> LLMs have already hit their limits
This is the most profoundly idiotic sentiment to ever come out of the AI discourse. Not only has it been sung for so long that it's been proven wrong mutliple times now, it also pulls the strawman of the century because NO ONE working in frontier AI labs is pushing the narrative that "we don't need anything to change to reach AGI". You can stick your head in the sand all you want, progress is happening and it's not slowing down as of yet. Your knowledge of AI being limited to talking to chatbots doesn't change a thing.