Experienced the system myself (Physics PhD -> postdoc -> PI at a National Lab -> Data Science). Some factors probably contributing to the situation are:<p>1. ECONOMY: Bleak economic outlook for research and necessity to find job, often manifested as looks of pity toward "researchers and postdocs" during family gatherings (you know you've experienced them ;) )<p>2. VISIBILITY: Optimizing for visibility favors performative science (i.e. announcing minuscule results on Linkedin/Twitter) over deep pursuit of scientific truths<p>3. ROI DIFFERENTIAL: Even marginal familiarity with coding, math, or stats (baseline for science PhDs) translates to decent earning potential in industry<p>4. PUSH FOR MONETIZATION: Young faculty and students in deep tech thinking more about "whats minimum I need to monetize my research" over issues of validation, reproducibility, etc (these can come post funding)<p>These and other factors lead to folks basically kicking the can down the road, a.k.a. the "Future Directions" section in papers.