Saving Artificial Minds in 2027: Some Recent Trends in AI Welfare Research – Leonard Dung

​Saving Artificial Minds in 2027: Some Recent Trends in AI Welfare Research

​Presentation by Leonard Dung, Postdoctoral Researcher at the Institute of Philosophy II, Ruhr-University Bochum, and author of Saving Artificial Minds: Understanding and Preventing AI Suffering (Routledge, 2025)

​This talk will, first, present key themes of my book Saving Artificial Minds (published in 2025) and then survey key developments in AI welfare research since then. I will also present some opinionated views on these topics. Topics are:

• ​The emergence of concrete behavioral and mechanistic empirical research programs on AI welfare

• ​Debates on AI identity

• ​The discovery of persona vectors and the relevance of personas to AI welfare

• ​Questions of legal rights on AI

​The talk and the Q&A are recorded. The video goes up publicly afterwards, with an edited transcript in which attendees are not named. If you’d rather not appear in it, put your question in the Q&A panel and I’ll read it out for you.

​--

​AI Welfare Seminars is a series of research presentations and discussion on AI welfare, consciousness and moral status.