I'm still not convinced that LLMs are as groundbreaking as everyone's making them out to be. They're just advanced language models, not necessarily AI breakthroughs, and we're still seeing the same old 'echo chamber' issues in their responses.