Nah. You're missing the point. Get the prompt wrong and it doesn't natter who/what is doing the output. These models are trained trained on the public internet

Nah. You're missing the point. Get the prompt wrong and it doesn't natter who/what is doing the output. These models are trained trained on the public internet. As your problems get more esoteric, AI's ability to recall answers becomes less accurate. Maybe Prompt Master, and then QA is the future of tech?