If your customer support LLM thing says "I've sent this conversation to our team of specialists" and doesn't actually do that, that's not a hallucination, it's an unforgivable bug

Using LLMs is like being a senior dev who is mentoring a junior dev. You need the experience to know what you need to explain and you need the language skills to explain things well.

When people say they don't find much use from LLMs I immediately suspect they are bad at mentoring too.

The future of AI is models that generate graphical interfaces. Instead of the linear, low-bandwidth metaphor of conversation, models will represent themselves to us as computers: rich visuals, direct manipulation, and instant feedback.

willwhitney.com/computing-in...