if you are a working engineer and are interested in participating.

We address the problem that despite all of the discussions about ethics and AI and a proliferation of guidelines, checklists and recommendations, we still know very little about how engineers go about dealing with the challenges they encounter in practice. This is not a new problem, my project VIRT-EU has explored this issue for three years and we produced a range of tools to support technology developers. Really though, we just scratched the surface of this very thorny problem.

The hackathon will explore the gaps between existing guidelines and recommendations for addressing ethical issues with respect to AI technologies and the realities experienced by engineers in practice.
The event will address following questions :

  • What kinds of ethical challenges do engineers encounter in practice as they work with AI technologies?
  • How are these ethical challenges understood and addressed?
  • How do we ensure that society can derive maximum benefit from AI technologies, while limiting the potential for negative consequences?

We are hoping that by bringing engineers working in a range of different areas together to talk about the challenges they experience and to discuss the practical solutions they employ, we can begin to develop a foundation for turning the high-level guidelines on ethics and AI into practical recommendations and specifications. What might engineers find useful as advice, recommendation or a best practice example? What tools are missing and need to be developed to help? Answers to these and other questions will hopefully form a basis for advocacy to ensure that ethical issues are appropriately taken into account in the development of AI technologies in the future.