RAI group members Dr Koorosh Aslansefat (2nd from left) and Kuniko Paxton (3rd from left) at the Pint of Science event in Hull, UK.
The tantalising and critically important topic for Pint of Science in Hull this year was "Trustworthy AI: Can We Trust AI for Skin Cancer? Are They Fair?" Kuniko Paxton presented their breakthrough research on fair machine learning along with Dr Koorosh to an eager and mixed audience!
What Is Pint of Science?
Pint of science banner displaying links to its social media platforms.
Pint of Science is an annual science festival that takes place simultaneously for three days in more than 500 cities in 25 countries. Researchers meet the public in local pubs and cafes to talk about their latest discoveries in an informal way. The event is meant to make a culture whereby people can engage with science in a relaxed and fun-like manner. It believes in making science accessible and interesting to the public.
Talk from Responsible AI Research Group
Kuniko Paxton delivering her talk at the Pint of Science event.
Kuniko Paxton and Dr Koorosh Aslansefat from the Responsible AI Research Group presented Trustworthy AI, with a particular emphasis on ensuring fairness and reliability in AI-driven systems designed to detect skin cancer. Kuniko raises current serious concerns about the fairness of AI algorithms, especially in critical domains like medicine. She thus called for increased scrutiny on fairness issues as used in solutions based on AI, healthcare being cited, and the level to which we can trust the AI systems. She cited the basic tenets as provided in the Responsible AI framework, one of them being fairness.
Audience and Their Questions During the Event
RAI group member, Dr Julius Mboli, making his contribution at the Pint of Science event.
The audience at the Pint of Science event consisted of AI researchers, developers, and healthcare professionals, along with a number of people from the general audience who had a good interest in science and related technologies. Their interest was palpable, and they asked many questions. They seemed to ask more on the ways through which bias in AI systems is detected and effectively mitigated, challenges that exist with skin colour in AI algorithms, and how such bias would play in real applications, such as skin cancer detection.
What Was the Impact?
Dr Koorosh Aslansefat delivering his talk at the Pint of Science event.
Kuniko made the presentation along with Dr Koorosh, bringing out a significant impression that the social value they created was out of fairness in AI, particularly in sensitive fields like healthcare. It showed current, forward-looking team research with innovative methodologies to achieve fairness in AI systems. This discussion added to the understanding of issues related to fair and non-biased AI, pushing the critical thinking of technologies with which people come into contact.
Conclusion
Finally, this event, organised in Hull under the flagship of Pint of Science, assured a clearer and more engaging reach to broader audiences for complex scientific issues. Kuniko and Dr Koorosh's talks about fairness in AI and its application in skin cancer detection were highly informative and motivating. Hence, this event shows that continuous research and dialogues ensure that AI technologies can be trusted and fair in their operations, hence creating an equal and reliable future in technology.
Get in contact with the Pint of Science speakers!
Kuniko Paxton
Web: web URL
Email: k.azuma-2021@hull.ac.uk
Dr Koorosh Aslansefat
Web: web URL
Email: K.Aslansefat@hull.ac.uk
Event Details
Date: 13th May 2024
Comments