i'm confused when i see people who don't like AI say stuff like "well eventually it'll be good enough to replace me"
and i don't know where that's coming from. sure the hypemen are claiming that, but where's the actual evidence? feels like AI has been plateauing pretty hard for the last couple years for me except maybe for in agentic code
but all the advances in agentic code are hacks built on top of the same shitty LLM model that's stopped improving ages ago and those will also have a ceiling