Karpathy: we'll spend far more time understanding LLM outputs — and shares ASD-STE100 writing trick
DavideCrapis · x · 2026-10-02
Andrej Karpathy posted that we'll be spending a lot more time trying to understand the outputs of language models, sharing tips including one concrete technique: ask your LLM to write or explain in ASD-STE100, a controlled language spec originally developed for aerospace maintenance documentation. LLMs are well-versed in it and its heavy constraints produce much cleaner writing.
A replier added that monitoring will become the key human function in engineering and governing these systems.
Related event: Karpathy Shares Four Techniques for Understanding LLM Output(7 posts)→
More from coding & agent
- Devs flag GPT's defensive programming for hiding bugs; OpenAI asks for /feedback cases — pvncher · 2026-11-19
- Flutter ships A2UI 1.0: agents respond with real interactive widgets, not chat text — rseroter · 2026-11-19
- Skills vs MCP vs RAG vs Memory: the four context mechanisms AI agents actually need — MaryamMiradi · 2026-11-19
- Anthropic details agent memory: 'dreaming' lets Claude organize its experience — daniel_mac8 · 2026-11-19
- The case for externalized memory as the real fix for continual learning — daniel_mac8 · 2026-11-19
- FT: US tech outages blamed on rush to fix bugs found by AI — nordicinst · 2026-11-19