Karpathy shares tips for reading LLM outputs, using the ASD-STE100 controlled language spec
daniel_mac8 · x · 2026-10-02
Karpathy posted a thread of tips for understanding LLM outputs, noting we'll spend far more time reading model-generated text. One trick he endorses: ask an LLM to write in ASD-STE100, a controlled language spec originally built for aerospace maintenance docs — heavily constrained, clean prose, and well known to models. danielmac8 adds that it's like reversing the harness so the agent interfaces with us instead of the computer.
Related event: Karpathy shares techniques for understanding LLM outputs(3 posts)→
More from coding & agent
- RAG fixes worked on test questions but not held-out ones: overfitting on 790 clinical PDFs? — Overall_Judge8086 · 2026-10-02
- Blogger weighs returning to Claude's $200 plan as Codex infra stumbles — StewartalsopIII · 2026-10-02
- Solo dev builds C&C-style RTS "Superintelligence" entirely with Claude Code — Icy_Distribution_361 · 2026-10-02
- Dev declares traditional LLM chats dead: modern harnesses like Cursor Projects are irreversible — js_craft_hq · 2026-10-02
- Row-Bot 5.0 rebuilds local-first AI assistant with persistent goals and per-conversation workspaces — Acceptable-Object390 · 2026-10-02
- Parallel C++ Textbook Shows C++17/20 Standards Can Replace OpenMP and MPI — TinfoilTricorn · 2026-10-02