I mean the point is missed in your response. You're talking about the current state of LLMs, but the premise of the question is a "what if" scenario where AI successfully takes over our jobs.
There's no need to define intelligence to point out the misinterpretation of OP's question.
Define “intelligence” with good enough precision and details to compare it against some complex algorithm (linux core etc).
Until then your statements bear no real meaning to argue about.