Unverified50% confidenceFactExact time
The dominant failure type in AI medical records is omission, not transcription errors.
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedFailures in AI agents are often semantic deviations rather than explicit error codes, and this kind of "successful failure" requires the orchestration layer to have semantic-level error recognition capability.78% similarUnverifiedAI修复的前提是原始信息尚存，如果录音质量过差，缺失的信息无法凭空生成，强行补全容易产生失真75% similarUnverifiedAI系统的静默失败表现为技术层面健康（返回200状态码、日志无报错）但业务质量持续下滑70% similarUnverifiedAI摘要在安全场景失效的深层原因是优化目标与信息论重要性之间的错位，混淆了信息频率与信息重要性70% similarUnverifiedAI hallucination is not a random failure but an inevitable result of low-quality context input; when the input context is sparse, the model lacks a handle to anchor itself to real information.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/926643API
curl https://kongchang.com/api/v1/knowledge/claims/926643MCP
get_claim(id=926643)