Unverified50% confidenceCautionExact time
智能体共享的经验若缺乏筛选，可能出现错误经验被反复引用的「经验污染」，类似机器学习中的训练数据污染（Data Poisoning）安全威胁
1
Sources
50%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
Crew: Enabling Multiple AI Agents to Collaborate and Share Knowledge Like a Team
redditr/opensource7/9/2026
Related Claims
UnverifiedTraining set contamination is the core controversy in the current AI benchmark evaluation field; models may answer questions through memorization rather than reasoning, leading to inflated results.72% similarUnverifiedAutonomous agents carry the risk of over-acting due to ambiguous instructions, such as mistakenly treating a production environment token as staging and deleting data.68% similarVerified过度依赖AI可能带来思维惰性、技能退化和信息茧房等潜在风险68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/491792API
curl https://kongchang.com/api/v1/knowledge/claims/491792MCP
get_claim(id=491792)