Do Big Lab Researchers Read Zero Papers? The ML Reproducibility Crisis
alex_peys · x · 2026-08-05
Responding to a claim that big lab researchers now read almost zero papers and view top AI conferences as full of overclaims, a developer argued this is actually a universal issue across all academic disciplines.
He noted that the reproducibility crisis just appears more obvious in CS and ML because the code is forced to be open, allowing people to try and fail to replicate results. Other disciplines, he argues, simply hide it behind "data available upon request."
Related event: Ex-Google Researcher Blasts AI Top Conferences for Exaggeration(5 posts)→
More from AGI Musings
- Top Researchers Launch 'Pax Machina' to Debate Institutions for Powerful AI — edelwax · 2026-08-05
- Pax Machina Launches: A New Publication Focused on AI-Era Institutions — ryan_t_lowe · 2026-08-05
- Restructuring Work in the AI Era: Become the Designer of Your Own Job — every · 2026-08-05
- Musk: AI and Robotics Will Trigger Unprecedented Global Bandwidth Demand — XFreeze · 2026-08-05
- Cognitive Offload in the AI Era: Augmentation or Surrendering Judgment? — cloudcrafterzNYC · 2026-08-05
- Using AI to Disguise Writing: Scholars Use Claude to Anonymize Peer Reviews — RobbWiller · 2026-08-05