Excellent counterexample! I have been using SponsorBlock and DeArrow for years and anecdotally I'm hearing more and more using SB.

[delayed]

> We have 17% of submitters every week being first time contributors

That doesn’t sound like necessarily a positive thing; it sounds like it could well be unsustainable churn (impossible to really tell from such limited numbers).

I have contributed in the past, but don’t any more because of how you went out of your way to make it tedious: https://github.com/ajayyy/SponsorBlock/issues/2106.

we also have 19% of constributors every week who have been submitting for over 3 years! I think that's a pretty impressive number, especially when it is so easy to reset your account by reinstalling the extension because we don't use traditional emails and passwords.

I think my Firefox profile is from twelve or thirteen years and three laptops ago. It was even used on Windows for a few years in the middle. Just copied the directory wholesale.

This is the most optimistic way to look at it. People who couldn't code before but had ideas can code now, can build things, can contribute to projects they had no business understanding before LLMs.

What I like about SponsorBlock is the statistics it presents. Every sponsor segment I've contributed has saved 682 people from seeing it. Per each submission I've contributed by spending maybe 1.5 minutes, I've saved 63K person-minutes (44 person-days). Wow. What an impact.

I didn't know about the stats, I just checked and found it's saved me over 3 days. I assume that's local to this machine, so that's just in the last year.

That is an incredible amount of YouTube or the sponsorships are way longer than I remember.

[delayed]

I'm convinced SponsorBlock could be completely automated with modern TTS and a fairly small LLM

Any are welcome, so far as I know having not checked their submission criteria, to spin it up and contribute tags to the SponsorBlock community. I'd wager there are people already doing so.

I think it would be more useful to more people to utilize the existing solution because it runs in a very small footprint for the end user. Most people wouldn't be able to run an entire TTS/LLM workflow without processing videos in the background or waiting some multiple of real-time in order to watch a video.

You could maybe do it with just LLM tagging the transcript, but that's still a ton of processing just to watch a video.

God that'd be so much wasted energy if every person wanting to skip in-video ads was running an entire LLM to do so.

Unavoidable for some stuff like podcasts where they inject random sponsorship into audio files and throws off all the timestamps. Though maybe wasteful llm adblock compute breaks ad serving compute, they somehow cancel each other out by making all the compute wasted on ads + telemetry defunct.

Circle back to original topic, eventually having local LLMs intercepting EVERYTHING in background to reproject a curated personal bubble filter feels like wasteful but sensible end game. Not just to hide ads / skip sponsors, run every image through anime filter, live editing xyz, if there'es some centralized DB to grab timestamps do it, but otherwise we could be burning tokens none stop to modify reality, and some add manipulation is just rounding error that comes along for the ride. I click a video, background processes play opens 10 instances of the video at 4x speed, processes, gives me version of video without ads, sponsorblocs, xyz content filtered out... or just generate video to conform to my biases because it makes me feel good.

It's not sponsorblock quality, but I have trained a model for podcasts. Turns out, at least for podcasts, you can get away with a surprisingly small model, small enough it can run on device directly.

No idea if it's good enough for all podcasts that are out there. It works for me and the podcasts I listen to, and I have tried testing it against a larger corpus, but I'm one developer and I just wanted to solve a problem for me.

https://github.com/Q726kbXuN/podcast-ad-tagger

Don't media plays already kind of do this with 'skip credits' or 'skip intro' ?

Thanks for the share. I read the slides and on the non technical side like the clipart

>SponsorBlock

I am routinely amazed at the speed and breadth of the SponsorBlock community.

Popular channels are tagged within minutes of posting. Obscure channels almost always get tags at some point.

The only place I've noticed significant lack of coverage is really long older videos, like say MIT OpenCourseWare. Not a complaint at all, not expecting anyone to sit through hundreds of hours of college to mark up a few tangents here and there, just an observation.

Lol, stack overflow actively scared people away by being toxic.

Huge fan of DeArrow; thank you for making it!

link to https://dearrow.ajay.app/payment#link=chrome is wrong on https://dearrow.ajay.app/

Thank you for these projects, keep making them!

However (and I haven't seen your talk) can't you agree that there is a difference between crowd-sourcing wikipedia and clicking on a marker for a sponsor?