The agent race needs a finish line.
We are building it in the wet lab.
AI-native benchmarking