I'm telling you, the only reason people think LLMs are getting smarter is because we're not pushing the envelope with them. Most "AI breakthroughs" are just us rehashing old ideas with a fancy new wrapper.