E
Endre
This is not of this world. The "Summarize Video" function is incredible but both for Youtubers and Vievers, it would open a whole new world to analyse all the hundreds or thousands of comments, would come from somewhere mankind has never gone before.
I did it myself. Held the bear of ChatGPT while it made me a Python app that downloads the comments into a TXT file so I could upload it to ChatGPT and prompted it up as a Social Media Assistant. IT'S INCREDIBLE!
Now, the required functions:
0: Extract the comments into a back-end TXT file, then give it to the LLM with a proper prompt*.
1: Make the TXT file downloadable for further analysis. Or just make us able to follow-up the comments file in Merlin's History, for a further chit-chat.
  1. Make a form for us for this function, to give our custom prompt. For example it's not the same prompt that should be used by a viewer as the Youtubers themselves.
MY UNIVERSAL PROMPT (for a default one?)
Be my social media assistant. In the file, you will find all the comments under a YouTube video. Each line starts with the first "@" symbol, identifying the user who made the comment. If you see additional user mentions in a comment, that means the original commenter is replying to someone else, indicating an ongoing conversation. Please group these conversations accordingly.
Analyze the comments based on the following criteria:
  1. General tone: What is the overall sentiment of the comments towards the topic or the video?
  2. Summarize: Write a brief summary of the most positive and the most negative comments overall.
  3. Highlight: Select twenty comments that are worth my attention because they provide useful information or are interesting from the perspective of the topic.
  4. Make a summary of the funniest comments.
  5. Summarise what the video uploader can learn from the comments in terms of further content creation.