Reddit user laments growing hardware barrier for local LLMs

A Reddit user argues that running local LLMs now requires far more VRAM than the 8-16GB norm of earlier days. The post reflects nostalgia for when gaming GPUs sufficed for experimentation.