Patriarchal blessing transcription via App or AI?

[dtphoto]

Is there a recommended tool for voice to text transcription of Patriarchal blessings?

Our stake set apart a new patriarch today, and he will be receiving the prior patriarch's hardware. The stake president is young and suggested using an app for converting voice to text, but didn't specify anything specific. I'm the stake high councilor assigned to technology, so I have ideas for solutions, but don't want to reinvent the wheel or break protocol.

Obviously, trust but verify. Technology is getting better and better with accuracy. I searched this community and found reference to "Dragon naturally speaking", but nothing newer than that. I'll open the conversation (and future searches) up to AI engines (ChatGPT prompt: "transcribe this audio file") or OneNote or a voice recorder app. Some AI engines will absorb the content to a server, and other programs will work locally with nothing leaving the device...the latter is required.

The only limitation I've found is the small paradox for text:

Blessings may either be transcribed directly into the text box provided on the screen or transcribed into a word processing program and then copied and pasted into the text box provided. No copies of blessings should be saved to the computer.

Since the browser window might timeout before transcription is complete, typing into MS Word would be good. When done, copy and paste. That word document would probably auto-save the draft copy, and I'm sure the letter of the law is to prevent the patriarch from amassing a library of the blessings they gave. (Possibly to avoid copying blocks of similar text from one person to another.)

[BrianEdwards]

I'm uncertain if you got your "paradox text" reference from this document Creating Blessings in the Patriarchal Blessing System. A bit earlier in that section it also states that a scribe can be called to help with transcribing blessings. This appears to indicate a human transcription is desired, although I guess maybe the scribe could be the one using the tech solution instead of manually writing it out. But the FAQ references another document "Calling a Scribe for the Patriarch" that I couldn't locate, which may or may not have additional details.

The date of the FAQ is Sep 2023, but of course with AI and translation apps, the Church continues to evolve in what it permits and supports. It may be that there is newer information that hasn't made its way into other support materials, or that may be only available to the Patriarch himself.

I personally wouldn't have an issue with using a non-cloud tool to create the transcribed text, since regardless of how the transcription occurs, there is clear wording in the FAQ about the responsibility of the Patriarch to carefully review the transcription for accuracy (grammatical as well as spiritual). And it may well be that other Forum members already know of Patriarchs doing something like what you describe.

[davidbcarter]

This is a very timely topic for me. I was ordained Patriarch for my stake 10 days ago so I'm Just beginning to learn the process. Being a long-time tinkerer/homelabber, I am very interested in learning what options I have when it comes to automated transcription that are consistent with Church guidelines/requirements. Being a former Ward and Stake Clerk, I immediately thought of this discussion forum.

When I log into the Patriarchal Blessing system on LCR, there is an FAQ section, and one of the questions is this:

Can I use my voice-recognition software?
The Patriarchal Blessings System does not accept voice files directly at this time. The voice-recognition system may be used with Microsoft Word, and the text it generates in Word can be copied and pasted into the transcription portion of the Patriarchal Blessings System.

I realize that the voice recognition (aka Dictation) functionality is Word is not the ideal because it does not accept audio files as input, but I have tested it by activating it and then playing back an audio file while it listens, and that seems to work pretty well. An app/tool that would allow me to point it to a local WAV or MP3 file would be great, and I hope we can find some acceptable solutions, but in the meantime, this seems to be a workable solution that is explicitly permitted per the quoted FAQ above.

I would love to see this topic continue to evolve. I recognize there is a very limited audience for this compared to many other topics discussed here, but it is one of great interest and importance to those of us who need it. For those looking for ways to help your Stake Patriarch, thank you! It means a lot!

[russellhltn]

I realize that the voice recognition (aka Dictation) functionality is Word is not the ideal because it does not accept audio files as input,

What version of Word are you using? I did a web search and found this: Transcribe your recordings for Microsoft Office 365.

Failing that, I would look for a "connector" that could pipe the file into a virtual audio card to feed into Word.

[davidbcarter]

I realize that the voice recognition (aka Dictation) functionality is Word is not the ideal because it does not accept audio files as input,

What version of Word are you using? I did a web search and found this: Transcribe your recordings for Microsoft Office 365.

Failing that, I would look for a "connector" that could pipe the file into a virtual audio card to feed into Word.

So, it looks like we can "upload" a file for transcription, but that word "upload" seems very iffy when we're talking about a person's patriarchal blessing. I'd want someone from the Church to give me the green light before I did that. Plus, the more I think about it, it's not a bad process to listen to it again as it's being transcribed. A thoughtful review by the patriarch that goes beyond just making sure the words are on the page seems to be the intent behind the process anyway.

Thanks for the additional info and suggestion!

[russellhltn]

Ah, I didn't read that carefully. I thought it was happening right on the machine.

But reading more closely, it seems like when it's going direct into Word, it is making a recording and doing the same thing (uploading it). Can you transcribe in Word when you're off-line?

The reason I suggested uploading or using a connector is to remove the "speaker to microphone" step that degrades the audio.

[davidbcarter]

But reading more closely, it seems like when it's going direct into Word, it is making a recording and doing the same thing (uploading it). Can you transcribe in Word when you're off-line?

You're right. I hadn't noticed that before. When I am offline, the voice recognition gives an error and doesn't work. That's interesting that the FAQ in the PB system specifically mentions this as an option if that's the case.