Hybrid Workshop on the Philosophy of Large Language Models

Image created with DALL·E by OpenAI (prompt: philosophers at a conference on large language models, digital art)

Jan 26, 2023 — Jan 27, 2023
Eindhoven University of Technology and online


Hybrid workshop on the philosophy of large language models (with a focus on communication and explainable AI). Hosted and sponsored by the Eindhoven Center for the Philosophy of Artificial Intelligence (@ephil.ai), the research consortium Ethics of Socially Disruptive Technologies (ESDIT), and the Philosophy and Ethics Group at Eindhoven University of Technology (@PhilEthicsTuE). The workshop is organized by Guido Löhr and Céline Budding.

In this workshop, we explore two aspects of the nature of large language models like LaMDA, GPT3 or PaLM. First, we will explore the question of what it can mean for these LLMs to make speech acts or to communicate with us. For example, what could it mean for a chatbot to make an assertion or make a demand? Second, we will discuss LLMs in relation to explainable AI. How can we make sense of the transformer models that make modern LLMs so successful?

The workshop is a hybrid event and open to anyone who is interested. If you are interested in attending online or in person, please email Guido Löhr and we will send you additional information.


Time is in Central European Time (CET).

Day 1: 26th January 2023

Room: Atlas 6.210

  • 13:00: Welcome (tea/coffee/sweets)
  • 13:15: Introduction (Löhr, Budding)
  • 13:30–15:15: LLMs and Explainable AI I
  • 15:15–15:30: Coffee Break
  • 15:30–18:00: Chatbots and Communication I
  • 19:00: Dinner at Gezana (map)

Day 2: 27th January 2023

Room: Atlas 1.652

  • 10:00–12:30: LLMs and Explainable AI II
    • 10:00–10:45: Atoosa Kasirzadeh (online) — Conversation with AI: aligning language models with human values
    • 10:45–11:00: Coffee Break
    • 11:00–11:45: Guido LöhrResponsibility gaps and proxy agents
    • 11:45–12:30: Céline Budding — Tacit knowledge and how-explanations of LLMs
  • 12:30–13:30: Lunch
  • 13:30–16:15: Chatbots and Communications II
    • 13:30–14:30: Keynote Herman CappelenExplainable and interpretable AI: perspectives from externalism & conceptual engineering
    • 14:30–15:15: Anna Strasser (online) — A new kind of comprehension in large language models?
    • 15:15–15:30: Coffee Break
    • 15:30–16:15: Kristian González Barman (online) — The understanding test
  • Conclusion/Drinks
Guido Löhr
Guido Löhr
Social Media Coordinator
Area Lead : Social Robots and Embodied AI
Céline Budding
Céline Budding
PhD Candidate