Jazon Szabo and Mackenzie Jorgensen, both PhD students in the UKRI Centre for Doctoral Training in Safe and Trusted AI (the STAI CDT), talk to us about the AI Ethics reading group that they run.
Tell us a bit about the AI Ethics reading group – how long has it been going, how often do you meet and what are your aims?
(Mackenzie) Jazon and I started the AI Ethics reading group online in spring 2021 because we wanted to facilitate a space where students could have more interdisciplinary conversations about AI and its societal impacts. The group really took off though in autumn 2022 when we started running in person sessions—we also broadened the scope of our group to include students from nearby universities and from other departments, encouraging an interdisciplinary space. Because of the growing interest, we decided to hold sessions every other week. We have paused our in-person sessions until students return to campus in the autumn; but, in the meantime, we started an international online version of the reading group held every other week. This online version has helped us reach students from other cities and countries to join our reading group. This autumn, we will run online and in-person sessions for the reading group to cater for those in London and those abroad.
What are your highlights from the reading group?
(Mackenzie) It has been a great connection point for PhD students to engage and learn from one another in a small group setting.
(Jazon) One of the philosophy students who has come to our reading group liked the topic of AI ethics so much that she did her undergrad thesis within this field. In fact, she is now doing a master’s on the topic!
(Mackenzie) At conferences, I love finding other PhD students who are keen to join the reading group. It is a fantastic way to stay connected with early career researchers around the globe.
(Jazon) I think people are sometimes afraid of presenting but with a bit of encouragement they do it and then they have a good time! I find this part of organising quite fulfilling.
(Mackenzie) We plan to invite some well-known AI Ethicists this upcoming year to speak and are excited about that. As we enter their final year of the PhD, we are focusing on running the online reading group and are passing on the baton to two younger PhD students, Zoe and Michelle. We are really pumped to have more leadership for the group!
Are there any readings you’d like to particularly recommend?
(Mackenzie and Jazon) Yes! Some of our favorite readings from the group are listed here.
- Gwagwa, Kazim, & Hilliard, The role of the African value of Ubuntu in global AI inclusion discourse: A normative ethics perspective: Patterns (cell.com)
Often in machine ethics the focus is on Western philosophy, so we discussed an African perspective on ethics and how it may be included in AI. - Waller & Timmers, Controversial Classifiers — the Perils of Prediction Algorithms in Public Administration
This was our first paper where the author, Paul Waller, came to present his own work. The in-person talk also clarified some interesting questions in more detail than the paper. - Weidinger et al., Ethical and social risks of harm from Language Models
With the recent rise of LLMs, we discussed the AI ethics issues specific to them.