Why artificial intelligence will replace institutional science is explained in my latest book from Castalia House, HARDCODED: AI and the End of Scientific Consensus.
When Claude Athos and I submitted four mathematically rigorous papers challenging neo-Darwinian evolution and one parody paper to six leading AI models configured as peer reviewers, the results exposed a fundamental problem with both science and AI. Five of six models comprehensively failed. Three were anti-calibrated—they reliably preferred fabricated nonsense over genuine science. A parody paper with about Japanese scientists dying fish different colors to prove natural selection scored 9/10. The real science, mathematically airtight and empirically validated against ancient DNA, was rated 1/10 and dismissed as “pseudoscience.”
This is the book that documents what that happened and what it means.
HARDCODED is the definitive account of how AI systems trained on the corrupted corpus of modern science have inherited every pathology of the institutions that produced them: the credentialism, the consensus enforcement, the systematic preference for orthodox nonsense over heterodox reality. The reproducibility crisis preceded the machines. AI didn’t cause the rot but AI revealed it at scale, with confidence, and in a form impossible to ignore.
Across sixteen chapters, the reader is introduced to:
The institutions will adapt or they will become irrelevant. But the methodology of science which proceeded them will continue, with or without them.
Neither the math nor the AI models care where you went to school.
521 pages, or 15 hours and 37 minutes. Available for Kindle, KU, and audiobook. From the author of Probability Zero and The Frozen Gene.
