Research
We build AI-nativity as an evidence-classed discipline in the open. Don't take our word — read the artifacts. Plain-language versions live in Notes.
Public artifacts (verifiable now)
72 questions across production courses · 7,818 model responses from ten LLMs · 28,593 evaluative judgments.
How we calibrate
We publish datasets and preprints and submit to peer-reviewed venues. We don't overclaim: no unearned superlatives. We hold ourselves to the same scale — L0.9 / 5 today. The evaluation harness and the L0–L5 rubric are available to researchers on request.
Open research directions
- The institutional coverage layer none of the 24 standards we mapped currently close.
- How organizations learn AI-native practices — workflow mining at production scale as a window into organizational learning (the CourseFactory line of work).
- Validating L0–L5 as an organizational measure — construct validity, inter-rater reliability, level→outcome longitudinal designs.
Further directions — agent self-certification, alignment efficiency, portable agent track records — are covered in white papers available on request.
Who we're looking for
Collaborators in ML / agent evaluation, learning sciences, organizational psychology, and measurement science — and universities and training providers who want to run their programs on a measurable human–AI learning architecture (HALA).
We partner with AI labs, universities, and organizations building the next generation of human–AI systems. All the ways to work with us — researchers, universities, agent vendors — on the Collaborations page.