I am using emergent ai to create a PDF annotation tool for estimators. To give you an idea, i'm a few grand in but its just as expected. I knew I was getting involved with rendering, optimization, GPU/CPU workload distribution, complex geometry, communicating with Windows, testing different PDF libraries, etc etc. I bought a little 50 page booklet to write all my ideas and features before I invested my time and money.
I had to hold the AI's hand step by step. I'm still ironing out things and expect at least another grand. The deeper you go, the easier it is for the AI to fix one thing while breaking another.
If you are creating something that is text based, you can surely find your way to success. I did this first to test the capabilities. Only a couple hundred bucks got me a simplified "Excel", again for estimators.
For reference, I am not a programmer but i know basics of programming.