Daniel's essay claims the century‑long stalemate in fundamental physics arises from methodology, not unsolvable puzzles. He splits the issue into a mathematical model that matches data and a physical theory that explains underlying reality. The perpetual Copenhagen‑Everett‑Bohm debate, he argues, shows that quantum mechanics is being used as a physical theory when it is essentially a probability calculus.

Since 1928, Standard Model physics has amassed a catalogue of fields and interactions but retained the same ontological scaffolding: spacetime from general relativity and wave‑like spinor matter from the Dirac equation. Additions such as gauge bosons, the Higgs, or supersymmetry merely populate that framework. No empirically confirmed ontology has supplanted it, and every successful prediction since the 1980s relied on pre‑1973 assumptions.

Daniel draws a parallel with modern machine learning, where researchers dispute whether large models build internal world representations or merely fit surface statistics. Yann LeCun maintains that today’s LLMs predict next tokens without understanding underlying physics, while others argue pattern‑matching at scale could constitute a form of understanding. The essay concludes that breaking the methodological impasse demands treating calculi as tools, not ontologies.