Partially Verified65% confidenceFactTime unknown
当前自然语言提示的最大问题在于歧义性，同样的需求不同人写出的提示词可能完全不同，AI产出的代码也可能截然不同
3
Sources
65%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
Unverified自然语言驱动的测试存在确定性问题，AI 对同一描述的理解可能存在波动，影响结果可重复性77% similarUnverifiedThe semantic structure of code differs fundamentally from natural language, making it difficult for embedding models to accurately preserve information such as identifier naming, call relationships, and scope.75% similarUnverifiedText generated by large language models tends to have low perplexity, with word choices that are overly consistent with probabilistic expectations, lacking the unexpectedness of human writing.75% similarVerified提示注入的根本原因在于大语言模型缺乏区分指令层与数据层的原生机制74% similarUnverifiedThe preference expressions and invented vocabulary of large AI models may reversely permeate and influence the natural evolution of human language.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/16939API
curl https://kongchang.com/api/v1/knowledge/claims/16939MCP
get_claim(id=16939)