The LLM part of the "new" problem is the speed at which it can proliferate and the trust people seem to have in AI answers. Idk