> if somehow designed not to hallucinate We know this isn't possible at the moment. Are we going to legislate for something that is not yet technologically pos

> if somehow designed not to hallucinate We know this isn't possible at the moment. Are we going to legislate for something that is not yet technologically possible? Should judges decide cases because maybe ML researchers will figure out how to reliably stop models from hallucinating?