Golden Datasets Are Dead
There's an instinct when you start building agent evals to replicate what the big benchmarks do. You see TerminalBench or SWE-bench or whatever, and there's this nice hill to climb. Model releases improve the score, progress is visible, stakeholders are happy. So you think: why not build an internal version? Start at 10%, iterate throughout the year, end at 80%. Show the chart in your quarterly review.
It doesn't work. Here's why.