UN Speeches Research

Deep dives and data analysis uncovering trends, patterns, and insights from over 75 years of General Assembly debates.

Methodology

AI-Assisted Humanities Research

This research was conducted using Claude Code, an AI coding agent, to explore whether such tools can meaningfully enhance research in the humanities.

Conversational Exploration

Instead of writing queries from scratch, research questions are posed in natural language. The AI agent translates these into SQL queries, iterates on results, and surfaces patterns that might otherwise require significant technical expertise to uncover.

Database Enhancement

The agent helped design and populate new database tables (like quotations and notable figures), write extraction scripts with proper Unicode handling, and validate results through sampling and temporal analysis.

Pattern Recognition

By rapidly testing hypotheses ("Who gets quoted more: philosophers or religious figures?"), the agent enables exploratory research that would traditionally require days of manual query-writing and data processing.

Iterative Refinement

The workflow involves constant iteration: finding false positives, adjusting patterns, re-running analysis, and validating against source material. The agent maintains context across sessions, learning from debugging to improve extraction quality.

The Research Question

Can coding agents like Claude Code make humanities research more accessible? Traditional digital humanities work requires expertise in programming, databases, and statistical analysis. AI assistants may lower these barriers, allowing researchers to focus on interpretation rather than implementation.

This project serves as a case study: every research page here was created through conversation with an AI agent, from initial data exploration to final visualization. The agent wrote the SQL queries, Python scripts, and React components—guided by human questions and judgment.