Taking Notes for Social Web CG taskforce meetings
Technical Discussion
11
Posts
2
Posters
0
Views
-
julian since you asked!
-
I'll need a bit more than upvotes here
-
thisismissem hah yeah, I was doing dinnertime and didn't have the bandwidth to answer, one sec 🙂
-
thisismissem hedgedoc or hackMD worked surprisingly well. Before I had a Google Docs sheet and participation was low because of the access control barriers.
I would not be opposed to an AI scribe, and I actually have access to one with my Google Workspace sub, but I think that comes with a lot of baggage that precludes its use.
The worst part of scribing your own notes is that you can't participate. All my notes for sections I lead don't ever get captured fully, and I have to do it post-meeting... about half the time it doesn't happen at all 😦
As mentioned offline, perfectly fine to use dmitri@social.coop's HedgeDoc instance.
-
Yeah, in the past I've gained consent to record a call & use a local model to transcribe which I then edited, but it was time consuming.
In the T&S taskforce grant I've applied for money to pay a scribe for their time
-
I think I'll give it a try. I did some light research and am going to give Whisper.cpp and some Ollama models a try. I'm not sure whether they'll have the ability to summarize and distinguish voices, but let's see... I'll report back in this thread.
-
I have been informed that there is an (outdated?) Etherpad instance at https://pad.w3.org that you authenticate with via Basic Auth using your W3C Account credentials. Unfortunately it needs that auth just to view or contribute to documents, which probably isn't helpful.
-
Okay, I spent the better part of an hour playing around with it (after compiling
whisper.cppand going down a weird rabbit hole — thanks Gemini!)I recorded some test audio, had it transcribed, and passed it over to ollama running llama3.1, with a prompt to summarize the meeting into bullet points and action items. It seemed to go over well, although I don't quite know how it will handle digressions, asides, and hour-long meetings :)
It might be we can adjust the prompt so that it isn't summarizing so much as efficiently condensing meeting discussions.
Definitely worth a shot if only to reduce the burden of scribing while chairing a meeting!
-
Okay, I spent the better part of an hour playing around with it (after compiling
whisper.cppand going down a weird rabbit hole — thanks Gemini!)I recorded some test audio, had it transcribed, and passed it over to ollama running llama3.1, with a prompt to summarize the meeting into bullet points and action items. It seemed to go over well, although I don't quite know how it will handle digressions, asides, and hour-long meetings :)
It might be we can adjust the prompt so that it isn't summarizing so much as efficiently condensing meeting discussions.
Definitely worth a shot if only to reduce the burden of scribing while chairing a meeting!
julian I used vibe.app when I did transcription, and it kinda butcher audience a few months ago and "fediverse" became like "fetty verse", but otherwise was okay.
-
julian I used vibe.app when I did transcription, and it kinda butcher audience a few months ago and "fediverse" became like "fetty verse", but otherwise was okay.
Whisper.cpp is surprisingly good. It is getting 99% of it... missing things it doesn't know about (like
PyFETinstead of Piefed), but it definitely knows what activitypub is, so that's handy.Thanks for pushing me (even if indirectly) to do this. Looking forward to trying it out for real.
Gli ultimi otto messaggi ricevuti dalla Federazione
-
@julian @mariusor @helge
fwiw, letterbook is fine with that
https://github.com/Letterbook/Letterbook/blob/6bf60b1b3d37a8b5d49a1d2554bfd5e4468f00de/Source/Letterbook.Api/ActivityPub/ActorController.cs#L309
-
The thing I'm adding to FIRES is the ability to label data as you import it.
-
trwnh@mastodon.social it's less so that I want receivers to do what I want (the ship has sailed on that), but rather that we have the opportunity to provide some guidance on preferred behaviour.
What's clear here is that there are going to be two separate actions, "Delete object and all children, recursively", or "Delete object but retain children".
We're discussing the best way to represent those two actions.
cc rimu@piefed.social nutomic@lemmy.ml silverpill@mitra.social
-
Since it's the end of the week, you might be wondering: where's FediMod FIRES 1.0.0?
It's coming, I just decided to add some more features, and most of my week was:
- meetings (7 of them!)
- sorting out things for the ActivityPub Trust & Safety Taskforce
- setting up some infrastructure for SWICG taskforces to use since one thing broke & I needed to update it.So I didn't get to work on FediMod FIRES as much as I wanted to.
-
@julian @rimu @nutomic @melroy @BentiGorlich the complicating factor here is not what other people do, but that you would be using the terms incorrectly or imprecisely according to their definition. this happens all the time in natural language where people sometimes use words they don't fully understand or use them with definitions not matching consensus. it's how we get people saying "literally" for things that are not literal, and other such slang.
-
@julian @rimu @nutomic @melroy @BentiGorlich is this a problem? it seems to only be a problem if you require the others to behave exactly as you do. the same "issue" applies to any activity in general. say you send a Delete; the others can do what they want:
- purge all children
- orphan all backlinked objects
- replace with a tombstone
- rewrite content to say "this post is deleted"
- ignore your activity as unauthorized or invalid or spamthe intent could be clearer...
-
@mariusor @julian @helge i don't think "all the ancestors" makes sense for inReplyTo. by doing that, you are claiming that your post is a response to every post in the thread above it. multiple inReplyTo still makes sense but should be used only where you are actually responding to certain things. if you want ancestors, define a property "ancestors" which is a list of ancestors ordered in a specific way (like in the mastodon api)
-
@thisismissem @nutomic it's not a JSON-LD thing. it's the lack of schematic constraint within the Activity Vocabulary. any property not defined as "functional" can have more than one value.
