ML is useful for many things, but not for predicting scientific replicability

How the veneer of AI is used to legitimize awful ideas

Science rests on replicability and reproducibility. Without it, we can’t separate useful insights about the world from artifacts arising due to chance, and even fraud and malpractice by researchers. There are other costs too: one study estimated the cost of irreproducible research at $28 billion every single year, in the field of preclinical research alone. Earlier this year, a paper in the widely read PNAS journal raised the possibility of detecting non-replicable findings using machine learning (ML). The authors claimed ML could be used to predict which studies would replicate, and…

related reading

• How much science is verifiable? Results from replicating ICML 2026 oral paperssai.science
• Scientists should use AI as a tool, not an oraclenormaltech.ai
• Sloppy Use of Machine Learning Is Causing a ‘Reproducibility Crisis’ in Science | WIREDwired.com
• Four more things we worked on in 2022normaltech.ai
• Introducing the REFORMS checklist for ML-based sciencenormaltech.ai
• Is AI Reasoning Right for the Wrong Reasons? | Quanta Magazinequantamagazine.org
• Machine learning needs better tools – Replicate blogreplicate.com
• Half of social-science studies fail replication test in years-long projectnature.com
• Predicting Empirical AI Research Outcomes with Language Modelsarxiv.org
• Reproducibility and transparency: what’s going on and how can we helpnature.com
• Can AI automate computational reproducibility?normaltech.ai
• What's Wrong with Social Science and How to Fix It: Reflections After Reading 2578 Papersfantasticanachronism.com