GPT-6 is multimodal, LLMs alone have no vision capability
"LLM" is now in practice a superset of "LMM"
"LLM" is now in practice a superset of "LMM"