We have already seen LLM misuse by counsels around the world. While this judge use is being tried out in the open for what seems like the first time in the world, I wonder how long until judges do the same thing as software devs and barely look at the output.