Transcribe Podcasts with Whisper AI & Elixir in Livebook

Looking for your next practical AI project? Join me as I walk through the process of scraping a podcast RSS feed and extracting the text from its audio using OpenAI's open source Whisper AI model. Along the way you’ll see that the Elixir ecosystem is perfectly capable of performing AI tasks--made even easier with Livebook. Whether you’re new to AI and Elixir or a seasoned veteran in both, this one is for you! Links: - Livebook can be found at https://livebook.dev - Project code is available at https://gist.github.com/ChristianAlexander/bea72b83cfe62926580619979708add8 Timestamps: DEMO - 00:00 The Project - 00:48 Installing Dependencies - 01:30 Scraping Episodes - 02:01 The AI Part - 03:10 Formatting Results - 04:47 Project Ideas - 05:10

Whisper transcription is a classic case where GPU choice changes turnaround time. This best GPUs for Whisper covers it in detail. Watch next: Ollama Tutorial | Run Llama2 locally | 7 billion parameter model | No GPU | LangChain Integration.