[jbc1]

Implementation is what takes that raw R&D and uses it to solve problems real problems, which is where it derives it value. Armies of people slightly modifying slightly different saas is how those foundational ML models end up in the everymans workflow.