Sorry, really silly nit. But curious if the 80k transcripts are auto generated or manually transcribed?
This one felt a bit harder to parse than the typical 80k podcast transcripts, I think because you left in pretty much all of the “ok”/”so”/”yeah” filler words.
Or might just be because it’s a public talk instead of an interview? Though I would actually expect the latter to be more conversation and less prepared.
Normally with the podcasts we cut the filler words in the audio. This audio was unedited so ended up with more filler than normal. We’ve just done a round of edits to reduce the filler words.
Sorry, really silly nit. But curious if the 80k transcripts are auto generated or manually transcribed?
This one felt a bit harder to parse than the typical 80k podcast transcripts, I think because you left in pretty much all of the “ok”/”so”/”yeah” filler words.
Or might just be because it’s a public talk instead of an interview? Though I would actually expect the latter to be more conversation and less prepared.
Normally with the podcasts we cut the filler words in the audio. This audio was unedited so ended up with more filler than normal. We’ve just done a round of edits to reduce the filler words.