Edit a transcript
Edit a recording's transcript the way the transcript reader does: name a speaker or take the name back, split a passage before a chosen word, and write the passages as plain words to copy or export
Edit a recording's transcript the way the transcript reader does: name a speaker or take the name back, split a passage before a chosen word, and write the passages as plain words to copy or export
Give passages ([{ start, end, text, speaker?, speaker_was?, words? }], times in seconds, each word { word, start, end }) and, in order, edits: { op: name_speaker, from, name } says every passage of one label under a name and keeps the label it had in speaker_was; { op: take_name_back, name } restores it exactly; { op: split, passage, word } splits one passage just before a word, the first half ending where the second begins. Answers passages (the edited list; words, times and order never change), speakers (in the order they first speak), named_speakers (the ones named by the person) and text (each passage as 'Speaker 0:05' over its words, the Copy transcript and Export PDF text). Each edit is read-data/edit-transcript.ts's own. Refuses no passages, a passage without start, end and text, an unknown op, an empty name, and a split at the first word or past the last. It reads, saves, plays and sends nothing; the reader screen also follows playback, searches the words and reprocesses in another language, which this run does not.
Nothing: your AI starts it with your words.
Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this tool by what it does and runs it.
Say to your AI: “Use Trillion to edit a recording's transcript the way the transcript reader does: name a speaker or take the name back, split a passage before a chosen word, and write the passages as plain words to copy or export”