As much as I’m not a fan of this, I’ll at least admit it was helpful in helping me quickly learn how to scale videos using FFMPEG. That, combined with the documentation and now I have a command saved to a reference text document to help if I ever forget.
Ffmprg is remarkably powerful. If it is a picture on your screen, you can use ffmpeg to do what you want to it. Recently discovered it could convert PNG, jpeg, and webp images back and forth. To think I’ve only used it for video.
I have zero interest in learning all the intricacies of ffmpeg so I find ChatGPT to be very useful. I’ve also used it for yt-dlp for downloading videos and converting the audio to mp3. Very useful. I personally save them as bash scripts so I can just input the file name or url as a command line argument. On Mac you can also wrap your bash scripts in AppleScript if you want to make applets for these functions. ChatGPT works great for apple script as well but I’ve had to feed it source code (eg from Apple Digital Mastering applets) to ensure it writes the new code correctly. You still must know what you’re doing.
Is yt-dlp really that difficult that you can’t take ten min to learn the syntax? Do you not worry that you might end up relying on this stuff too much?
For example, I know I’m completely reliant on GPS, and I would be useless without it
The syntax is easy, but the options change a bit depending on what you want to do. My entire job I requires me to use my brain so I don’t mind cheating when it doesn’t really matter as in this case. In my case I wanted a SQLite database to store URLs and playlist IDs for recording attempts and to make sure I don’t download the same video multiple times. I think I also had songs run thru music brainz for audio fingerprinting and mp3 tags. ChatGPT doesn’t get it right the first time but often gives a reasonable boilerplate piece of code as a template to start from.
As much as I’m not a fan of this, I’ll at least admit it was helpful in helping me quickly learn how to scale videos using FFMPEG. That, combined with the documentation and now I have a command saved to a reference text document to help if I ever forget.
Ffmprg is remarkably powerful. If it is a picture on your screen, you can use ffmpeg to do what you want to it. Recently discovered it could convert PNG, jpeg, and webp images back and forth. To think I’ve only used it for video.
I have zero interest in learning all the intricacies of ffmpeg so I find ChatGPT to be very useful. I’ve also used it for yt-dlp for downloading videos and converting the audio to mp3. Very useful. I personally save them as bash scripts so I can just input the file name or url as a command line argument. On Mac you can also wrap your bash scripts in AppleScript if you want to make applets for these functions. ChatGPT works great for apple script as well but I’ve had to feed it source code (eg from Apple Digital Mastering applets) to ensure it writes the new code correctly. You still must know what you’re doing.
Is yt-dlp really that difficult that you can’t take ten min to learn the syntax? Do you not worry that you might end up relying on this stuff too much?
For example, I know I’m completely reliant on GPS, and I would be useless without it
The syntax is easy, but the options change a bit depending on what you want to do. My entire job I requires me to use my brain so I don’t mind cheating when it doesn’t really matter as in this case. In my case I wanted a SQLite database to store URLs and playlist IDs for recording attempts and to make sure I don’t download the same video multiple times. I think I also had songs run thru music brainz for audio fingerprinting and mp3 tags. ChatGPT doesn’t get it right the first time but often gives a reasonable boilerplate piece of code as a template to start from.