I am now convinced that 99.95% of the time (lower bound, by the way), when someone says 'AI will solve our problems!', not only are they lacking an actual understanding of what LLMs actually are, they also don't understand what their actual problems are.