Do OCR across all videos

[yarikoptic]

and make it into CC based format with keywords of what was found on the slide.

Most frequently slides contain more information that potentially is spoken. I think OCRing all of them and providing a complimentary to CC file in CC/subtitles format would be useful to search/mine for keywords etc.

Again -- only for content under permissive licenses

[yarikoptic]

with AI advent and recent RF to use annextube which makes organization more structured it is even more feasible now!