In most professional industries getting to the right answer is only half the problem. You also need to be able to demonstrate why that is the right answer. Your answer has to stand up to criticism. If your answer is essentially the output of a very clever random number generator you can't ever do that. Even if an LLM could output an absolutely perfect legal argument that matched what a supreme court judge would argue every time, that still wouldn't be good enough. You'd still need a person there to be accountable for making the argument and to defend the argument.

Software isn't like this. No one cares why you wrote the code in your PR. They only care about whether it's right.

This is why LLMs could be useful in one industry and a lot less useful in another.