Last time I contacted them I received a reply ostensibly from a human but which was so obviously LLM generated it was almost funny.

It was about an A4 page of text and spent most of it being empathetic to “how frustrating” my issue must be and how they were really keen to help. But it was extremely patronising, barely addressed my problem, and it felt like a really long winded and super polite way of saying “fuck you” to a paying customer.

At least that’s how I took it.

I’m sure LLM’s can be used really effectively in customer service but not by just blindly copy/pasting whatever it spews out.