Real output from the pipeline, generated ahead of time. Every upload is converted, transcribed, summarised and split into chapters; a line can be spoken in the speaker's cloned voice, and the whole video dubbed into another language.
The voices here are synthetic text-to-speech voices, not real people.