Commentarii Roamani

In this issue we'll do a wrap-up of the main ideas and systems we explored in Commentarii Roamani in 2025.

If you're a new subscriber, read on to see where to go for content you've missed! And if you've been around for the whole journey, revisit the issues that could help you turbocharge your entry into 2026! Either way, think of bookmarking and sharing this issue as a useful reference tool (we linked all the issues to make it handy)

The newsletter issues we've sent out fall into three categories: building personal thinking systems, learning and knowledge construction, and AI as a thinking partner.

Building personal thinking systems

A large part of this year focused on building workflows and systems that improve our daily life, in Roam.

• ​[[Experiments]] is a page where you log new things you are trying with deliberation, along with the changes you notice over time, and it's meant to document what works and what doesn't to improve your day-to-day life. Over time, this works to amass empirical evidence for your choices and iterate moving forward on what habits to keep.

• ​[[Via Negativa]] is a page where you write down habits, projects, or goals you have decided to stop pursuing. Writing these decisions down helps clear mental space, because once something is explicitly abandoned, it stops resurfacing.

• ​The [[Daily 5]] is an organized daily workflow in Roam built around a fixed template with five sections: review, plan, log, journal, and catch. Each day, you open the same structure in your Daily Notes and fill it in as the day unfolds, removing the inertia from starting each day, but also offering a predictable structure to follow.

• Team Workflows (pt1 and pt2) is a system that focuses on using Roam as a shared space. The issue covers setting up a graph for collaboration and adding lightweight structures for daily check-ins, weekly progress, meetings, and projects, so teams can share context and track work over time without forcing everyone into the same workflow.
  ​
• ​Algorithms of Thought is an extension that structures reasoning itself. It provides explicit thinking patterns for breaking down problems, comparing options, and clarifying decisions. Instead of keeping reasoning implicit or scattered, it leaves a clear trace of how conclusions are reached, which makes decisions easier to revisit and revise over time.

• ​Better Tasks is a task system in Roam where each task is a block with its own dates and repeat rules. Instead of rewriting the same tasks or relying on memory, tasks automatically resurface when they are due or available.

• ​[[Focused Goals]] is a planning workflow built around a fixed time window and a small number of priorities. We explored this workflow during the Winter Arc, but it can be reused for any focused stretch of work.

Learning and knowledge construction

This was another recurring theme: we posited that learning takes shape over time, from initial setup to reading, studying, and reuse.

In Back to School (pt1 and pt2), we looked at how to organize courses, schedules, syllabi, lecture notes, and exams in Roam.

We also created the [[Learnloop]] workflow to look at how notes can be created and revisited regularly through short reviews and spaced repetition. This way, understanding builds over time instead of accumulating only around deadlines. This was our first peek at using AI from within Roam, which is covered in the next section.

Finally, we look at reading as part of the same process. In Roam Reader, we show how reading can happen alongside your notes, with highlights and comments captured in context and saved directly into the graph, making what you read easier to reuse later for writing, synthesis, or further study.

AI as a thinking partner

This section focuses on how AI is used inside Roam as a support layer for thinking, writing, and exploration.

The Live AI issues (pt0, pt1, pt2) cover how AI operates directly on blocks, pages, and the graph itself, and because AI works on selected context or the graph, thinking stays anchored to your own material rather than starting from scratch.

We also looked at how AI changes search and discovery inside large graphs. With tools like Search+ and Roam Portal, search is not limited to finding a specific block. You can filter by time, page, or structure, save searches, and step back to see how ideas develop and connect across months or years. This makes search part of the building process.

Thank you all for this wonderful year!!! If you have ideas for features, suggestions, or need any help, send us a message: support@roamresearch.com.

♥️