

LLM Personas As Potential Moral Patients – Patrick Butlin (Eleos AI)
🧠 Sentio
Sentio is building the digital minds field in London & beyond. We host the Sentio Series - fortnightly London events that combine a talk, Q&A, and networking. Talks explore empirical, governance and moral questions in areas such as AI consciousness, sentience, agency, and welfare. A home for those working on or interested in digital minds to gather regularly.
🎙️ Overview
LLM personas, such as the 'assistant' persona, are characters that appear repeatedly in model outputs and are supported by dedicated mechanisms. They are notable candidates for moral patienthood because they have coherent values and other psychological traits.
In some respects, it makes more sense to think about the interests of Claude, the helpful AI assistant, than those of Claude Opus 5.5, the large language model. In this talk, Patrick will discuss some open questions that bear on whether personas could be moral patients.
Followed by Q&A and (in-person) networking, with pizza (vegan) and drinks included. Whether you work in digital minds, are considering it, or are curious from a related field like AI safety, come along!
👤 Speaker
Patrick Butlin is a philosopher of mind and Senior Research Lead at Eleos AI Research, a non-profit research organisation studying the potential for consciousness and moral standing in AI systems. Patrick has contributed to influential publications on AI consciousness and welfare, including as a lead author on the 2023 report ‘Consciousness in Artificial Intelligence: Insights from the Science of Consciousness’. He previously worked at the Future of Humanity Institute and Global Priorities Institute at the University of Oxford.
🧍 Attend In-Person
6:30pm – Doors Open, Pizza (Vegan), Drinks
7pm – Talk
7:45pm – Q&A
8:15pm – Networking & Drinks
10pm – Wrap-up
💻 Attend Online (GMT)
7pm – Talk
7:45pm – Q&A
8:15pm – Stream Ends
✉️ Contact
Message Joe at team@sentioai.org