Gpt can produce pretty realistic looking generated text. If you wante to analyze the text and work out what it's saying, or reduce it down, or simplifyit,. Or lik check for sentiment, whatever. So transition totsorof the producty aspect, is there, what's possible now that wasn't before? In te sense that ik, say, dolly, can i take a prompt and produce an image? Gpt has given building blocks that start us can actually work withor tird making products.