As an AI researcher myself, I welcome LLMs hitting a scaling wall. Not because I hate them or anything, but because reaching the “peak LLM” would …