As I understand there are some models in both Econ and Finance where they literally took equations from thermodynamics and just said “what if this variable represented money instead of heat”.
99% of models developed in open view in both fields are either generally wrong in practice or are "correct" insofar as they were correct at one point in history or set of circumstances, but have since been factored into the efficiency of the system and are no longer relevant. Thermodynamics is a nice analogy for a lot of effects in economics, but it's important not to stretch that analogy like butter over too much bread
but humans are not atoms in brownian motion. they need reliable models of human behavior informed by neuroscience/psychology