Investigating with AI

The course bridges the gap between traditional social science research and contemporary AI capabilities, empowering students to gather, transform, and analyze textual data at scale - turning them into active investigators who can extract meaningful patterns from the digital noise that surrounds us.

Day 1: Welcome, intro, course scope, logistics; state of LLMs; hands on : data analysis on google colab

Day 2: LLMs, JSON and pandas

Day 3: APIs

How the Web works; APIs; The Wikipedia API; The NYT API, Duckduckgo API

Day 4: NLP with Spacy: NER, POS, topic modeling, text classification, tokens, stop words, lemmatization, …; NYT API; more python;

Day 5: embeddings; LLMs; RAG; Spacy.io;

Day 6: deploy streamlit app on github

Day 7: from LLMs to Agents

Day 8: Data visualization and inference

day 9 networks, scraping

day 10 machine learning, deep learning, reinforcement learning

day 11 buffer

day 12 projects presentation