Investigating Documents using AI

July 13, 2024, Abraji Congress, São Paulo

Jonathan Soma, Knight Chair in Data Journalism, Columbia University.

Director of the MS in Data Journalism (1 year Master’s degree) and Lede Program (10 week summer program).

Find me at @dangerscarf or js4571@columbia.edu.

Notes

This only covers the small, small portion of content that I talked about at Abraji! If you’re interested in hearing more in-depth or broader uses of AI in journalism, look at my 12-hour video series Practical AI for Investigative Journalism.

If you’re interested in more programming stuff, take a look at jonathansoma.com and Everything I Know, two places where I put a lot of my content.

Slides

You can find the slides on the slides page

AI in a Spreadsheet

AI in Python

  • Instructor for using Python to extract structured data

RAG (“chatting with your PDFs”)