A 3D galaxy slowly rotates on the screen, with each light dot representing a hidden connection between Fischer’s 1972 chess game against Spassky and today’s trading strategies. This isn’t a sci-fi scene—it’s Andrej Karpathy’s daily work interface: a self-thinking knowledge galaxy built from scratch.

While ordinary people use Notion as a filing cabinet and Roam as a spreadsheet, the former Tesla AI director implemented the topological structure of cognition with code:

- Click any node, and the AI automatically reveals all hidden connections

- Zoom out to see the entire thinking universe

- Zoom in to discover the causal chain between the 1972 chess match and current trading strategies

ChatGPT forgets conversations in two hours, but this system permanently remembers all connections. After a developer replicated this architecture, they got a dynamic network of 378 notes, 1854 nodes, and 3856 edges. The gap between knowledge consumers and knowledge compounders has never been talent—it’s the underlying architecture.


In another widely cited guide, developer CyrilXBT revealed the operating logic of a similar system: using Claude AI as neural synapses to turn Obsidian from an archive graveyard into a growing organism. The key breakthrough lies in a four-layer architecture:

1. Zero-friction capture layer (Readwise/Telegram bot)

2. Automated pipeline layer (N8N workflow)

3. Permanent storage layer (local Markdown files)

4. Intelligent connection layer (Claude daily briefing)

While most people are organizing folders, true cutting-edge thinkers are building cognitive feedback loops. In six months, such systems will evolve disturbing capabilities—it can point out the internal connection between an idea you jotted down three weeks ago and last night’s dream, when you didn’t even realize you were thinking about this problem.

As one user put it: People who use Notion are building libraries, those who use Roam are compiling dictionaries, and Karpathy is building galaxies. This may explain why the top AI researchers are obsessed with rebuilding their own thinking infrastructure—in the age of intelligence, cognitive architecture itself is a competitive advantage.