Prof. Dr. Aimee van Wynsberghe
Alexander von Humboldt Professor for the Applied Ethics of Artificial Intelligence and Director of the Institute for Science and Ethics at the University of Bonn
»There are many researchers, civil society organisations and industry leaders looking at how we can use AI to achieve the sustainable development goals and I think this is a great initiative, but we have forgotten to look at the sustainability of AI itself.«
Prof. Dr. Aimee van Wynsberghe is the Alexander von Humboldt Professor for the Applied Ethics of Artificial Intelligence and Director of the Institute for Science and Ethics at the University of Bonn. As a member of the European Commission’s High Level Expert Group on AI, she is one of the most prominent AI and robot ethicists. In 2021 Prof van Wynsberghe launched the Bonn Sustainable AI Lab to foster interdisciplinary research concerning environmental justice and AI. We asked her about the opportunities to use AI for the benefit of humanity.
In 2021, you opened the Bonn Sustainable AI Lab. What goals are you pursuing with your Lab?
There are many researchers, civil society organisations and industry leaders looking at how we can use AI to achieve the sustainable development goals and I think this is a great initiative, but we have forgotten to look at the sustainability of AI itself. The development of AI has a large environmental impact from carbon emissions, the mining of precious minerals, water usage to cool servers, electricity consumption, and land usage. In other words there is a massive physical infrastructure needed to develop AI and there are people and the planet being exploited within this infrastructure. I want my lab to do research in this area, to raise awareness, and to encourage people to stop and think before rushing forward with AI.
Your team is very interdisciplinary. Why are different perspectives so important when it comes to Sustainable AI?
The sustainability of AI requires a deeper look into the technical details of the technology, what kind of servers are used, how long are algorithms trained for. But because the environmental costs are often a burden for the most vulnerable and marginalized demographics of the world (e.g., those working in the mines to acquire minerals) it is more than a technical issue, it is a moral issue. I am brining together environmental ethicists with political science and agriculture to address the many complex questions that arise when unpacking the topic of sustainable AI.
What exactly does sustainability mean in the context of AI, or when is an AI system “sustainable”?
I refer to sustainable AI as a kind of umbrella term with two branches. Under this umbrella you have two divergent questions or areas of researcher: 1. how to use AI to achieve sustainability in some way, for example to achieve the sustainable development goals, and 2. the sustainability of AI. To answer the question of ‘when an AI system is sustainable’, we still need to do a lot more research to fully understand the scale of environmental consequences resulting from the development and use of AI. This is one of the goals of the lab, to be able to find an answer to this question.
You also conduct research in the field of robot ethics. How can robots enrich society?
Robots are different from AI; robots are embodied technologies that we interact with in a different way, AI is better understood as software. Because robots have this embodied aspect, and can be designed to appear ‘cute’, there are some researchers that want to make robots in a social way so that we may believe a robot is our friend. I think that is dangerous. Instead, I think we should be focusing on how robots can contribute to the global community and again to help with environmental concerns. For example, there are many people who are exposed to toxic chemicals when they dismantle electronic waste, and these people are often in poorer countries. If we had robots to do this kind of work, I think it would provide a good on a massive scale. But we mustn’t forget that we have to have a plan for recycling any robot we make and/or use. This is another issue that is under studied and requires a lot more attention.
Assuming we could use robots to (co-)care for patients as early as tomorrow to curb the care crisis, how would they have to be programmed so as not to harm humans?
First I think it’s important to really question whether robots are the best solution to the ‘care crisis’. I would think that treating care givers with more respect and adequate pay and support would be a better solution to the problem. But if we were to use robots we also need to study what is it that makes us say we have received ‘good care’ so that we can understand how the robot should be made. For this we need to have nurses and other medical professionals involved in the design process. They are the ones that can really tell us what has to be done to make a robot will not harm a human because they truly understand what care looks like in practice.
You are also a member of the EU Commission’s high-level expert group on artificial intelligence. What does your work look like there and can you incorporate the findings from your research well there?
Being a member of the EU HLEG on AI was an incredible opportunity. It was a multi-disciplinary team with the goal to begin the process of how AI should be regulated in a European context where human rights and European values are the core. One of the lessons that I was reminded of here was that ethics as a process is dynamic, meaning it’s not possible to make a checklist in 2018 and for the checklist to remain as relevant in 2022. I say this because back in 2018 no one was talking about the environmental consequences of AI but now that we know more about the technology, we see how important it is to prioritize this as an ethical concern.
What do you particularly like about NRW as a research location and what can NRW learn from other locations?
I am incredibly honored to come to the University of Bonn. There is a vibrant research environment with many self-driven academics looking to generate high quality research as well as make an impact on real world issues. There is an incredible network of organizations in this area looking into sustainability issues which makes this such an exciting place to be, each day I learn about another researcher that I must get in touch with!
Having come from the Netherlands, working at technical Universities where natural sciences and the humanities worked together quite often, I think this is an important part of learning both at the bachelor/masters level as well as an established researcher. I am pleased to see that many academics at Uni Bonn are also interested in this kind of collaboration so it’s now just a question of making it happen!
Canadian Prof. Dr. Aimee van Wynsberghe studied in Canada and the Netherlands and is an AI and robot ethicist. She is co-founder and co-director of the Foundation for Responsible Robotics and advisor to the European Commission on AI issues. In 2021, she was awarded the Alexander Humboldt Professorship. Since February 2021, she has been working at the Rheinische Friedrich-Wilhelms-Universität in Bonn. A new professorship in Applied Ethics of Artificial Intelligence was designed for her. She is director of the Institute for Science and Ethics (IWE) and director of the Bonn Sustainable AI lab at the University of Bonn. She is author of the book “Healthcare Robots: Ethics, Design, and Implementation”. In 2018, she received the L’Oréal UNESCO For Women In Science Award for her contribution to scientific dialogue and research.”