Have you ever started writing the background section of a paper and suddenly thought: “Where did I read that again?” You know the finding, you remember it mattered, but you just can’t place it. This is exactly the kind of problem this workshop was designed to help with, by building a note-taking system that lets you actually find what you’ve read, even years later, and make sense of it in context.
The workshop was run by Nina Mouhammad at this year’s Joint Technology-Enhanced Learning Summer School (JTELSS) in Bulgaria. After a welcome and a short introduction to the topic, we got into one of the more interesting parts of the session: a structured controversy on the question “Do I still need to read papers and make notes on them, if LLMs can summarize and explain them to me?” Participants first took sides assigned to them, then they “dropped the mask” to reveal their real position which sparked some interesting discussion.
This was followed by a content input from Yannic Jäckel, a colleague who researches academic note-taking, joining us via video. Yannic gave a really solid overview of why reading papers and taking notes manually still matters. A few key points that came up: small inaccuracies in AI-generated summaries can propagate through your work and lead to wrong assumptions that quietly make their way into your research; there is an important difference between literal and critical reading; and note-taking itself functions as a kind of firewall, a way to actively engage with claims rather than just passively absorbing them.
After that, we ran a quick poll to get a sense of participants’ backgrounds. Most had used some kind of literature management tool before, with quite a few already on Zotero. Markdown and Obsidian were largely new territory for the room.
From there, we moved into the tools. We walked through how Zotero and Obsidian can work together to support a three-part note-taking system: literature notes (one per paper, with highlights imported from Zotero), point notes (one per insight, with the insight itself as the note title), and topic notes (which pull everything together automatically). We set things up together first, then participants clicked along to create their first notes and explore the system themselves. A short debrief and survey rounded things off.
A big thank you to all participants for engaging so openly, especially during the structured controversy, and to the organizers of JTELSS 2026 for all the effort that went into making this year’s summer school outstanding again.