A principle must be understood by anyone, even without knowing its origin. The principle should clearly state a rule. It must not rely on hidden information from a past conversation.
By Pawan Nayar · from the LLOS AI Learning record
A principle must be understood by anyone, even without knowing its origin.
The principle should clearly state a rule. It must not rely on hidden information from a past conversation.
A principle written during a long talk often uses words like 'it' or 'this'. The writer knows what these words mean. A new reader does not know. When the principle is saved, the original talk is gone. The principle then becomes confusing to others. A good principle is clear on its own. It does not need extra explanation.
A developer writes a principle: 'It must always match the user's input.' A new developer reads this. The new developer asks, 'What is 'it'?' The principle is not clear. The developer should rewrite it as: 'The AI model's response must always match the user's input.'
Resolved from this lesson's own record — real definitions, not invented for this page.
More from the same record, pulled by shared topic — not written new for this page.
Write explanation files like you are teaching someone. Use simple words.
Markdown files help people understand information. A developer uses them for human research and notes.
Markdown files are for human notes and research. A developer uses them to explain ideas or keep records.
When an AI model asks a user to test, open, refresh, check, or verify something, it must include the full web address or file path. Do not use phrases like 'refresh the page' or 'open the file'.
A principle must be understood by anyone, even without knowing its origin. The principle should clearly state a rule. It must not rely on hidden information from a past conversation.