The world is being quietly rearranged by people who write very long documents.


The title they went with Symmetric observations without symmetric causal explanations Noisy translates that to

Symmetry in data doesn't guarantee symmetry in causes


A mathematical paper proves that when you observe symmetric patterns in data, you cannot assume the underlying causes are symmetric — you may need to consider radically different causal explanations. This matters because scientists often use pattern symmetry as a shortcut to eliminate impossible causal models, but this shortcut is mathematically invalid.
If causal inference relies on symmetry shortcuts that don't actually work, scientific reasoning built on those shortcuts — from physics to biology to social science — may be discarding valid explanations and keeping invalid ones.

If you insist
Read the original →