Nowadays, I prefer AI CS agents to humans. I just had a chat with an AI yesterday, it understood me perfectly even when I made mistakes, I was impressed.

In contrast, humans tend to paste me the same barely-relevant macro over and over, no matter how much time I spend explaining my issue.

Yeah, at least LLMs read everything you write (for now). Human first level support agents are incredibly frustrating if you have to explain anything with more than one logical step.