Celebrating International Women's Day
Published: 24 May 2023
Inspiring us all - Simone Stumpf - Reader in Responsible & Interactive Artificial Intelligence
Growing up in semi-rural Southwest Germany, I was always keenly aware of how I felt different. As a kid, I loved books, I was interested in opera and thought school was great! This was not something expected of the daughter of a welder and a deli assistant, coming from farming stock. I was the first in my family to attend university and, having switched between employment in industry and academia several times, I finally settled on making my career in research and education. You can learn more about my path by watching this short video on youtube.
Diversity and inclusion is really important to me personally and in terms of my work. As a cis-woman who is part of the LGBTQ+ community, I have first-hand experience of how necessary it is to break down barriers between people and the strength we can gain together rather than apart. I have worked in making technology gender-inclusive – whatever gender someone identifies with – and accessible to people with specific needs. I have worked with people who have no or low vision, who are living with HIV, Parkinson’s Disease or dementia. All this is based on my inherent curiosity about people, what makes them tick and what they want and need to live joyful and fulfilling lives.
My recent work is focused on Responsible AI. I am particularly interested in how we can involve everyone in developing AI, and one of my projects, ORBIT, has investigated how people who are blind can teach an object recogniser to find their personal things, like their keys, wallet or backpack. This involved them taking short videos of objects that they wanted the AI to recognise, and that is obviously not easy for them to do. However, we managed to collect nearly 4000 videos which can be used by researchers to develop better AI systems.
My other research concern currently is fairness of AI systems. How can we prevent AI that makes biased decisions? For example, how can we ensure that a ‘HR AI’ doesn’t throw away job applications for software engineers just because they are from women? How do we know if the health AI we are developing systematically discriminates against black people? These are really difficult issues to tackle which means raising awareness about AI in general but also providing tools to be able to check and steer the AI system, ideally by people who are affected by the decisions that these AI system make. You can read more about the efforts to make AI more inclusive for people with disabilities in this article.
We definitely need more people working in this area, and this space needs to be diverse and inclusive, to have any hope of making a difference and ensuring that as a society we benefit from AI.
First published: 24 May 2023
<< Spotlight