this post was submitted on 02 Dec 2023
786 points (97.9% liked)
Technology
59414 readers
3442 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I have some background in tech but admit I'm a long way out of touch now. I wouldn't be at all surprised if they're working on back end stuff to have personalised ads "baked in". I know the resource implications of this are huge, but it still wouldn't surprise me.
The resource implications are the problem. The cost - in terms of compute time - to bake those ads eats into the profit earned from advertising as a whole. Since only a fraction of users adblock, they would probably lose more money than they gained.
They'll consider it once the compute cost inserting the ads is low enough that it's worth it. I have no idea if we've reached that point yet, but I'm guessing not, since otherwise they'd already be doing it.
Most of the formats served by YouTube are already chunked, which means they can easily insert different chunks of video (ads, etc) at various points in the stream by changing the manifest. This is trivial, computationally. The complexity lies in building the mechanisms to make it work.
The non-chunked formats are only used by older devices, and are lower quality. Those would require re-encoding to change, but few users see them anyway, and those users probably don't adblock.
Platforms can now insert ads directly into the manifest file into totally random timestamps. The file chunks' names follow the same pattern as the original video. You cannot filter or prepare for it. Probably that will be the future. (AWS MediaConvert can do this for example.)
And they only create the manifest file upon starting the stream so you can inject personalized ads too.
I guess we will have to compare the last video frame and/or audio sample of every segment to the first frame and/or sample of the next segment or something like that?! Maybe the effects of "the loudness war" in ads will help us detect ads solely by the loudness change within the audiostream?