YouTube is experimenting with server-side ads

If they're predictable with the timing and length then sponsorblock will still work.

I think it's more like extracting raisins? ad contents are still separate from the dough. finding the boundary conditions or ads hashes is guaranteed to work. whether it is feasible for adblockers is a different matter yet.

It's not that simple. Right now it's still separate video streams but presented as if they were the actual video, put in a queue of sorts.

Ublock Origin released a script to block them yesterday btw

That's why I said "What took them so long?"

Adapt to what?

I don't know, man. I hope they succeed. If they don't, then I will never visit YouTube again.

Some other frontend that would allow me to fast-forward them would be fine, though.

Even if adblockers aren't able to remove the ads, I'm sure they can still make it so you can skip over them with the arrow keys or video timeline.

Google is required by EU law to show what is an ad and what isn't. Adblockers could somehow detect that and skip forward.

I’ve never understood why they haven’t just provided a method of doing this for all their customers. Like a Google Ad service that meshes together everything on the page with the ads server side, so it’s harder to target them client side.

The value that Google has always provided as an ad platform is that they're targeted ads. You can target estimated age, geographic location, gender, estimated income. You can target your ads so narrowly that only a single person ever gets them even.

To bake ads into the actual content stream they have to expend compute editing and re-rendering the video for as many times as they have ads that they intend to run on those videos. They can do it once with once batch of ads but then it's only as targeted as who clicks on that video. Realistically they'll want to do it 5x, 10x or more per video (and store every copie of the video, unless there's some tech to store it as segments and seamlessly stitch then together as a single stream) to continue targeting the ads which gets very expensive fast

You need to kill the market first, if you make ad riddled shit first, no one uses your system. Now there is no real competition, which means they will monetize their position. It is what corporations do. We need alternatives, and I know Fediverse has some.

Introducing our new surveillance-based, dynamically generated, native sponsored video ads with mandatory interactive minigame engagement.

Careful - if we ever detect evasion, that's a lifetime IP ban.

I didn't say it would be easy. But yup.

It doesn't require any compute on the server at all.

Index the content of the ads, identify it, and drop that data from the served video file? There may be a more clever solution, but that'd definitely work. It should be possible to checksum or just straight up store the data for the first couple of kilobytes of video data that would uniquely identify each ad.

Youtube obviously must have a rota of however many ads which they can display, so eventually they'd all get identified although you'd be playing whack-a-mole forever as they release new ones. Isn't Sponsorblock partially crowdsourced anyway?

This would be challenging and fairly expensive, but worth it if you were motivated by sufficient spite.

Don't ads need to be legally highlighted as being ads in many countries?

Would make detection easy.

Would love to see AI ad recognition thrown back in their faces.

There will always be a way

Image Recognition could attatch the first frame of an ad to the length of time the ad plays for, then add it to an online DB a la sponserblock.

They might try to block seeking during these sections, but YouTube usually has raw mp4 streams available under the hood. You can even pull them using invidious or newpipe. Take that out and we might be fucked.

Train an AI to detect ads and voila

In the extremely rare event that I watch a youtube video on a my phone, and an ad comes on, I mute sound and literally turn my head away. Advertisers can't do shit about that lol.

Two ads before the video starts, which was a movie trailer.

"To watch this ad you first must watch two ads!"

This specific example is one thing that self hosting is arguably better for. I’ve made a few shitposting memes and the like that are five seconds long and uploaded unlisted just to share with friends that get immediately flagged and banned for DMCA that I’ve taken to just self hosting them. They’re getting like three views anyway because the world was never meant to see them.

People sharing videos with friends and family seems like a problem that’s already solved, if you really don’t want to use YouTube. Big channels that get millions of views are the real issue.

It's not just file size either. Video basically has several different things going on, where improving on one aspect tends to require compromise on the others:

• Resolution
• Frame rate
• Quality
• Bit rate (file size)
• Encoding complexity
• Decoding complexity (which affects battery life of mobile devices viewing the content)
• Robustness for dropped or corrupted data

Over time, the standards improve, but generally benefit from specialized hardware for decoding (thus making decoding complexity a bit more complicated when serving a lot of people with different hardware).

Netflix, for example, serves a small number of very large files to many, many people on demand. That means they benefit from high encoding complexity, even if it shaves off a tiny bit of file size, because spending a few extra hours on encoding a movie that's 10mb smaller is worth it if 10 million people watch that movie, as that's 100 terabytes of traffic saved.

But YouTube/Facebook and the others with a lot of user-submitted video, they're ingesting hundreds of hours of content every minute, chopping it up into like 5 different resolutions/quality levels.

Then YouTube has a shitload of processes for determining which video gets which treatment. A random upload of a kid's birthday party might get a few hundred views at most, so YouTube cares less about file size and more about saving that computational complexity up front. But if a video hits 1000 views in a few minutes, that means it's on the cusp of going viral, and it might be worth re-encoding with the high cost encodings that save space/bandwidth.

If a service doesn't scale, it won't be necessary to have that kind of complexity in the service. But those videos will load a bit slower, use a little more battery and bandwidth to watch, be more prone to skipping/distortion, etc.

Video is hard. User submitted video is harder. Especially at scale.

An engineer who needs a line item on their CV to get promoted.

Seriously though, 1080p is not a lot if you're on a big monitor or TV. At 720p you can start pixel counting on some displays

Your equating the software development with the running costs.

People have made OS and people have made YouTube alternatives. But that's nothing compared to the quantity of servers, networking infrastructure, storage, power usage, and labor to maintain and update it.

P2p isn't a valid alternative because that's just shifting costs onto your users. Just because a central entity isn't taking on the burden of cost doesn't mean the cost isn't there.

Pictures and text are rather low usage, both in storage and networking but video isn't. Especially when millions are watching videos at the same time.

Not downvoting, but I just think you're way too optimistic. It's like believing we, humans, could stop fighting wars. Sure, theoretically. But the difference between theory and the practical is that in theory there's no difference.

For the idiots downvoting explain why, or I’ll just believe you’re YouTube shills

Fuck you.

Storage and maintenance. OSes are miniscule in comparison to the data YouTube stores, we're in the multiple exabyte range here. Someone's got to pay for it somewhere. Floatplane might be a decent comparison as to what a FOSS YouTube might look like - they have a dedicated dev team and charge per channel to view, following more than a couple of creators would become cost prohibitive for me personally.

You absolutely need storage in a P2P network, the data doesn't just magic into existence, not only that but if there are insufficient peers in the network then you're not watching the video, smaller creators and older content would likely suffer as a result.

peertube uses webtorrents. it's viable. it works. owncast is fully self-hosted. it works. all the people downvoting are repeating a talking point, and have never implemented these projects.

We can barely keep Mastodon / Lemmy instnaces floating that host text, gifs and pictures.

That doesn't include paying the content creators.

Just because you're getting shit for free, doesn't mean that other people will want to do it for you for free.

Fuck you.

Someone could say the same for an OS, but we have many open source alternatives.

An OS requires significantly less resources. The only online features you need for an OS is a website to market the OS and host ISO's. Then you need a server to distribute packages to users. Packages which are significantly smaller then HD or 4K videos

Considering I remember some project in the past tried something like that in the past and found that because you can't control when people log off you can't guarantee files will transfer in one piece not to mention how expensive it was having everyone's computer constantly using Internet and computing stuff. For that reason I think the main problem here is that we are trying to centralize video sharing onto one platform. I instead propose we encourage people to make their own platforms. Like if you want to watch idk PewDiePie you go to PewDiePie.com and encourage people to explore the Internet instead of just sitting on one site. I suppose as a step in the right direction I propose that we get people to make online data bases using laptops/desktops that have nothing but xzamp and the videos you wish to upload to the web. Then we all collectively promote a sort of aggregation site that promotes everyones videos that way the aggregation site only has to store a bunch of hyperlinks and handle all the traffic while you the content creator just have to handle the traffic your content generates now the only challenge is making this idea profitable because if content creators can't profit few if any will make content and if the aggregation platform can't break eaven then we are back to square one of no one knowing where to look for content.

a) I downvoted you b) I am a YouTube shill c) Fuck you

that means I get money from them, right? I'm still waiting for my check. YouTube is the best.

I can't point to any one hobby that truly picked up my reddit time, but I do feel like everything I'm doing in its place is more productive

You do know physical books exist, dont you? :) I did not say ebook.

Not OP, but I would definitely pay for premium if they offered a lower cost version that was only ad-free YouTube. But I won't pay when they justify the higher cost with forced bundling of other services I am not interested in and have no use for, e.g., YouTube Music.

$13.99/mo is pretty steep, and realistically I'd have to get it for my wife too which effectively doubles the price and would make it the most expensive streaming service I've ever subscribed to (behind SeriusXM which at least has to finance literal satellites in space and delivers me radio when I'm in dead zones with no cell towers). More than my budget right now will allow for sure (I just cancelled every subscription after rechecking my budget)

Nebula has most of the content creators I would pay money to support, and more of that money would be going back to them.

Similar reason as people moving from Spotify to Tidal. The creators get paid more per view on Nebula than on YouTube.

Besides, I imagine there's quite an overlap of people that watch the type of content that goes up on Nebula and the people that are willing to pay for the content.

Nebula is only cheaper if your currency is the USD

Nebula isn't supported by ads AFAIK though

Kind of bootstrap paradox

I just went and looked to see if peertube is remotely viable.. technically seems working. I found an app on f-droid, got on a bigger instance (1k users seems about the biggest). Videos load and play. There isn't much content at all. A real shame. So yeah we don't have an option

I upload my videos there. Started off because professor wanted us to record ourselves then I just uploaded whatever. It's not much but it's honest work.

PeerTube doesn't have the allure given by the chance of getting paid for what you upload

Not because they couldn't but because they willfully operated at a loss to quench potential competition. The reason there is no replacement on YT is that all the content is on YT and creators won't shift to other platforms because their whole audience is on YT as well.

YT is not a video sharing platform anymore, it's a market. And that's why it sucks so bad.

There's no way to know, Google doesn't report YouTube profit separately in their financial statements. The closest department is "Google services" which does have a 34% profit margin.

And this will be more expensive to implement than regular ads, thus margins will be even thinner...

I could sit here and throw out all the bad shit YouTube's done but quite honestly the fact that its owned by google is enough justification to not give them money lmao.

comskip.exe does it well on free to air TV, but I suspect the methods it uses might not work so well for Sponsorblock etc. That said, maybe a hash can be made of the video every ten seconds, and when the playback hash differs, skip that ten second block. Computationally intensive I suspect, but might work for embedded ads.

There are plenty of open source ai, especially these single purpose one.

You could probably train something like that on semi-reasonable consumer hardware. Ads often have a very distinctive style and tone, and you need only a single output - the probability of it being any given second being an ad. It would probably take a lot to run though, you better hope the people who install the extension have good PCs. And, it would probably never get 100% accurate, you'd have to put up with still seeing some ads and having to rewind when it skips over valid video.

Of course, I use uBlock Origin.

But sometimes, I'm on a fresh install or on a friend's computer and I forget that I don't have an ad blocker.

Mostly just hide video and mute audio until it's over. Alternate player for twitch is the name of the extension.

I use video swap new from the GitHub the other commenter posted.

Not that I've really found.

It's like lemmy before may last year - just weird

there are alternative apps to access peer tube instances, like for lemmy. even newpipe has support for one of the bigger ones

I agree in that its probably not the right time in 2024.

I also don't think bittorrent is the right tech.

That said, as yt becomes more toxic there's more demand for alternatives.

It's certainly possible (e.g. take a hash of the first few frames of the ad and you can detect it pretty much anywhere and cut the right amount out of the stream).

But it's a lot more involved than just hiding an element on a webpage or blocking the same bit of JS every time.

And while I can see ways to automate it (take two streams for different users, compare differences, etc), it will likely end up being quite intensive on resources.

The only long term solution is "stop using Youtube". We need some fediverse style P2P replacement, where we pay for the videos with our outgoing bandwidth, and we're not there yet. Being a trillion dollar corporation sure does give you a lot more options in how you host things.

and twitch is still adblockable. it does look harder though, it bugs out sometimes. i'm sure there will be a great solution, there always is.

and fuck google while i'm at it, adblocking is only growing because of how egregious ads are getting.

Displaying ads on or within - definitely

Modifying content and distributing the modified content? That's a trickier one.

Not technically true - the movie reel itself wasn't altered.

It was swapped out for ads, and the same is true for digital formats. Here, they'd be actively modifying and distributing a modified file.

With Avidemux, cutting videos is super easy

Yes. What I'm also wondering though, why hasn't Youtube started injecting ads directly into the video feed earlier?

they can show an ad-specific timeline or even disable playback controls while the ad is visible.

Which adblockers could detect (at least in browsers not infected with manifest v3) and use to skip the ads. 🤷‍♂️

If it's impossible to fast forward the ads, that means the timestamps of the ads have to be send to the browser, so adblock should be able to use that data.

Not necessarily true. Premium subscribers are worth a ton more to content creators because they get a cut of the premium price for every premium view. And it's not insignificant compared to a non-premium viewer.

For a non-premium viewers to provide value to a sponsorship they need to either use an affiliate link and purchase, or at least look. Creators often don't get paid for viewers that don't click the link. And only a small percent of the users is going to do that. And even non-premium viewers already use ad and sponsor block.

Don't be confused - content creators would likely benefit greatly from a higher percentage of premium subscribers on their videos. They are guaranteed income, and sponsors are often a lot more volatile.

This is a great point, not sure why you're downvoted. Something else I thought of is the fact that YouTube is very good adjusting streaming quality on the fly, meaning unless everything is pre-cached it would be difficult to stream the way most podcast apps can "stream" audio by essentially downloading it faster than it can be listened to.