Oct 30, 2024 at 1:40 PM Shotcut 24.10 adds initial AI features through OpenAI's Whisper audio transcription model
Shotcut has released version 24.10 of its free, open source, and cross-platform video editor, now built on FFmpeg version 7.1. This update introduces AI capabilities through OpenAI's Whisper, integrated via the whisper.cpp project. A basic model is included, with an option to download an enhanced model for improved transcription. However, users should anticipate occasional errors, and the AI-powered speech-to-text feature currently only supports Apple Silicon GPUs.
The update also brings enhancements to transitions. The Ripple Delete transition now fully restores clips, while the Lift transition automatically fills gaps with adjacent clips. Adjusting clip positions now extends transition durations instead of creating gaps. This release also includes various other improvements and bug fixes, improving the overall user experience.
OpenAI's AI transcription tool hallucinates excessively - here's a better alternative A researcher 'found hallucinations in eight out of every 10 audio transcriptions he inspected' from Whisper.
https://www.zdnet.com/article/openais-ai-transcription-tool-hallucinates-excessively-heres-a-better-alternative/
Why is shotcut more popular than kdenlive