Our series in AI evals continue with an exploration of the types of benchmarks. The opinion series explores the trend of all the major AI labs creating the same primitives( research, reasoning, sea...