Comment by lelag

Also, if we discard ChatGPT for just being a "fancy" autocomplete. What do we do when we discover we too are just "*really* fancy autocomplete" machine too ? LLM can output untruth but humans are really good at that too.