We live in a time where AI chatbots speak way better English than most native English speakers.
I'm wondering whether soon we'll see the outsourced CS chats run through a 'small' efficient LLM to 'clean up' (or in some cases, just re-localize) the English of such workers to match the locale of the caller.
unless the bots have been trained on outsourced support worker chat threads?
I dont like this timeline
They were trained on both and can do both, it's a matter of asking for either in the system prompt.