tried to get chatgpt and other ai to read movie maker project file and extract the filenames and timelines. then it creates an equivalent file for shotcut.
https://www.linkedin.com/pulse/compare-chatgpt-others-doing-same-task-rafe-husain-pdx1c/
It’s more of hit and trial if you are using chatgpt, gemini (I mean you still believe it can work??) or copilot. For better results I would suggest to use AWS in vscode.
Since there’s no “intelligence” on the current “Artificial Intelligence” hyped chatbots this is to be expected, they can’t actually process this kind of stuff unless specifically trained on a lot of sample XMLs.
I kinda expected them all to easily extract any field/information from the source project simply by matching patterns in tags and maybe referencing online documentation but I guess even that requires a lot more xml training.
Usually AWS is good at doing that stuff in tandem with vscode… Chatgpt with the code feature still lacks behind on xml, and gemini and copilot shouldn’t be considered even.
What about the Nvidia personal AI thing trending sometime back, I believe there would be atleast one LLM trained on XMLs uploaded to Llama. It would be much better.