Normative Conflicts and Shallow AI Alignment

Raphaël Millière

The progress of AI systems such as large language models (LLMs) raises increasingly pressing concerns about their safe deployment. This paper examines the value alignment problem f…