Dataharvest 2024 - the European Investigative Journalism Conference

Have you ever found yourself wondering: what if my documents could talk back to me? The ability to extract, interpret, and utilize information from documents efficiently is crucial, not only for journalists but also for their audience. In this session you will learn the basics of Large Language Models(LLMs), how they work and how you can build a tool with them for your newsroom (or maybe even for your audience), as well as practical and ethical questions. Bring along a set of documents (preferably in .txt format), along with your creativity, and you will leave this session with a template to be used whenever you want to use AI to "chat" with a set of documents. Participants will leave with an understanding of basic concepts behind LLMs, how they work, and what exactly is happening behind the screen when you use a tool like ChatGPT. Participants should not be strangers to coding. Some web development and R knowledge is recommended, but not essential. All the necessary software will be pre-installed on the school computers. If you are bringing your own laptop you will need to install Node.js (https://nodejs.org/en/download), a web browser, a code editor of your choice, and  R/Rstudio.