Couldn't agree more with you. At the end of the day the people building the most successful products, are too be busy to be formalizing their experiments into research papers. While I have respect for academic researchers, I think their perspective is fundamentally very limited when it comes to AI engineering. The space is just too frothy.