Unverified50% confidenceBenchmarkExact time
有研究表明在自发语音场景下WER从5%提升到8%，用户感知的任务完成率可能下降超过30%
1
Sources
50%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
The Real Bottleneck for Conversational AI: Model Capability or Training Data Quality?
redditr/LanguageTechnology7/10/2026
Related Claims
VerifiedIf the end-to-end latency of a voice dialogue system exceeds 800 milliseconds, the conversational experience degrades significantly.75% similarUnverifiedVoice cloning technology can reconstruct a target person's timbre, intonation, and emotional characteristics from just a few minutes of audio samples, and products from companies such as ElevenLabs and Resemble AI have already lowered the barrier to a level accessible to ordinary users.72% similarUnverifiedResearch indicates that as many as 60%–70% of syllables in native speakers' everyday conversation undergo varying degrees of reduction.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/496984API
curl https://kongchang.com/api/v1/knowledge/claims/496984MCP
get_claim(id=496984)