And we would absolutely do it more often if compiling would cost as much as training of an LLM costs now.

I considered adding "normally" to the binary modifications expecting a response like this. The concepts are still worlds apart

Weights aren't really a binary in the same sense that a compiler produces, they lack instructions and are more just a bunch of floating point values. Nor can you run model weights without separate code to interpret them correctly. In this sense, they are more like a JPEG or 3d model

JPEGs and 3D models are also executable binaries. They, like model weights, contain domain specific instructions that execute in a domain specific and turing incomplete environment. The model weights are the instructions, and those instructions are interpreted by the inference harness to produce outputs.