Again you're making unsupported assertions ("trained most on Python code") and coming to unsupported conclusions ("they seem to prefer it") -- my own very quick experiment shows that it depends on how you ask ("app" versus "code") at the very least.
The choices of JS and Python were pretty solid for the prompts I gave. Maybe the LLM is, well, making a reasonable decision in context, rather than just defaulting to Python!
"LLMs prefer Python" is an over-simplification. Python was already very popular, so I agree that LLMs are likely to entrench that popularity, but that doesn't mean Python will grow into other areas where it wasn't being used much before.
Why are you verbosely repeating my own replies? You asked why, I told you a theory that MAYBE because they were trained on Python code. You do know that's how a LLM or even a simple neural network works right?
Of course it is an over-simplification. Should I do an empirical scientific study before I can reply that MAYBE LLMs seem to prefer Python? I was talking from my own personal experience.
Are you using a LLM to write your replies? Because they seem very odd to me.
