Transcribing a song with a 15-minute script

AIs being AIs. I wanted to read the lyrics (I have a hard time understanding songs just by listening). Look at how much YouTube’s transcript helped me (image). lol

NotebookLM didn’t help either: “The lyrics you asked me to transcribe include: [excerpt omitted]. You have previously identified the song as “Homem da Meia-Noite” by Banda Eddie.”

So I went to Google Colab and asked “how do I extract the audio from a YouTube video” + “now I want to call OpenAI’s Whisper API with this mp3 file”. Lots of errors — which Gemini itself fixed — and 15 minutes later I had a muuuuch better version!

[The full lyrics transcript is omitted in this version.]

I’m really impressed by how much easier it is to write a script today (REMEMBER, it’s not a full app, it’s a script!)

For anyone who wants to see (the size of) the script:

Links

• youtube_2_audio_2_transcribe.ipynb (GitHub)

Originally written in Portuguese and published on LinkedIn on February 19, 2025. I translated it into English with the help of AI (Claude): it let me bring my LinkedIn posts to people who don’t read Portuguese without rewriting each one from scratch. I reviewed the translation to keep my original meaning and tone.