Theo Browne revisits his earlier assessment after reviewing his own coding-agent tracesEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria.. He reports substantially longer task times in his usage and says request-start delays compound across multi-step tasks.
Theo Browne discusses caching and tooling behavior, while acknowledging his measurements reflect his own usage and that some slow tasks were interrupted. He also describes technical discussion with Google staff about possible causes and improvements.
Watch on YouTube
