Podcast: Play in new window | Download | Embed
Watch on Archive / BitChute / Odysee / Rokfin / Rumble
Did you know that there are full, hyperlinked transcripts available for all of my documentary productions? Well, there are! Did you further know that Substack just added an automatic machine transcription which is now transcribing all of my audio uploads to their platform? Well, they did! You can see the transcripts on the latest uploads at corbettreport.substack.com. While you’re there, why not sign up for my free email list and/or become a Corbett Report member? That way, you can get access to my weekly newsletter (which, by the way, is NOT an email!)
Cool!
Yes very cool indeed, and very much appreciated!
It’s another great research tool to have.
Nice! I sometimes have trouble understanding some of your guests, this will be very helpful.
Also James, I listened to another creator’s podcast that had Dr. Robert Epstein as a guest and I think that he would make an excellent guest for your podcast. Dr. Epstein is solutions focused and making very good progress on giving the people transparency into the inner workings of the behemoth Google, and capturing their data for public consumption. Here’s his website if you are interested https://mygoogleresearch.com/
My experience with transcripts and closed caption has not been great. I have seen places where the text is the exact opposite of what the speaker actually said and realized that if someone were relying exclusively on the transcript/closed caption for information, they would have the completely wrong idea of what the person had said. Yes, it has been that bad. Sometimes it’s just nonsensical.
The WAPF podcast has had some real doozies in their transcripts and I have intended to let Hilda know, but I keep forgetting. I only rarely look at the transcripts – usually only when I want a bit of information and don’t want to have to listen to the whole podcast again.
I scanned this transcript and thought they did pretty good on it. There was a sentence they split in the middle, beginning the next paragraph in the middle of the sentence. They just deleted the half-word false starts instead of trying to transcribe them, and left out an “um.” Okay, that’s pretty intuitive.
Then I decided to listen again and read the transcript at the same time. It is interesting that the transcript highlights each word being spoken as it goes along.
There are a couple sentences they left out entirely and the wrong words in the next section are incorrectly highlighted as James is talking. Then, while the other voice is talking, nothing is highlighted, so it catches up with itself to get back on track below that.
At one point, James says, “But … but…” to emphasize a change (there are now transcripts). They leave out the second “but.”
One time James says a word, then goes back to add in another word. They leave the first part out and just put in the final “take.”
So this could be interesting … or entertaining … or misleading …. Time will tell. My suggestion is to never rely on transcripts. If something sounds off, it probably is. Verify it with the audio. The audio is better, more reliable … in my opinion.
I also think that if we have trouble understanding someone, the computer-generated transcript is going to be even LESS reliable.