The Obsidian Generation

In 2022, Obsidian had roughly 500,000 users. By 2025 it had crossed a million. Notion, which covers a broader use case, has over 30 million. Roam Research, the more esoteric option, has a smaller but intensely committed following. Logseq, Capacities, Bear, Apple Notes with tags, a dozen more.

Something happened in the early 2020s. A significant cohort of people decided their thinking was worth systematically recording.

What they built

The PKM (personal knowledge management) community developed sophisticated systems for capturing, connecting, and retrieving ideas. The Zettelkasten method, popularized through Niklas Luhmann's legendary 90,000-card system, was adapted for software. Tiago Forte's "Building a Second Brain" — PARA method, progressive summarization, projects-areas-resources — sold hundreds of thousands of copies and spawned an industry of courses.

These systems work. People who use them seriously report better thinking, better output, better memory for what they've read. The notes-as-thinking-infrastructure thesis has been validated.

But the systems were almost entirely individual. You build your second brain for yourself. Your vault is private. Your graph is yours alone.

The invisible graph problem

The result is a strange situation: hundreds of thousands of people have created the most detailed external representations of their thinking that humans have ever built — more detailed than any profile page, more honest than any bio, more revealing than most published writing — and almost none of it is visible to anyone else.

Your Obsidian vault, if you have been using it seriously for two years, contains more information about how you actually think than every social media profile you have ever made combined. It has your real obsessions, not your performed ones. It has your actual questions, not the ones you want to be seen asking. It has your working frameworks, your intellectual influences, your recurring confusions.

And it is completely invisible.

Why people kept their vaults private

The privacy was natural, even rational. Notes are messy. They contain half-formed ideas, embarrassing confusions, contradictions you haven't resolved, opinions you haven't stress-tested. Sharing them publicly feels like sharing a draft marked "do not distribute."

There is also no infrastructure for selective legibility. You cannot easily say "show someone the pattern in my thinking without showing them the specific notes." The graph is either open or closed.

So people kept it closed, and the mapping project that the PKM community was collectively running — the project of figuring out how individual minds actually work — remained entirely siloed.

What changes with extraction

AI changes the equation because it makes extraction possible. You don't have to share your raw notes to make your thinking legible. You can run your vault through a model that extracts the patterns — what keeps recurring, what you're wrestling with, what your frameworks are, who shaped your thinking — and represent that as a structured fingerprint.

The fingerprint is derived from your notes but is not your notes. It's an abstraction: legible enough to be compared with someone else's fingerprint, private enough that you haven't handed over your raw thinking.

This is the infrastructure the PKM generation has been missing. Not a way to share your vault — a way to share the shape of your mind.

What happens when the maps talk to each other

The Obsidian generation has been building maps of individual minds. The question is what happens when those maps start to connect.

The most obvious implication is the one Rodin is built around: matching. If you can compare the fingerprints of two people and measure how similarly their minds work, you can surface connections that no credential-based system would ever find. The philosopher in Seoul whose thinking maps onto yours, even though you work in different fields and have never heard of each other. The researcher in Berlin who has been stuck on the same question as you for three years.

But there is a longer-horizon implication. If enough people externalize their intellectual fingerprints, you would have, for the first time, a detailed map of how ideas actually distribute across minds at scale. Not survey data on beliefs. Not citation networks in academia. A real map of which frameworks travel together, which questions cluster, which intellectual lineages are still alive in people who have never read their supposed originators.

That would be a genuinely new kind of knowledge. The Obsidian generation built the raw material. The question is whether anyone builds the infrastructure to make it collective.