Extract vs Load: comparing the same knowledge graph built two ways, edge by edge
adnan_hashmi · x · 2026-10-30
A side-by-side comparison of building the same knowledge graph two ways: Load syncs from systems of record with zero model calls (227 nodes, 487 edges, all known-true), while Extract has an LLM read 105 pages against the ontology—and introduces invented nodes, wrong values, duplicates (one thing, two names), missing edges, and dangling references, in quantities that vary with the model and can't be distinguished from inside. The takeaway: load structured facts from authoritative sources rather than LLM-extracting them.
More from coding & agent
- llmman: run Claude Code, Codex and other agents on any local model with one command — weskambale · 2026-10-30
- Context Engineering with DSPy: a practical guide to programming—not hand-crafting—LLM context — kieranklaassen · 2026-10-30
- Former games founder protypes an elemental fighting game with 'vibe UI' built on Claude and GPT — trq212 · 2026-10-30
- Proof of Agent rolls out 15-minute checkpoints anchored to Solana and Base, capping validation costs — adamamcbride · 2026-10-30
- Dev updates his site's showcase by @-replying a bot — "build in public" redefined — jonathan_wilke · 2026-10-30
- ~10x speedup for leafcutter as maintainers turn to AI to validate unreviewably large PRs — jmschreiber91 · 2026-10-30