Unverified50% confidenceFactExact time
Tianji personally tested 22 experts this season, of which 13 can only talk, and only a minority can actually get work done.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/6/2026
First Seen
Valid until: 1/4/2027
Sources
Related Entities
Related Claims
UnverifiedOpenAI's published blog posts only tested up to around 16 agents and did not push experiments to a scale of ten thousand agents.64% similarUnverifiedTheoretically reliable clustered standard error estimation typically requires at least 30–50 clusters, whereas this study has only 19 independent time units, making the sample size relatively small.60% similarUnverifiedAIME has only 30 samples, and tiny score differences derived from running a small number of seeds are not statistically significant.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/979533API
curl https://kongchang.com/api/v1/knowledge/claims/979533MCP
get_claim(id=979533)