Allow adding new speakers after transcription and improve speaker tagging flexibility

Users are currently unable to add new speaker entities after an audio file has been transcribed. The initial number of speakers appears to be hard-coded, preventing the addition of new speakers (e.g., an audience member joining a discussion). While renaming and managing existing speaker assignments is possible, the ability to introduce entirely new speakers post-transcription is missing. This limitation also affects the flexibility of speaker tagging, as users cannot easily adjust the speaker count if it changes during the audio. Additionally, adjusting specific parts of the transcript to change the speaker is "wonky," as users cannot select more than one paragraph at once for speaker assignment changes. Improving the ability to add new speakers and enhance the overall flexibility of speaker tagging, including multi-paragraph selection for speaker changes, would significantly improve the usefulness of this feature.