Tools for thought

The umbrella term for software meant to augment thinking through networked, atomic, interlinked notes rather than filed documents. Created during a lint pass: the phrase recurred across many pages without a hub of its own.

The lineage (this wiki's cluster A)

• Intellectual roots: vannevar-bush's memex (associative-trails), douglas-engelbart's augmentation/NLS hypertext, and niklas-luhmann's zettelkasten — three independent arrivals at connection over hierarchy.
• Digital tools: tiddlywiki5 (~2004, self-contained file) → roam-research (2019, bidirectional links + graph) → obsidian / logseq (local-first plain markdown) → notion (cloud/structured-DB counterpoint) → tana (typed supertags + AI-native).
• LLM era: llm-wiki, gbrain, llm-wiki-agent — the LLM doing the linking, synthesis, and maintenance.

Shared traits & the through-line

Atomic units, liberal linking, emergent (not hierarchical) structure, a personal knowledge-graph, and — increasingly — typed notes (Tana supertags, GBrain schema packs, this wiki's schema.org @type) and AI over your own structure. The arc ends where this wiki's thesis lives: the tools-for-thought movement and the LLM-as-knowledge-partner idea are the same story (see synthesis).

The branch that runs the other way

Every tool above stores knowledge outside the head. spaced-repetition is the half of the lineage that puts it in — SuperMemo and Anki schedule reviews at expanding intervals so a fact is recalled rather than looked up. It belongs here on Engelbart's own terms (software augmenting a cognitive faculty), and the two halves are usually practised together: notes worth keeping become cards worth retaining.

The tradition beside it

This lineage is one of two. information-science — the discipline formed around the documentation-movement, information-retrieval and knowledge organization — has worked the same problem since 1895, from the institutional side: shared catalogues, designed classifications, trained staff. Tools for thought works from the personal side, with association in place of classification and the owner in place of a librarian. The two literatures rarely cite each other, which is why this wiki reached the llm-wiki pattern without meeting either Otlet or the word retrieval as a field information-science-wikipedia.

Related

zettelkasten · spaced-repetition · roam-research · tiddlywiki5 · obsidian · logseq · notion · tana · memex · llm-wiki · knowledge-graph · information-science · documentation-movement