Thank you for saying this! Every now and then I start to feel like I'm either the only sane person left or I must have gone insane and don't realize it.

LLMs amazed us and we've been trying to find a good use for them since and there's a lot things like this, replacing a transpiler with a non-deterministic machine that just guesses when it doesn't know the answer.

Shouldn't we be targeting LLMs at problems that traditional software absolutely can't handle?