The Aesthetic Cognitivism project aims to provide a robust theoretical and empirical demonstration of the strong view of aesthetic cognitivism—that aesthetic experiences yield essential knowledge and understanding. We will employ Large Language Models (LLMs) and natural language processing methods to conduct an in-depth, historical analysis of art criticism. This analysis will track the evolution of critical concepts used to articulate the cognitive and moral value of art, establishing how art criticism has informed and transformed the nature of artistic experience. The aim is to use LLMs for metadata extraction from a huge volume of text necessitates access to high-performance computing resources. Goal: To analyze large datasets of art criticism to identify, extract, and track the evolution of key critical concepts (e.g., knowledge, understanding, moral value). Method: Leverage fine-tuned open-source LLMs (e.g., Llama, mistral, or similar) suitable for text classification and sophisticated metadata extraction tasks. Requirement: This process involves both finetuning and extensive inference runs on large text corpora, requiring access to high-memory GPUs to handle large models and batch sizes efficiently..