Fraunhofer IAIS NetMedia Lecture

I was honored to be invited to give a lecture at Fraunhofer IAIS NetMedia, where I had the opportunity to present my work titled Minds and Machines: Towards Aligned Human-AI Interaction.

Lecture’s Abstract

In light of the evolving paradigm towards human-centric AI, this presentation covers the challenges of effective human-machine communication. Historically, AI research emphasized surpassing human capabilities. The focus has gradually shifted towards utilizing AI to facilitate daily human activities. The prevalent issue, however, resides in the alignment problem: AI systems often lack the understanding and contextual awareness inherent in human communication. My research, structured in three steps: explores emergent communication in artificial settings, fine-tunes language models to align with human values using reinforcement learning, and addresses misalignments by applying the cognitive ‘Theory of Mind’ to artificial agents.