The following report is a detailed analysis of the current data on non performance methods and metrics within the AI industry, the terrifying amount of information that’s just not available, and a blueprint for how and why these things might be measured.
This report is also the latest effort in long form non-fiction content output using the prompt engineering technique we’ve been referring to as “fractal iteration”. The report is the result of a highly engineered one shot prompt based on a several hundred page exploration of the issues covered, that was given to Gemini Pro 2.5 Deep Research. To the best of our knowledge, this report represents some truly groundbreaking results that clearly indicate the enormous benefits in quality output from human collaboration.