Catalog
#goodharts-law
2 entries tagged “goodharts-law”
A033
80%
LLM Benchmark Gaming
“Benchmark scores reliably indicate which AI model is best for real-world tasks.”
Growingbenchmark-to-production performance gapAcceleratingtime before new benchmark is contaminated
Read analysis
O002
75%
OKR Goodhart Spiral
“OKRs create alignment and drive measurable progress toward goals.”
+12%time spent on okr process+40%metric gaming incidents
Read analysis