A.C.T. Stack v2.0 : Content Archiving, Transcription, Translation, Diarization and Summarisation

[Note: this Topic began as an AIP Idea, and has been subsequently withdrawn and moved to the General category at the request of the author.]

A.C.T. Stack v2.0 : Pioneering the Future of Apecoin DAO

TLDR :

Essential tools to enhance user experience, streamline operations & foster collaboration and cohesion.

Description :

Abstract : The proposal envisions a suite of tools to enhance the user experience and streamline operations within the DAO ecosystem.

The ACT Stack v2.0 tools will include a lot of different tech, the comprehensive integration of which promises to exponentially amplify value creation while elevating the overall quality of user engagement.

About me : u know who I am :eyes:

Aim : Cause why not !? Also, was getting bored in sem 1 of uni

Team : Currently, I’m the only one on the team but I’ll be looking to expand shortly.

Where Was I ? : cooking :cook: :shallow_pan_of_food:

Status :

  • Ready for a Live demo (Youtube side) :white_check_mark:
  • Undergoing Validation (Twitter side) :brain:

This proposal Includes :

Scraping Content links → Live :white_check_mark:

Scraping the content which needs to be summarised. Scrapes all the required information from the link we provide and then exports all that information as a csv file which we’ll need to refer to at a later stage.

Context : I provide the link to HubermanLabs youtube channel and it returns me the file hubermanlab.csv with the information I need in the next step

Archiving → Live :white_check_mark: :question: (twitter side needs to be validated)

Archive the media files you want using the aforementioned CSV file at the click of a single button. We use youtube-dlp for archiving youtube files and twitter-dl for twitter spaces.

Context : The videos are being archived (Pic 1) || The videos have been archived successfully (Pic 2)

Diarisation + Transcription (ASR + DZ) → Live :white_check_mark:

Transcribing an Audio with multiple speakers leads to a lot of context being lost by virtue of absence of contextual data. Diarization helps to segment auditory stimuli according to who spoke what and when. It can easily differentiate between 10 people at a time and provide the much needed breath of fresh air.

Context : Each directory contains a video file, a summary in .txt and it’s diarized transcript in both .txt and .vtt file format

(PS : HubermanLab ASR + DZ was taking too long, I added the screenshots for ParkerNotes which I did beforehand)

just cause of the sheer amounts of individuals included
Summarisation → Live :white_check_mark:

Summarises each video and extracts the key takeaways and actionable insights from the provided media

Context : Each directory contains a video file, a summary in .txt and it’s diarized transcript in both .txt and .vtt file format

(PS : HubermanLab ASR + DZ was taking too long, I added the screenshots for ParkerNotes which I did beforehand. Also, there was only a single speaker in this clip so that’s why there’s no other speaker and it isn’t segmented)

Transcription → :question:

Bridging the Linguistic gaps so that our biggest strength (multiculturalism) doesn’t become one of the biggest barriers in our communication.

Nothing to show for it rn, except for this

Use Cases : (will update in 2nd draft)

  • SaaS : Providing it as a Service to Content creators (youtube, twitter etc)
    This will need a frontend
  • Extracting Insights : Extracting Actionable Insights from data and bundling it for use on an enterprise scale
  • Assistants : AI assistants primed on an ever expanding library of content (fed from the transcription + Diarization side of things)
  • B2B : Building AI powered audiobook platform (like blinkit or shortform.com) with inbuilt features like summaries, key insights, audio intelligence and being able to query a specific book for answers to your questions.

Benefits this would bring to the DAO and the ecosystem :

  • Users would be able to circumvent the language barriers (promoting intra group cohesion)
  • It’ll help us onboard new users who don’t have as much context as the vets, thus facilitating (external) growth
  • Collection of valuable data would allow us to fine tune our systems and improve
  • A central place to track all the discussions which ever took place within our premise and then refer to them at a later date

The introduction of the proposed tools stands to bring multifaceted benefits to the Apecoin ecosystem. Enhanced user engagement and streamlined operations will catalyse productive collaborations among community members. The advanced communication infrastructure, including the unified content indexing system and automated audio transcription, will facilitate seamless information exchange, improving decision-making and knowledge sharing.

With the Ape Assistant powered by similarity search, users can access relevant data swiftly, boosting efficiency. The incorporation of Dynamic NFTs for tracking user performance metrics not only incentivises participation but also refines outreach efforts. Grant accountability ensures transparent fund utilization.

Overall, these tools forge a cohesive environment, nurturing growth, innovation, and adaptability within the Apecoin DAO ecosystem.

What does the future hold for us ?

This is the logical precursor to us having a Unified Content Index or as I like to call it “The Hive mind”.

Unified Content Indexing : Hive mind :brain:

Think of it as a central Library which houses all the discussions which are held. The content inside which would be sorted accordingly and would be searchable using our search engine.

By keeping all Transcriptions and forum discussions and specific discord conversations in a single place one would be able to search through the content library for relevant discussions and then be able to refer to that particular discussion with the backlink for the forum discussion, discord thread or the audio transcription (space or assembly call) and the timings of that conversation in the future.

Effectively tracking how an Idea evolves overtime.

All this information is stored in a vector database according to its semantic meaning.

This would then Further Tie into Ongoing Forum Discussions (forum.apecoin.com), Documentation (apecoin.com) and Transcriptions (Audio Transcriptions)

This Hive mind can then In turn Power our Next gen Ape Assistant (if we decide to go that way)

Relevant Forum link : 🧠 Unified Content Indexing : ACT stack 2.0

User Input and Feedback :

If you have any suggestions or scope for improvement, please let me know.

Questions →

What are the Major Languages we need our content translated to ?

What more can be done with this to make it better ?

What’s left for me to do ?

  • Translation : To bridge the linguistic gaps and circumvent the communication barriers

It’s implementation wasn’t upto the mark, so It’s been benched for a bit.

  • Frontend : Providing the user (and their audience) a more convenient way to display and access all of this information

Seamless Integration between the Framer / Webflow CMS and Data generated from this script which we’ve automated to populate its respective position in airtable using custom code.

Help Required :

Currently, I need help with these things :

  • CMS guy for Framer
  • Helper to integrate data from Airtable into the aforementioned Framer CMS

Any help is appreciated, if you think you can help me in any way, let me know down below.

Timeline :

Rollout to creators will begin sometime around EoM (end of month) or in January. Have end sem practicals and Exams throughout december so I’ll be taking some time off for that.

We’ll rollout this functionality in phases, before which we’ll demo this to a couple of people and projects within our fold so that we can gain valuable insights to iterate upon. Then we keep on improving and expand :saluting_face:

Closing Statement :

Looking forward to gathering all the feedback the DAO members are giving me and the improvements they’re suggesting so I can make those proposed changes and make it even better.

See you guys tomorrow onwards, more regularly now fr fr :x::billed_cap: no backsies.

Hi @CEOofWeb3.0,

The community feedback period for your proposal would be ending in less than 24 hours.

  • If you’re content with the feedback received, your next steps are to finalize your proposal using the AIP Draft Template.

  • A moderator will reach out to the author to finalize the AIP Draft. Upon receipt of the final Draft, we will review and provide instructions on the next steps.

  • Are you ready to proceed to the next phase or do you wish to extend community discussion for another 7 days?

We look forward to hearing from you.

-@Facilitators

1 Like

Hi ApeCoin DAO Community,

@CEOofWeb3.0 has requested to extend the community discussion period for this AIP idea. This topic will automatically close a further 6 days from now. We encourage the community to continue to engage in thoughtful discussions through constructive criticism, honest feedback, and helpful suggestions.

Follow this Topic as further updates will be posted here in the comments.

-@Facilitators

1 Like

Amazing idea, but doesn’t suit well with me

2 Likes

Thanks for the feedback, might I know why?

1 Like

Update 1 :

Added functionality for the twitter side of things, now can archive multiple twitter spaces at the same time, although it takes significantly more time in comparison

It takes around 20-25% of the audio file’s duration to download it and then convert it to a suitable format

x-arc functionality has now been validated for use @scale

The spaces link scraping for twitter can be done by two methods, either by Twitter’s API which costs 100$ a month or by scraping links manually which is significantly more complicated as it also includes automated authentication and bypassing the anti scraping contingencies Twitter’s put in place to combat just that.

Don’t have access to the Twitter API basic version and currently debugging the link scraping using selenium method. It’s working well for @0xSword ’s use case of scraping replies but glitching for my use case of scraping links, some issue with scraping dynamically loaded content, will have to troubleshoot later

Lemme know if anyone can share access foe a sec so I can test out the

1 Like

Hi @CEOofWeb3.0,

The community feedback period for your proposal would be ending in roughly 24 hours.

  • If you’re content with the feedback received, your next steps are to finalize your proposal using the AIP Draft Template.

  • A moderator will reach out to the author to finalize the AIP Draft. Upon receipt of the final Draft, we will review and provide instructions on the next steps.

  • Are you ready to proceed to the next phase or do you wish to extend community discussion for another 7 days?

We look forward to hearing from you.

-@Facilitators

This topic was automatically closed after 5 days. New replies are no longer allowed.

Hi ApeCoin DAO Community,

@CEOofWeb3.0 has requested to withdraw their application. This AIP will be moved to and remain in the General category, as per the author’s request.

Kind Regards,

@Facilitators