Your Notes app is solving the same problem as a Roman wax tablet. The problem hasn't changed. The tools have changed dramatically. And each generation of tools solves one layer of the problem, exposes the next layer, and hands it to the next generation to address.
The wax tablet and the problem of impermanence
Roman scholars, merchants, and schoolchildren used the tabula cerata — a wooden frame filled with wax, written on with a stylus, erasable and reusable. Caesar is described carrying one into battle to dictate correspondence. Cicero referenced his writing tablets repeatedly. The medium was ubiquitous in the ancient world for exactly the same reason a phone note is ubiquitous now: it was the fastest way to capture something before you forgot it.
The problem with the wax tablet was the same problem with everything temporary: it didn't last. Important things got transcribed to papyrus or vellum. Unimportant things got erased. The distinction between "important enough to preserve" and "just a passing thought" had to be made at the point of capture, which is the worst possible time to make it — you don't yet know which thoughts will matter.
This is the first layer of the note-taking problem: the gap between the moment of capture and the moment of knowing whether something was worth capturing.
Commonplace books and the problem of retrieval
By the Renaissance, educated Europeans had developed the commonplace book — a personal compilation of passages, quotations, observations, and ideas gathered from reading and experience. The practice was taught in schools. Erasmus wrote instructions for keeping one. Francis Bacon kept one. John Locke not only kept one but invented a novel indexing method for it in 1685, published as "A New Method of Making Common-Place-Books," which used a two-variable index (first letter of a word + its first vowel) to reduce a large book to a manageable lookup table.
Locke's indexing method was widely adopted and reprinted for over a century. It was, essentially, an algorithm. He had identified that the problem with commonplace books wasn't filling them — it was finding things in them. The retrieval problem. You can capture everything. You cannot find it again without a system.
The gap Locke identified in 1685 — between capture and retrieval — is the gap that Google, search engines, and every notes app full-text search feature has been trying to solve ever since. The 2024 version is Obsidian's Dataview plugin. The 1685 version was a two-letter index in a leather-bound book. The problem is the same.
Vannevar Bush and the problem of association
In July 1945, the director of the U.S. Office of Scientific Research and Development, Vannevar Bush, published an essay in The Atlantic called "As We May Think." It is one of the most prescient documents in the history of computing, written before most of what it describes was technologically possible.
Bush's central insight was that the library — the dominant knowledge organization system of 1945 — was organized by hierarchy: subject, then subfield, then topic, then subtopic. This was physically convenient for shelving books. It was cognitively wrong for how humans actually think. "The human mind," Bush wrote, "operates by association. With one item in its grasp, it snaps instantly to the next that is suggested by the association of thoughts, in accordance with some intricate web of trails carried by the cells of the brain."
He proposed the Memex: a desk containing microfilm-stored documents, with a mechanism for the user to create "trails" — associative links between documents that could be followed and reshared. The Memex would remember not just the documents but the connections you made between them. It was never built, but it directly inspired Douglas Engelbart, Ted Nelson, and Tim Berners-Lee.
The problem Bush identified — hierarchical storage fails to capture associative thought — is the problem that bi-directional linking in Roam, Obsidian, and Logseq is attempting to solve today. Eighty years later, the same problem. Progress is real but slow.
Ted Nelson and Xanadu: the system that never shipped
Ted Nelson coined the words "hypertext" and "hypermedia" in 1963. His Project Xanadu, begun in the 1960s, proposed a global network of linked documents where every piece of text could link to its source, every link was bi-directional, and nothing was ever deleted — only versioned. Nelson wanted transclusion: the ability to include a piece of one document inside another while preserving the link to the origin, so you always knew where a quote came from.
Xanadu was never shipped. Tim Berners-Lee's World Wide Web was — in 1991, as a simpler, unidirectional, deletion-permitted version of what Nelson proposed. The web became the dominant hypertext system not because it was better but because it shipped. Nelson has spent decades pointing out what the web got wrong. He is largely correct about the technical deficiencies. The web exists anyway.
The lesson of Xanadu is repeated throughout the history of knowledge tools: the more complete the solution, the harder it is to ship. Every generation's ambitious knowledge system — Xanadu, the Semantic Web, Google Wave, Notion's relational databases — runs into the gap between what the system can theoretically represent and what users will actually maintain. Most people can't consistently annotate every note with structured metadata. The perfect knowledge graph requires a level of discipline that real humans don't sustain.
Evernote, Notion, Obsidian: the modern cycle
Evernote (2008) solved capture at scale — clip anything from anywhere, search everything, access on any device. 200 million users discovered that once you can capture everything, you have a new problem: you have too much captured, not enough organized, and none of it connected to action.
Notion (2016) solved organization — relational databases, linked pages, structured properties. Users discovered that the organizational system required more maintenance than the notes justified, and that organized notes disconnected from tasks were still disconnected from action.
Obsidian (2020) solved linking — bi-directional links, graph view, networked thought. Users discovered that linking notes to each other was not the same as linking notes to action.
Each tool solved the problem it found at the frontier of the previous tool's solution. Each solution exposed the next unsolved layer. The layer that keeps appearing — in every generation, in every tool — is the gap between captured knowledge and executed action. Bush saw it in 1945. It is still there in 2024.
The gap that no tool has closed
The history of note-taking technology is, at its core, the history of solving one problem at a time: impermanence, retrieval, association, organization, linking. Each generation of tools makes genuine progress on the layer it addresses. And each time, the next layer appears underneath it.
The layer that has not been solved by any dominant tool is the gap between the note and the task — between the thing you know and the thing you're going to do about it. Notes and tasks have lived in separate systems for the entire history of personal information management. The wax tablet was for capture. The scroll was for reference. The calendar was for scheduling. The action and the knowledge lived in different places.
Every personal productivity crisis — the overwhelmed inbox, the abandoned Notion workspace, the Zettelkasten that never produced output — is at least partially a symptom of this gap. You know things and you don't act on them. You act on things and you don't have the context. The knowledge and the action are in different containers, and the cost of bridging them is paid every time you pick up either one.
This is the problem that hasn't been solved yet. It's 2,000 years old. It will probably take more than one more app to close it. But closing it is the direction we're pointed.