Hacker News new | ask | show | jobs
by msravi 607 days ago
>Theoretical physics is highly mathematical, but it aims to explain and predict the real world. Theories that fail at this “explain/predict” task would ultimately be discarded.

This isn't really true, is it? There are mathematical models that predict but do not explain the real world. The most glaring of them is the transmission of EM waves without a medium, and the particle/wave duality of matter. In the former case, there was a concerted attempt to prove existence of the medium (luminiferous aether) that failed and ended up being discarded - we accept now that no medium is required, but we don't know the physical process of how that works.
2 comments
westurner 607 days ago
We have lower error predictions but don't know why.

Methods for explaining why include Counterfactual inference and/or now quantum Causal interference.

All or some of quantum statistical mechanics, fluid dynamics, and quantum chaos intend to predict with lower error too

link
westurner 607 days ago
On the Cave and the Light,

List of popular misconceptions and science > Science, technology, and mathematics : https://en.wikipedia.org/wiki/List_of_common_misconceptions :

> See also: Scientific misconceptions, Superseded theories in science, and List of topics characterized as pseudoscience

Allegory of the cave > See also,: https://en.wikipedia.org/wiki/Allegory_of_the_cave

The only true statement:

All models are wrong: https://en.wikipedia.org/wiki/All_models_are_wrong

Map–territory relation: https://en.wikipedia.org/wiki/Map%E2%80%93territory_relation :

> A frequent coda to "all models are wrong" is that "all models are wrong (but some are useful)," which emphasizes the proper framing of recognizing map–territory differences—that is, how and why they are important, what to do about them, and how to live with them properly. The point is not that all maps are useless; rather, the point is simply to maintain critical thinking about the discrepancies: whether or not they are either negligible or significant in each context, how to reduce them (thus iterating a map, or any other model, to become a better version of itself), and so on.

link