I've had several good conversations about AI/LLMs today.
After a few weeks of thinking I could see what they're useful for, I think I'm officially a skeptic now.
Every conversation has had:
– They’ll get better though!
– Yeah, sure they make mistakes, but you learn to deal with it
Or has been about a use case where a simpler, deterministic and far more efficient method would be better.
Mostly though: I'm just sick of talking about and hearing about it. It's just become word salad now.