> Copyright claims is a perfect example.

Yep.

One of my favorite subscribers to watch is some guy (Harald Baldr[0]) who travels around remote areas of the world and vlogs it.

But he's said a number of times how even 10 seconds of background music can cause his video to get demonetized and then all ad revenue for the video goes to the record label who owns the copyright and there's absolutely nothing he can do to combat it.

This happens even if his video is 30 minutes long and has nothing to do with the music at all -- it just so happens he's walking past a store or hotel lobby that has music on.

[0]: https://www.youtube.com/channel/UCKr68ZJ4vv6VloNdnS2hjhA/vid...

They make it hard to correct those things too. You have to re-upload the entire thing as a new video instead of being able to edit or replace existing videos to deal with copyright claims. But then their algorithms also seem to demote content creators from deleting or unpublishing videos so correcting those things actually hurts your channel in the long run.

One of my more popular videos could never be monetized because I added a song (via the old YouTube editor interface) that turned out to have a copyright claim but the interface doesn't allow me to remove or change the background music in post. So I'm given the option to delete something with 911k views and annoy my viewers with a duplicate upload or just cut my losses.

i thought they have the option of muting audio now?

people are downvoting a question?

so i looked myself. seems like there is that option now. they also have an option to just "mute" the infringed song but it's in beta

https://support.google.com/youtube/answer/2902117?hl=en

Upvoted you to compensate :) Also, thanks for sharing the link

Comments like this one are outside community guidelines. You really shouldn’t be begging for votes on here.

That's not a question. That's a statement with a question mark at the end. Also, the statement ignores the fact that the parent comment asserted edits have no impact on demonitization, so even if it was correct it would be moot.

In many languages the question mark can be the only difference on print between interrogative and declarative sentence, and as my native language is one of them, I hope that English will develop in that direction.

English can work that way too, converting statements to questions with inflection or the addition of a question mark. e.g. "You understand?"

However, in this case, interpreting this as question rather than a statement leads us to see a question about the inner thoughts of the commenter, which clearly is not what was intended.

What I believe was intended was to make a statement that was to be understood as uncertain. However, the "I think" prefix is all that is necessary for that and the addition of a question mark just creates uncertainty about the it being interrogative/declarative.

I interpreted it as 'upspeak'

YouTube are actually adding some tech to help with this. In future when making a claim you have to add the timestamps in the video where the copyright occurs. Then the creator will get that info as well as a few simple quick-fix options to cut out that portion of the video or replace it.

It's long overdue but they do seem to be making some strides to help creators at last.

Interesting, is there a source about these changes?

Yes, but will it detect the claims and allow you to fix them without stealing all your revenue?

Yes, but subject to claimant's approval if you did not remove the offending sections.

A part of me feels like claimants will find a way of gaming this and making their approval a series of endlessly nudged goalposts...given the way IP holders are already taking advantage of the tools already at their disposal.

No, two rounds max; both parties lose (suspect's video deleted and claimant gets a strike) it is not resolved in two rounds.

So the claimant still, invariably get what they want, which is a pulled/not published video? That doesn’t sound like both parties losing to me.

You seem to have forgot the 3 strike rule.

No I really didn’t-could do without the implication, why don’t we and leave the personal jabs about what I forgot or didn’t out of this and stick to the issue.

I don’t believe as designed that even three strikes does well at (pardon the pun) striking a balance between content creators and content ID claimants. YouTube’s own behavior, leading up to where we sit now, and the behavior of claimants gives me no reason to give either party the benefit of the doubt or faith in how they purport to be handling content claims.

I forgot nothing, friend.

I don't understand why they won't put the money in escrow instead of just shunting it to the claimant. Presumably because it would require people to make a determination. Seems irresponsible.

It's not really a DMCA so this supposed legal process is really just made up by YT who don't have any obligations

Along with some other reasons, the fact that they have implemented a system which effectively strips DMCA protections from content uploaders should result in there being no safe harbor provisions in effect for them.

An internet service provider can impose any rules they want unless a client can produce a contract saying otherwise. Safe harbors are an indemnification against being sued for contributory infringement. That doesn't obligate them to host non-infringing content.

Given that the DMCA is not a fundamental property of the universe, rather it's legislation. That legislation was written in a way to give both copyright holders, people who want to put content on the internet, and service providers a framework for dealing with copyright. If that framework isn't working out for all the interested parties, then it can be changed. When the framework isn't doing it's job, it should be changed. And in my opinion, part of that is that there should be no safe harbor for:

* content that is upload by users who cannot be affirmatively identified. * content that is uploaded by users who are not within the jurisdiction of the united states. * if a service provider implements mechanisms that defeat the intent of the law

Ultimately, yes, I get it that no one is required to host others' content. But at the same time we're not required to provide get out of jail cards.

> * content that is uploaded by users who are not within the jurisdiction of the united states

Really? And how do you think that works out for the millions of content creators in Europe, Asia, Australasia, Africa and South America?

I think it works out that Google spends more time ensuring that the content they import into the US and redistribute is something they can legally import and redistribute.

I think that's what they do already. You just have to make sure you dispute the claim within 5 days.

https://support.google.com/youtube/answer/7000961?hl=en

Oh that's good! Didn't realize it.

Doesn't quite solve the spam problem, but seems like a step in the right direction.

Nope, the problem is that disputes can be (and routinely are, by default) rejected by the claimants with no risks to them, again. Escalating beyond that risks a copyright strike (three of which mean your channel is gone, but even a single strike costs you access to various features).

The best approach is to only escalate if you understand the specific legal situation of your video and can make a case that would likely hold up in court. Needless to say most creators aren't well-versed in legalese and the risks to the creator are much greater than the risks to the claimant so most creators avoid this for "less important" videos.

Here's a nice explanation from a channel that has experience with being taken down and receiving strikes: https://www.youtube.com/watch?v=HybK82zw4us

They have a general assumption that claimants (or rather, Big claimants) are in the right by default; even when a big claimant is wrong, it's better to be on their good side and at worst lose favor with a smalltime content creator

Not to mention that that's not a copyright violation if it's incidental and under 1 minute or so.

Not quite true. Incidentals under Fair use are entirely dependant on the context of the created content (at least for the UK).

News/educational - likely okay as incidental.

Everything else is up to the courts to decide.

Under this example, I’d argue that this is not incidental usage. The audio could be considered integral to the “scene” that is being recorded. It gives the audience an idea of what place x looks, feels and sounds like. And this is certainly not news/educational content.

> Under this example, I’d argue that this is not incidental usage. The audio could be considered integral to the “scene” that is being recorded. It gives the audience an idea of what place x looks, feels and sounds like. And this is certainly not news/educational content.

But if you watched his videos this isn't the case.

It's him walking through some public outdoor market in Sri Lanka and one of the vendors happens to have a radio on which is captured by his camera's microphone as he walks through while narrating what's going on. He didn't go out of his way to capture the audio separately and then overlay it afterwards. It's background noise that was recorded in the same audio stream he uses for his voice's microphone, which makes it pretty much impossible to cut out later short of muting all of the audio or making artificial cuts to remove the content.

In any case, it's a shitty situation for both the content creator and the viewer.

How it was recorded has nothing to do with copyright law (in this case, mechanical rights are a whole thing).

A musical work was publicly performed (via the radio). He recorded that musical performance (making a copy of the musical work in the process) and then distributed that recording of it to a wider audience. Each time someone watches that video, that is a public performance of that musical work.

Whether there was intent or not, it doesn’t matter. How it was recorded doesn’t matter.

Copyright law doesn’t care about intent. It cares whether a musical work has been performed or copied. Both have happened in this case.

And it can be argued that it does not fall under news programming (he’s not a journalist breaking a story) nor educational (he’s not teaching anything). Therefore it will struggle to be considered fair use.

This is actually copyright law finally being applied correctly on the internet. Its just taken years for it to actually happen on platforms like YouTube.

I agree with recording by accident or on purpose should not be a deciding factor in anything but...

The problem is 10 seconds or even a minute of a recorded song shouldn't mean a record label or the original music owner can swoop in and reap 100% of the ad revenue (minus Google's cut) for a 1 hour video you recorded which has ads relevant to your content (not the song).

By the way, you could argue he is educating because he often visits historical sites and gives some back story on it. You could also argue he is a journalist because he has a running series called "Harald invades" where he'll go into some family's house (with permission of course) and learn about how they live and work (aka. it's a story). Actually most of his videos feel like stories, since it's not over edited garbage with a spin for profit. He just goes through towns and documents the process raw. I would say he is one of the best journalists I've ever seen.

I agree with the 10 seconds problem. But that’s a problem with Youtube’s implementation - nothing to do with fair use, incidentals or copyright law.

Edit: having replied to another comment, I think I’ve realised something. IIRC The original YouTube license deal with PRS meant that they only sent over video level data - no cue sheets of music tracks etc. They specifically wanted this in their license. So their current implementation might be a remnant of that.

News and educational content actually have quite narrow scope. And those episodes/segments probably wouldn’t fall under that scope.

And then you have to take into account the overall purpose of the channel. Which is not one of news/education.

People watch it for entertainment much like the old Michael Palin travel documentaries. None of which are considered news or educational.

Just playing the devil's advocate, "Fair Use" could be an interesting question here...I think it would fail the "academic use" rule, but worth consideration, if the criteria (from standford.edu) is:

1. The least amount of copyright material as possible should be used.

2. "Fair use" work must have significant new and unique material added (not be a compilation).

3. "Fair use" work must not harm future potential markets for the copyright work. (ex: not a highlight video)

4. Work must be either a parody, criticism, review, or "academic use" to qualify for "fair use".

If we're talking a 10s clip of audio where the original is significantly longer, I think the most significant question is whether the work could qualify under the legal term category of "academic/educational". A work can only be considered "academic/educational" if it meets all of the following (also from stanford.edu):

1. Noncommercial instruction or curriculum-based teaching by educators to students at nonprofit educational institutions.

2. Planned noncommercial study or investigation directed toward making a contribution to a field of knowledge.

3. Presentation of research findings at noncommercial peer conferences, workshops, or seminars.

I don't know the legal muster required to meet this, but from what I've read, this is where "almost all" youtube videos are going to be disqualified, especially by the intent of the rule, which is to provide an out for teachers/instructors and students.

The whole argument is rendered null by the fact that youtube has to comply with the DMCA, which requires that work be taken down if it contains work created by other people (clips, background music, photos), though.

Also youtube seems to have a fairly flexible amount of power here, can take down pretty much any content it wants, and if it chooses to side with the copyright side by default, they have the power to make that consideration.

My opinion is that if you're going to be creating content, and advertising and/or monetizing them, you really shouldn't have any copyright work in there. Saying "it's only 10 seconds of the work" may provide some legal footing for the "must have significant new work" rule, but it seems like you're just drawing an arbitrary line in the sand and saying your side is okay, whereas youtube owns both sides and the whole beach.

So to take this to its ultimate conclusion: These copyright holding corporations are just blindly optimizing for maximizing their grasp on any content whatsoever. If there was any more subtlety to it they wouldn't claim a 10 second bit of accidental background radio, that the maker of the video never wanted or asked for. I can't wait until they realize they can build really large speakers and be able to "audio bomb" and lay claim on anything recorded, anywhere ...

> These copyright holding corporations are just blindly optimizing for maximizing their grasp on any content whatsoever.

This is what the reality is:

These copyright holding corporations are just optimizing for maximizing their grasp on their own recordings within any content whatsoever.

They legally own the recordings. What happens with the recordings is ultimately up to them. If they want to be paid for the use of their recordings, it’s their choice and they are free to do so.

> the maker of the video never wanted or asked for

Let’s flip this on its head... Did the rightsholders ask for him to be recording at a specific moment in time when their content was playing?

No. He made that choice. He chose to record that musical work.

He could have done plenty of things to avoid a recording a copyrighted work.

A better solution might be mechanical licenses calibrated to the usage and revenues.

30-minute (1800s) video plays an 18s of a somg? Mechanical right by song's author to 1% of monetisation.

Used in samling or simultaneously with other copyrighted works? Mechanical split among works.

The percentage might be tweaked -- vieo holder splits 50-50, so songwriter gets 0.5% rather than 1%, in this hypo.

This is actually similar to how it works for traditional TV and Radio Broadcast in copyright societies:

(Total revenue of license period x peak/non peak weighting x duration of song played) / total music duration for license period.

This is something YouTube vehemently fought against. At least in the original 2014 contract with PRS for music.

IIRC YouTube would never send cue sheets (list of included tracks per video) in the usage reporting data. It was just the video title. That was the old deal though - I left before the new deal was signed with ICE services.

Edit: oh, and sampling has its own process. The content creator has to negotiate with each individual rights holder as to the splits when the song is registered.

I don't think it's fair if copyright holders can just inject their audio by having it played in the background somewhere as you are recording (like recording in the vicinity of a bar, which is what one of the commenters talked about), and then get to monetize that.

Which is exactly what these corporations are doing when they buy air time on the radio: part of it is definitely marketing strategy to have that audio be in the background or present in as much places as possible, purely for exposure.

It's something they are actively injecting in our shared/public spaces, basically making ambient sounds "copyrighted", and it's kinda ridiculous they expect to be paid when successful.

I know, this is how copyrights works. But it's a relatively new thing they actually take action when something is recorded that has unwanted background sounds, that the video maker unwittingly recorded, while the copyright holder is explicitly paying to have that sound be playing in the background of as many venues as possible, with the purpose to expose it t as many people as possible.

Also this crazy idea is unheard of (haha pun) in other kinds of media: If one of my graphic designs would be shortly visible in the background of a camera pan ... can you imagine trying to claim even 0.1% monetisation?

Having recently been enlightened as to the incompetence of record labels when it comes to handling revenue splits in the simple model (e.g. they can't even get them right for Tier 1 artists with a simple split between authors), I'd not hold out much hope they'd be able to cope with a more complex model like this. And that's before you get to antiquated places like the PRS who would also need to be involved.

The other thing to remember is that media cartels don't want you to remember that Fair Use is a thing, so they try as hard as possible to make sure no automated system has a provision for it and to argue that there's no good standard to follow so you can only use it if you're willing to pay for lawyers and court time.

But in YouTube's case it is kind of moot since the system isn't designed with fair use in mind and is not technically copyright claims. It's entirely up to YouTube to decide what they do, and their default and only position is to side with the media companies with the expensive law firms over their actual content creators.

Actually fair use falls firmly on the side of copyright holders when a claim is made, until it is disputed in court.

The provision for it is to take it to court. That is the only option. There is nothing YT can do if a claim has been made.

So it’s right that there isn’t a counter claim system in place for fair use here.

I'm not sure why this is being downvoted because it's exactly correct from a the current legal position (as far as I understand it). An automated system can't reasonably be expected to establish, for example, whether copyright content is being used for the purposes of parody. Similarly, a copyright holder has every right to disagree with your determination of what is parody or transformative works, there aren't any simple rules to get around that.

Whether that is a sensible position for us to find ourselves in is another question.

> Under this example, I’d argue that this is not incidental usage.

This seems hard to buy - surely courts are aware that copyrighted material necessarily exists in public spaces? If this guy filmed himself walking around then presumably he filmed all kinds of illustrations on t-shirts, text on signs, people whistling to themselves, etc. Wouldn't the copyright owners of all those things be able to make the same claims as the owner of whatever was playing on the radio?

If the law works the way you're describing here and in replies, it sounds like every video ever made in a public place would be one long string of copyright violations from start to finish.

The problem is that he is making money of the back of the recording of the musical work that has been registered as copyrighted material. That’s literally the whole point of copyright. Someone is making money off my musical work? I deserve to be remunerated for its use. (Very high level description).

If it’s Someone whistling an on the fly melody then that’s fine. The musical work isn’t registered as being copyrighted. It’s just some bloke/lady whistling. So that doesn’t count and is totally fine.

If they were whistling a popular tune, well, that’s getting into the whole idea of arrangements. They’ve created a new arrangement of the musical work, which might be copyrighted. Depends on context with arrangements.

If this were a holiday video he shared with friends and family, it would totally fall under fair use. He’s not making money off the back of it.

As to all the other signs etc, no idea. My background is music copyright.

> The problem is that he is making money of the back of the recording of the musical work

I don't follow. Can any reasonable person really suppose that a 30 minute travel video's ad revenue might be materially affected by the presence or absence of a random ten second snippet of a random song at some random part of the video, briefly audible in the background (at presumably low quality)?

I'm no lawyer, but as stated the claim seems facially absurd.

The degree of material gains from including a copyrighted musical work don’t matter. It’s a blanket thing.

The copyrighted work has been included in content that is monetised. The guy is making money and the musical work has some part to play in that.

Here’s a counter example.

Imagine if I recorded Bono from U2 chatting about random stuff between playing live tracks at a concert. But I inadvertently caught the last 10/20 seconds of most songs. I make these recordings publicly available and monetise them.

I was only doing it to catch Bono’s musings between tracks. But people can hear the copyrighted material as well.

Under copyright law, I have made a copy of the musical works (mechanical right) and have made them available for consumption (performance right).

Just because I only caught snippets doesn’t matter. It might affect the amount of money I need to pay rightsholders, but there is copyrighted material in those recordings. So the rightsholders should be remunerated.

That’s it. That’s the whole fundamental point.

> The guy is making money and the musical work has some part to play in that

I understand that a lawyer for the music owner could conceivably argue that claim, and that a court could conceivably agree with it. I'm sure everyone in the thread agrees on that.

What I'm questioning is your presenting "he made money from the music" as a clear fact, and not something a court would decide - suggesting that the presence of audible music necessarily guarantees that theft has taken place, no matter how strong the arguments to the contrary. It's hard to believe the law on this famously contentious topic would be that simplistic.

He made money from his content. His content included a recording of a copyrighted work. Therefore he made money from that recording. It’s that simple.

He could have:

* scrubbed the audio

* removed the section including the musical work

* overlaid a royalty free musical work to replace the copyrighted one

* asked the market vendor to turn the radio off

* waited until the radio played a work that was not copyrighted

* etc etc

He’s chosen to include copyrighted material in his content by not doing any of that. Some of which would have been relatively simple to do.

In court I would think it would be difficult to prove there was nothing he could have done to remove the offending audio.

And he also falls foul of Fair Use (there’s another comment somewhere that digs into the definitions of educational content in this case) as this is not incidental usage.

If he falls foul of Fair Use, then regular copyright laws come in to play, which takes the simplistic binary viewpoints:

* Is there copyrighted music or not?

* Is money being made from the use of music or not?

Yes to both => remunerate rights holders.

I understand that you're trying to be constructive, but nothing here replies to anything in my comment.

> He made money from his content. His content included a recording of a copyrighted work. Therefore he made money from that recording. It’s that simple

That's your position; mine was that there are strong arguments against that claim (I listed several), and no apparent reason why a court wouldn't consider them.

(Note that I'm not claiming the arguments against your claim are correct - I haven't even watched the video. I'm saying that whether infringement occurred is not a simple matter that can be determined by algorithms, as you've presented it. It's a nontrivial legal question that a court would need to rule on.)

> He could have: [removed the audio various ways]

I don't know why you've argued this. Of course he could have removed the audio, but what does that have to do with fair use?

You’re trying to argue that just because the music happens to be there that it has no relevance to the fact he makes money or not. I get that.

Fair use applies for cases like holiday videos shared between family & friends (not the case here), and incidentals.

You’re trying to argue that this is incidental. And I see what you’re trying to get at. But unfortunately the definitions of what constitutes news and/or educational programming are usually pretty clear cut. And this does not fall into either of those criteria. And they are the only things protected under incidental usage. There’s a comment somewhere in the thread that details a specific definition of educational content. The author of that comment agrees it would not fall under that category.

Once you remove fair use, it becomes completely black and white.

He made money from the video. The video contains music. Rights holders must be paid.

That’s it. That’s the copyright law viewpoint once you eliminate fair use.

> I don't know why you've argued this. Of course he could have removed the audio, but what does that have to do with fair use?

We wouldn’t be having this discussion about his specific video if he’d taken action to remedy the problem before it became a problem.

If I were a lawyer for a rightsholder, I would make the argument that there were many options for him to avoid using my clients copyrighted material and he did none of them. He could have avoid this situation but chose not to.

What makes him so special? Why can’t he edit his video like every other media company?

Edit - this is why the incidental provision was created in the first place. News & educational content may not be able to edit out the audio without jeopardising the core of the content - e.g. a live news report where there’s cars driving by playing music in the background.

> You’re trying to argue that this is incidental

I specifically said I'm not.

> > I'm not claiming the arguments against your claim are correct - I haven't even watched the video. I'm saying that whether infringement occurred is not a simple matter that can be determined by algorithms

> [news and/or educational programming] are the only things protected under incidental usage.

In a casual search, every source I checked disagreed with this claim. Examples:

https://www.copyright.com/blog/music-licensing-fair-use/

> There are certainly people who argue – and there are credible arguments to be made – that the incidental picking up of the music in the background during a documentary film can qualify for fair use. When you’re talking about incidental music picked up, you’re more likely to qualify for fair use if you’re not focusing on that music, it plays for a short period, and it’s in the background.

https://cmsimpact.org/code/code-best-practices-fair-use-onli...

> Fair use protects the creative choices of video makers who seek their material in real life. Where a sound or image has been captured incidentally and without pre-arrangement, as part of an unstaged scene, it is permissible to use it, to a reasonable extent..

Ah. I think I know what the problem is here.

I’m based in the UK and it seems Fair Use applies somewhat different here thanks to the Copyrights act in 1988.

We have a fair dealing clause[0] here.

> Certain exceptions only apply if the use of the work is a ‘fair dealing’. For example, the exceptions relating to research and private study, criticism or review, or news reporting.

[0] https://www.gov.uk/guidance/exceptions-to-copyright

Fair dealing comes from the POV I’m trying to get across.

> > I'm not claiming the arguments against your claim are correct - I haven't even watched the video. I'm saying that whether infringement occurred is not a simple matter that can be determined by algorithms

Yeah I somewhat agree with this. I’m pretty sure I said so somewhere (it might have been a slightly different comment thread, apologies if I didn’t make that clear).

The existence of a copyrighted work can be detected by an algorithm (although the accuracy is debatable in some cases). Whether it falls under fair use cannot.

As stated, Fair Use is decided on a case by case basis. But it’s assumed the rights holder is correct in their claim until the court decides.

In any case, before a counterclaim is approved by the court or if the court sided with the rights holder, then black and white copyright law comes into play.

Still, if the content creator took action to avoid including potentially copyrighted material, like everyone else does, then we wouldn’t be having this discussion.

Edit - before you pick up on it, criticism/review is a specific critical review of the work in question. As that doesn’t apply here, news/educational fair dealing then becomes the question. And then all those points I’ve been trying to make come into play. Could he have done things differently? What was the purpose of the content? Etc etc.

> I’m based in the UK .. We have a fair dealing clause here.

I know, and I'm familiar with the distinction. I assumed you were as well, and would be using the term of art for whichever system you meant.

> The existence of a copyrighted work can be detected by an algorithm .. Whether it falls under fair use cannot.

Thanks for the information.

Also, from one of your linked articles.

> If you’re going to rely on fair use, the bottom line is, you need to factor in how much risk you’re willing to take. Why? Because some uses are riskier than others and the risk of a failed fair use defense is copyright infringement.

Some of those are pretty ridiculous expectations of any content creator.

Scrub the audio? I don't think he's going to go in there and literally separate and remove the song's frequency from his voice which are interweaved together. Something like that takes a tremendous amount of knowledge and time and always comes out sounding bad because any time you remove shared frequencies it'll make your voice sound very choppy and robotic (I know because I've done things like this). His job would slowly become being an audio engineer instead of a traveler.

Overlaid a royalty free musical work to replace the copyrighted one? Not possible, his voice and background audio is on the same track. To do that would essentially mean muting the entire track. This is the easiest option and often what he does. He just cuts it or mutes it, both of which makes his video content worse.

Asked the market vendor to turn the radio off? He often does this in closed stores where he plans to spend a bit of time in, but it's unrealistic to expect anyone to do this in an open outdoor market with ~200 vendors spread around everywhere for 1,000 feet in every direction. You're asking to basically shut down the whole market and control every audio source while you walk around. Not happening for a regular person.

Waited until the radio played a work that was not copyrighted? Not possible due to the above reason but this is also a little more unreasonable because there's no reasonable chance a regular person will be able to determine which songs are copyright and which are not, especially not in another country where they don't even know what song is playing. Also by the time he researches the song (assuming he somehow figured out how to even determine if it's copyright or not), chances are a different song would be playing and now he's stuck in a research loop for the rest of his life, or ends up in an another infinite loop waiting for a copyright free song.

What about overdubbing? That’s the trick the majority of the industry uses.

...

> Asked the market vendor to turn the radio off?

> Not happening for a regular person.

And here were get to the crux of the matter.

I defer you to Charlie Brooker’s excellent video on “how much making one episode of a TV show costs”... skip to 5:05 for the section on clearance.

https://youtu.be/LbrmaeiZ4RE

All this stuff I’ve talked about is looked at by teams of people in media companies because they have to exist within the scope of the the law.

And that’s the issue you seem to have. It’s unreasonable to expect one person to do 20 extra jobs.

Yes it is. Which is why media companies hire 100s of people. Because they have to comply with the law.

Just because this is one person doesn’t mean he is outside the scope of the law. He doesn’t get any special treatment for being an individual. He’s got to suck it up like the rest of us.

He sometimes overdubs it while saying he has to mute the original audio due to record labels.

That video seems like an advertisement to try and convince people not to record anything unless they are a TV station.

I'm happy we live in a day where someone like Harald can just point a camera around a place while talking and it ends up being more entertaining (to me at least) than anything running on TV.

> briefly audible in the background (at presumably low quality)?

Yep it's both low quality and low volume. It's hard to apply exact numbers but let's say speaking volume is an 8 but the background music is at a 3 or 4. It's present and loud enough to make it out but it's not crystal clear. Quality wise it suffers because it's not a direct audio input into the video. It's being picked up by a small positional mic from ~50+ feet away.

Just curious, do you have a source for that? I'd like to read more about it

Why blame YouTube? They're doing it at the request of the right holders, who would otherwise have a claim to pull the whole video down. They also provide an automated tool to scrub claimed music from already-uploaded videos (https://support.google.com/youtube/answer/2902117?hl=en) You might say the music rights holders lack empathy though...

Issuing a real legally binding DMCA takedown holds legal liability. Youtube id claims are not dmca takedown claims, and youtube is very clear on this distinction. You can not counter-sue a company for issuing a false id claim.

Issuing a DMCA takedown that is both false and meant in bad faith is an act of perjury. In theory the copyright holder could get jail time.

A interesting case is the video of a toddler dancing on youtube, 29s long, to a song. The Ninth Circuit held that copyright owners must consider fair use before issuing take-down notices. The case got settled afterwards.

This difference between DMCA takedown and ID claim is often discussed by youtube creators as both a boon and a curse. The good thing is that id claim does not impact any future standing with google or legal punishment beyond the loss of revenue from the video that get claimed. The bad side is that they get abused.

> Issuing a real legally binding DMCA takedown holds legal liability.

I've seen several documented cases of DMCA abuse but I haven't seen a single case where a major company was punished for it, and I doubt they'd face any consequences substantial enough to overshadow the profits they've made abusing the process. If you've got examples though I'd love to see them.

There can be a lot fewer cases if they don't have to go through the process!

It seems like there should be a way to escalate from an ID claim to DMCA -- to reject the ID claim as a content creator and force the claimant to send a DMCA. Then you can dispute the notice if it's issued in error, or counterclaim if they're abusing the process.

If the creator has deep enough pockets to fund a lawsuit, a tortious interference claim could be filed if someone falsely asserts a copyright violation and messes with your monetization.

> Issuing a DMCA takedown that is both false and meant in bad faith is an act of perjury.

But extreme negligence is 100% allowed, so that clause doesn't matter.

Thanks for the references. So basically the bargain of the ID claim is that if someone's use wasn't fair, the penalty is relatively lenient vs what it could've been in court, but then the disadvantage is that you don't get a proper appeal when you consider something to be fair use?

This is like DRM in a way: it's a circumvention of the copyright system that removes protections/mitigations from the side of "the people" (the demos).

Such actions should void the copyright on the claimed media, either make a DMCA or put up with it. The use of this "side-channel with a supposition of benefit to favour media corps and no proper defense review or resolution" perverts the copyright system's balance in providing a deal between the demos and the creators (and other rights acquirers).

Google's facilitation of this (instigation?) IMO shows they've come along way from 'don't be evil'.

Problem is that YouTube has automated the process to favor the fake claims filers.

Real case that happened to my friend(with maybe 2k subscribers):

1. Film a seaside with normal sea breeze and wave sounds (recorded at the same time) no music added.

2. Some Sony Music subsidiary claims that video infringes on their song from a different region which apparently contains similar sea sounds(facepalm here). Ok stuff happens algos are only human..

3.This is where the evil of Youtube comes in:

You file a dispute, point out the obvious in a professional manner -> YouTube seemingly reviews your dispute and claims that the video is infringing still!! It is obvious to anyone that the response was automated.

4. The only way to win this nonsense is to escalate to the next level of dispute which most normal humans are loathe to do (scared of lawyers, scared of YouTube, etc).

Again, the YouTube is evil for taking out the humans out of the review process while faking the review process.

We know they do it for everything else and that is Google modus operandi - to automate everything requiring human interaction.

In step 3 all that happens is that Youtube sends a second request to the original bot asking if they are correct. All of the bots immediately return yes. There is no penalty for the media cartel lying to YouTube so they don't even care.

For the cartels the decision tree is simple. Issue a demonetization claim and take the money for the channel. Doesn't matter if they own it or not. If the user successfully contests it then you're just out the money going forward, but nothing else. There's literally no reason for them to act ethically, so they don't. If YouTube does push back they have the weight of Copyright Law to bring down on YouTube, making even relatively small violations worth potentially trillions of dollars in damages.

Could someone just automate this escalation process in the same way people were automating the fighting of traffic tickets?

What's the upside of doing so? Is there a cost associated with it that you're going to have to pay even if you win (eg attorney or even just filing costs, or travel to a venue)?

The rights holder most definitely would not have a claim to pull the whole video down. If a human looked at that and saw it was music playing at the mall that was incidental to the video, it would be a clear case of fair use.

Fair use is entirely dependant on the context of the content. It is never really a “clear case”.

It’s decided on a case by case basis by the courts.

Sure, but historically, no one would bother taking such a case to court because the supposed damages from a video in that kind of case are obviously non-existent.

The ability to spam these kinds of take-downs at essentially zero cost and with no ramifications for the chilling effects and false claims create an environment where it is the norm.

Sounds like an opportunity for a classic class action lawsuit.

Blame YouTube because they made the system. And if you dispute the claim, the clamer gets to decide if the clame was false. The system is totally wrong.

Exactly. The DMCA itself, as bad as it is, is less broken here. Under the DMCA, if you post a DMCA counter-notice, the site puts the content back up, and the claimant has to sue the person who uploaded it rather than the site. As bad as DMCA notices are, that system would still be much better than YouTube's current system.

You have a copyright claim against the claimant if they decide wrong. A lawyer would be needed to peruse it though, so it probably ins't worth it.

funny how they can detect that, but all the unofficial pirated music on there is not possible to detect

That's what I was thinking initially too.

But I don't think their goal is to have the videos removed. It's to collect any ad revenue generated from the video.

If they collect all of the ad revenue then the last thing they want is for the video to be removed because now it's free advertising for them. If 100 people upload a song and those songs get millions of views, the record label keeps all of that ad revenue.

I believe that's why you can find nearly every song ever made uploaded by many people on Youtube. The record labels and Google love it because they get paid.

Also, I believe thats why a lot of those videos show the lyrics in the video portion - it changes the audio's use.

That's rotten.

Is it even possible to scale a human operation to this level, even with youtube's checkbook?

It's not that AI needs to replace empathy alone, it needs to replace culturally and legally nuanced judgement about copyrighted works, among other things. That's hard to teach fast to a mechturk type workforce, and staffing a team of people competent enough to understand that nuance at youtube's scale has to be prohibitively expensive if it's even possible.

Humanizing operations buys quality at the expense of putting a upper bound on scale. Few companies are willing to accept that ceiling for fear of losing market share.

No but they should be putting more humans towards their larger creators. Anyone with over 100k subs shouldn’t have to beg on social media just for YouTube to take action. Some creators have direct contacts but even those contacts are limited in what they can do.

Youtube could also offer a fee-based genuine review process for epople who care about their channel enough to get their pocketbook out. And to make it even better, if Youtube is as sincere in their claims of "doing their best", make the entire negotiation public, so we can start to finally see the actual nature of what's going on, rather than rely upon unconfirmable promises and pure speculation, as is the current state of affairs.

Do they have to do this? Of course not. Being a private company, they can do whatever they want.

However, by not doing things like this:

- they are constantly losing public support and trust

- they are setting a track record of evasive and non-transparent behavior, that could be used in potential future government legislation

- they run the risk of a cashed up honest/transparent competitor suddenly bringing a new, honest platform to market, catching them with their pants down both technically and ~morally (in the high ground sense of the word). This may sound impossible with the scale of YouTube, but think about this: what if someone brought a platform that only publishes (and maintains) quality (for some definition of the word) content? Like a YouTube for ~serious content? This way, they could leave the expensive hosting of nonsense to YouTube, and cherry pick the quality content, upon which a future self-sustaining business model will likely be very dependent. I think there are probably a number of people in China who have both the money as well as understanding of human nature (including deceitful abuse of legislation and limitations of human intelligence) necessary to make this happen. Whether it's worth the risk I have no idea, but some day user-produced content is going to be a huge cash cow, not to mention the propaganda power inherent in owning the #1 platform.

I agree, but the fee should be a deposit placed by both parties willing to have a review, with the loosing side paying. Direct payments for reviews would create bad incentives to not improve their AI.

Or strait up go for external arbitrators. But "I'm willing to put money behind my words" seems like an easy signal against abusive mass claims.

I really like that idea. Even if you only make both sides put up $5 each, loser pays, that would add up a lot for automated bots making fraudulent claims and hopefully not too much for people being claimed against.

If a single human reviewer takes 15 minutes on average to process each claim, then that's $20 an hour coming in per reviewer, which could make it viable.

Hopefully that would have the dual effect - firstly making it too expensive for bots to make thousands of fraudulent claims, massively reducing the volume of claims, and at the same time having humans hopefully making fair decisions instead of the AI just assuming the claimant is correct.

That makes some sense. I'm not sure where I stand on it.

On the one hand, it seems a bit unjust that smaller creators should be snuffed out for things that wash off the back of established players. A two tiered justice system isn't just.

On the other hand, it's a reasonable (and more cost effective) way to do more than nothing.

A bit gross, but it would probably work in YT's favor to suppress cries of injustice early. Would this even be a scandal if it only hit small creators?

> A two tiered justice system isn't just.

Luckily, this isn't the real legal system.

The way I look at it, it's really about the users affected, rather than creators. A creator with 100k fans has more people depending on them than a creator with 100 fans. It makes sense that Youtube would be more careful with the former than the latter, because of the customer impact.

I agree, but fortunately the smaller channels rarely have issues. People go after the larger creators because of the deep pockets and such. Better to start somewhere, at least.

They should be putting humans towards problems that large creators have but not specifically towards the large creators. That way they can help everyone. Those people would naturally be dealing with the large creators more often just because they are heavier users.

That's easy to say and sounds good, but doesn't solve the scaling problem. Every video creator wants some human judgement as to whether their background audio is fair use or not before the video is demonetized. If you give that treatment to every video creator, you are soon going to be employing half the population of the planet as copyright judges

No you use the AI to help the human. And you use the human to spot check the AI. It’s totally scalable.

So 3 years ago, there were 24,000 channels with more than 100k subscribers, and more than 40 channels per day were hitting that mark. Even with over a million subscribers, channels make less than $20,000 a year.

There are simply too many people, and they don't make enough even on the big channels to justify that sort of individual attention. In fact, the only way they can make ANY money is to have so little human attention.

https://socialblade.com/blog/youtube-milestones-four-channel...

https://www.inc.com/minda-zetlin/even-youtube-stars-with-14-...

If they hired just ten people to do 15 minute manual reviews for >100k channels, that's 250 reviews a weekday and more than enough to take care of all the biggest problems affecting them.

It gets actually pathetic when it comes to channels with a million subs. Sure, some of them don't make a ton of money, but there's only a few thousand total. One small team could do so much. If a channel makes $20k a year, with a significant fraction going to youtube, why can't it get $50 of someone's time?

>Few companies are willing to accept that ceiling for fear of losing market share.

The question I find interesting is: is this assumption correct? Is service scale the dominant factor in market share, or could you win a significant piece of the pie with a "Smaller YouTube by Humans"?

In theory, a more human touch could attract content creators, who in turn could bring their audience. Whether that would translate to significant market share, though, dunno.

> The question I find interesting is: is this assumption correct? Is service scale the dominant factor in market share, or could you win a significant piece of the pie with a "Smaller YouTube by Humans"?

At least so far, the answer to your second question has been "no," at least if you define "significant" as "that which attracts significant VC investment." (I'm not suggesting that's the best metric, but it's what an awful lot of tech companies use in practice.) This is because the answer to your first question is demonstrably yes: when it comes to services that heavily rely on network effects, service scale is the dominant factor in market share.

It's possible that you could carve out a sustainable niche if you really leaned into the "by humans" part, but it'd still be a niche. I don't think that's necessarily a reason to avoid it.

This already exists. For example there's stream.cz, content creators are curated and may get some help with creating content if it's good.

There are other similar online "TVs", like https://www.mall.tv/.

> possible to scale a human operation to this level

Here is one stat I found from 2019:

- 400 hours of video are uploaded to YouTube every minute

So if videos were reviewed at 1x speed, they would need 24000 humans working 24x7. (Not sure how to scale that to actual jobs, but seems manageable)

Related: "Facebook had more than 30,000 employees working on safety and security — about half of whom were content moderators." Not a fun job:

https://www.theverge.com/2019/2/25/18229714/cognizant-facebo...

Videos at 1x speed seems to be a very naive approach. Only viewing videos detected/reported with copyright claim would be a more sensible approach, as that is where nearly all of youtube's issues lie. Even putting a few hundred people manning disputes from large customers would be hundreds of times better.

I don't think that stat is relevant. This is about reviewing only those videos that face copyright claims.

Ah, that seems trivial in comparison.

For Facebook, they mention a contractor salary of $28800/yr. Multiplying by 30k is just under $1 billion/yr (assuming no overhead).

>Is it even possible to scale a human operation to this level, even with youtube's checkbook?

YT probably doesn't make money. We don't know because Alphabet doesn't put it as a separate line item in their earnings reports. Insiders say it about "breaks even".

Can you explain exactly the level of scale you're talking about? How many humans do you think would be needed to handle copyright claims?

80 years of video is being uploaded to youtube every day so do the math

80 years, or 700,000 hours.

At $10 an hour that would be $7m a day, $2.5b a year.

Alphabet's revenue is arround $144b a year, so it would cost 2% of revenue to review every single uploaded video. Clearly that's not going to be needed -- you only need to review those which

1) Have a claim by $BIG_CORP

2) Have a counter claim

I don't want to do math with flawed data. Every minute of uploaded content is not going to face copyright strikes, and does not need to be reviewed.

if it works for large scale and profitable webhosting companies, it should work for youtube.

granted, this is not exactly apples/apples, but it's not far off

Until the new EU copyright law with the upload filters, YouTube didn't have to make the copyright takedowns as aggressive as it did. It went way above and beyond what the law was requiring. Why did it do that, you ask? Because if was part of whatever deal Google made with studios in order for them to give it access to songs for its failing music services.

One could also argue that if YouTube's takedown fitler wasn't as "good" (where good doesn't actually mean objectively good, but aggressive) as Google made it be, then EU's upload filter wouldn't have passed either, because then there would have been no example of anyone "doing it right" (read: taking down anything that smells like a cousin of a copyrighted work, including stuff like public works, bird chirps, etc -- just to be sure).

My point is, YouTube wouldn't have needed as many humans to check if people's taken down stuff was needed to be taken down, if its algorithms weren't designed to be so aggressive in the first place.

Google dug its own grave here. Now it's stuck between the creators who increasingly see it as a hostile/too risky service, and the people who keep calling for YouTube to censor stuff that "offends them", and who will never ever be satisfied with whatever censorship regime YouTube puts in place, just like the copyright trolls never will be either.

I've seen videos analyzing the YouTube trending tab and suggesting that human moderation suppresses videos from 'controversial' creators like Joe Rogan and PewDiePie.

It's dangerous to become the arbiter of taste, I can understand Google's aversion to more human involvement beyond just the problems with scale.

I've seen others that allege that the YT related-to algo is being systematically gamed to raise the prominence of such videos, so the removal is a return to norm, rather than a suppression.

What's frustrating is that we don't have many ways to find out what's true.

Youtube's algo is an arbiter of taste whether they're happy about it or not. I'm not sure it's possible to be apolitical here.

Yes it is. Just recommend videos similar to ones you've watched, and attempt to cater neutrally to your sensibilities. Use views and engagement to determine when something is trending. That's apolitical. Arbitrarily excluding Joe Rogan or PewDiePie is political.

PewDiePie is literally the most popular vlogger on the internet by a wide margin, and Joe Rogan is the biggest podcaster on the platform, I find it hard to believe that any "suppression" they are facing is at all significant.

They do apply humans but not where it matters to creators. A lot of creators hate their automated compliance schemes.

Curation is terrible too, when it comes to kids content. I won't let my kids use YouTube kids anymore, it's awful. It used to have mostly appropriate kids content, but now people have gamed it.

The algorithms are a proxy for some youtube engineers' tastes

That's a popular conspiracy theory, but it doesn't hold up to Occam's Razor. The simpler explanation is that content from all sides gets buried to some extent, but it appears biased to people because they only notice when their own content is buried.

Why is it hard to believe that an engineer would engineer something that agrees with their own beliefs?

It’s because of the way the conjecture is structured. It proposes that we believe a multi billion dollar corporation would risk their profit margins on the tastes of one engineer in the shadows.

Occam's razor would say that such a large corporation does not have the direct controls to monitor every single actor. Is it so hard to find actors acting against the interest of their employer? I've even personally experience cases where the actors are rewarded for such actions due to flaws in the incentive structure.

Saying AI does funky things from time-to-time is hardly an extraordinary claim.

Meanwhile, an individual injecting bias without the hack's true purpose being obvious to co-workers would be difficult. The more subtle and circumstantial the hack, the more likely it will be broken/ignored by someone else's changes.

A hack is still possible, but I'd want evidence of intentional bias before entertaining the claim.

Algorithmic bias does not have to be an intentional "hack" to have drastic effects.

Because most people don't work in silos that allow them to operate with impunity. I'd assume at a company the size of google any feature is going to have minimum of 3-4 different sets of eyes on it between Engineering, QA, UX, and Product.

Have you never heard of algorithmic bias?

https://en.m.wikipedia.org/wiki/Algorithmic_bias

It's quite real, and it is not illegal (in the US), so asserting this is a "conspiracy theory" is silly.

Software engineers are human beings, so of course they have inherent bias.

The very term "conspiracy theory" is frustratingly flawed in the first case, in practice being nothing more than a perjorative for anything that runs contrary to the official narrative, where 'official' is highly context sensitive.

In practice it means 'a theory about a conspiracy' with the implication that conspiracies aren't real and therefore the a theory involving a conspiracy is wacky bunk. But sometimes the official narrative is that there was in fact a conspiracy, and in those cases the term 'conspiracy theory' is not typically used. A prime example is the 9/11 Commission Report which describes in great detail a conspiracy to commit mass murder. When it comes to 9/11, everybody agrees there was a conspiracy. I've yet to even hear rumor of anybody who believes that day was a series of unlikely coincidences. Whether you think it was bin Laden or Dick Cheney, it's obvious somebody was doing some conspiring.

I think the difference between the logical and practical meanings of the term is rooted in a specific case in which popular theories were theories about conspiracies, but for which the official narrative is that the guilty party didn't conspire with anybody. Obviously I'm talking about the assassination of John F. Kennedy.

It's no conspiracy, we all encode our beliefs into what we do. I don't think the poster is suggesting it's on purpose. Occam's Razor would suggest it's more work to take bias out, I would think.

“You know, you can remove men like Alan and me from the system, but we helped create it. And our spirit remains in every program we designed for this computer.”

— Dr. Walter Gibbs, Tron, 1982

> That's a popular conspiracy theory, but it doesn't hold up to Occam's Razor.

Occam's Razor isn't some sort of physical law that governs human behavior.

https://simple.wikipedia.org/wiki/Occam%27s_razor

> Occam's razor (or Ockham's razor) is a principle from philosophy. Suppose there exist two explanations for an occurrence. In this case the one that requires the least speculation is usually correct. Another way of saying it is that the more assumptions you have to make, the more unlikely an explanation. Occam's razor applies especially in the philosophy of science, but also more generally.

That doesn't refute my point. A razor is a tool for working with limited evidence, which is exactly what I'm doing.

Statement #1:

> That's a popular conspiracy theory, but it doesn't hold up to Occam's Razor.

This comment is open to interpretation. A reasonable interpretation is that something must hold up to Occam's Razor, or else it is not true, especially with the inclusion of "That's a popular conspiracy theory".

In fact, the "conspiracy theory" charge is also flawed. The speculative claim "The algorithms are a proxy for some youtube engineers' tastes" may actually be true. We do not know.

Statement #2:

> The simpler explanation is that content from all sides gets buried to some extent, but it appears biased to people because they only notice when their own content is buried.

This persuades the reader, perhaps based on a presumed knowledge of the phenomenon of confirmation bias, that any such beliefs are likely an example of such confirmation bias. Again, the truth is not known.

If you don't like your comments being interpreted, you could tighten them up to minimize the need or possibility for interpretation.

I think he's saying that positing Occam's Razor isn't proof of anything. If he's not, then I'm saying it, and that it gets over applied to the extreme.

Sure it does. I'm hypothetically designing a filter that checks for objectionable content on my site. You find any use of skeletons objectionable. I don't, if they're cartoon skeletons. My filter will not prevent cartoon skeletons from being posted, but will post real skeletons.

Bias has been introduced.

The fact that "both sides" are affected doesn't mean there's no bias, it means there's > no bias. It's literally impossible to censor something without SOME bias, because somewhere, someone doesn't find the thing you censor objectionable.

My right wing friend argues that Google straight up manipulated search results to harm Trump and I don't have any evidence to argue against him.

https://www.judiciary.senate.gov/imo/media/doc/Epstein%20Tes...

I wouldn't be surprised if there's an anti-alt right bias (which frankly I don't feel so bad about because many of those folks are toxic).

Also, to be pedantic that's a misapplication of Occam's razor. The simplest explanation to explain nature is likely true. Nature likes simple laws. People and organizations, however, can be endlessly complex.

Although bear in mind here that Epstein has a long-standing beef with Google that may be introducing a bias here.

> In 2012, Epstein publicly disputed with Google Search over a security warning placed on links to his website. [...] Several weeks later, Epstein admitted his website had been hacked, but criticized Google for tarnishing his name and not helping him find the infection.

https://en.wikipedia.org/wiki/Robert_Epstein

Here is a thought experiment: there are two political parties vying for control over the government of a country. Ahead of an election, leaked documents are published on the internet describing a plan by one of these parties to tear down the institutions of democracy in that country and install themselves permanently as a dictatorial elite. This party's political enemies will be rounded up and killed on an ongoing basis by a militarized police force, women will be stripped of their rights and property and reduced to chattel, possession of subversive materials will be punishable by torture and death, and so on. This is all outlined in the document, which was written and reviewed by the leaders of that party. They've been caught red-handed!

The other party has its skeletons-in-closets, but no analog to the above-described document exists. So, people using popular search engines to research their options ahead of election day find this horrifying information about party A and nothing comparable for party B, and they end up voting for party B mostly on this basis.

So, those search engines, despite providing results that faithfully reflect reality without a priori bias toward either party, will (hopefully) greatly influence the outcome of the election. Is that bad? Should they have censored this document in results so that nobody comes away from a web search with different opinions than they had going in? Is finding the truth an unacceptable thing for "big tech" to facilitate?

It's like when politicians like Steve King complain that search results for their names make them look bad. To me, if you conduct yourself as a racist, bigoted lunatic on the public political stage, that's what a web search result about you should say. I recognize that statements like "racism is bad" are subjective value judgments, but that's rolled up in all of this too: if as a society we don't think racism and other bigotries are bad, then this sort of search result won't be seen as negative. So there's some admission here that Steve King and his ilk know their words and actions are viewed negatively by a large segment of the population and so they would like them to be suppressed. By not suppressing them, are major search engines exhibiting a bias?

No, not really. I don't think so.

P.S. The burden of proof falls on the purveyors of conspiracy theories, not the other way round.

That's funny, because the only political video recommendations I get are right wing videos, despite actively avoiding them for years.

I guess when it comes to politics the Youtube algorithm just shows whatever the user doesn't want to see.

Youtubes PS4 UI "breaking news" section is exclusively videos about brexit/boris johnson (this was even before he became PM) for me despite me not living in the UK and having zero interest in the topic. It's weird.

I've always found this argument strange. Your comment isn't specific enough, so I don't want to put words in your mouth, but in general –

Do people really believe youtube engineers are encoding their values into the algorithm, intentionally or unintentionally?

Intentionally, I just can't imagine. What might this look like? How does it get through code review? Is it broad enough to apply to the entire ranking algo, or does it only act on tiny pockets that the offending engineers specifically care about?

In the unintentional case, certainly they're twisting specific knobs like engagement time, ad revenue, vids watched per session, etc... but is that encoding a specific person's taste, or a broader company's thirst for revenue? Do they even care about taste, or just metrics?

In other words, I find it suspect that youtube's policy is "More Jordan Peterson conflict vids", and more likely that it's "float high activity posts to the top", creating an echo chamber.

To be clear, I think there's some blame that rests on youtube eng (and let's be realistic, the broader product teams bossing eng around).

But it's not for encoding their tastes, it's for indiscriminately allowing the tastes of their markets to echo in a feedback loop that becomes a screech of extreme content. Not always anti-social, or even politically directional, just extreme flavors of whatever it is.

Unintentionally. Of course they are doing it. Youtube's content policy is so laughably bad that they probably have no choice but to take their own interpretation and use that.

It is definitely possible. Imagine an engineer gets approached by some other corporation that wants to boost a few videos, ads, etc.

It would not take much to nudge the system in that direction. The resulting payoffs could be in the billions.

They have so many goofy things going on over there. One example is certain bands will never make 'trending'

This is a video I recall watching recently that backs up your point.

Not sure if anyone has refuted this person's points but they are pretty alarming no matter what side of isle you sit.

https://www.youtube.com/watch?v=fDqBeXJ8Zx8 (What 40,000 Videos Tell Us About The Trending Tab)

I think it's pretty clear that "trending" videos are manipulated. I've seen a video with 10k views and five times more dislikes than likes show up in trending.

Unless you know the algorithm for how "trending video" is selected -- I certainly don't -- you can't say that that's "manipulation." Is it a number of clicks over a certain period of time? What's the period of time? How many clicks? And most importantly, how much of what any given user sees as "trending" is based on contextual information: what they've most recently watched, what they tend to watch on average, what channels they're subscribed to?

An awful lot of discussion around YouTube and human curation/moderation elides just how much content they're dealing with; the most recent figures I've seen -- which are over a year old at this point -- show that about 500 hours worth of video is uploaded to YouTube every minute. If we figure an average video is about 15 minutes (a 2017 statistic I found suggested that "the average length of a first page YouTube video is 14 minutes, 50 seconds"), that's 2000 videos a minute -- over 30 a second.

Is it impossible that YouTube employees have secret levers to raise and lower the profile of individual videos and channels? Of course not. But it's absurd to think that such manipulation is common, let alone corporate policy.

The actual fact is that YouTube has publicly committed itself to doing exactly that (building levers to raise and lower the profile of individual videos and channels), and continues to face media pressure to keep doing so as corporate policy: https://youtube.googleblog.com/2019/01/continuing-our-work-t... https://www.huffpost.com/entry/youtube-conspiracy-theories_n...

To me the problem isn't that flat earth videos exist on Youtube, but the implied "legitimacy" of the recommendations, along with the typically addictive UI elements like autoplay. Flat earth / Nibiru / etc insanity has always been on the internet, but when you saw it, you knew that you were in some obscure corner of the internet that was not normal. If anything, YouTube's insistence on mixing brand-safe, front-page recommendations along with content-adjacent recs lends credibility to the (many fewer now) "rabbit hole" recs. If you're on "Tony Blair is a reptile.mp4" and all you see are things like "Tony Blair is ACTUALLY a hologram.mp4" with 2k views, it's like "whoa, this is weird". But if you get 3 reptilian recs and 5 TED talk recs that have millions of views, maybe it seems more normal? I'm not sure.

Not only is it manipulated, it's flagrantly obvious. PewDiePie for example trends in many countries except the US [e.g. 45 times in Canada, 42 times in Germany vs 1 time in the US], and it's not just him either but many "controversial" channels.

Source: an analysis of 40,000 trending videos [7 months of trending tab videos]: https://www.youtube.com/watch?v=fDqBeXJ8Zx8

Apparently dislikes count almost as much as likes for popularity - videos that evoke an emotional response are considered better by YouTube. It's a general trend in social networks, because it leads to more views, and one of the causes of the current environment of polarization and toxicity.

Dislikes are another engagement metric.

While I agree with you, I think you over emphasize the technology and under emphasize the flawed contracts.

Eg, the copyright claim system seems fundamentally designed for abuse. You can't throw humans at it and expect it to work properly, it's working as designed now. You could change it, have humans handle the decision making or w/e - but if you are going to change it you could also "simply" write a better contract. Something not so flawed and easily abusable; of which there have been many ideas, discussions, etc.

YouTube does have human moderators though. YouTube just make a point to pretend they don't exist and the PR tries to make you think all moderation is algorithmic, because the alternative is angry YouTubers trying to retaliate against human moderators.

They can hire more people, but they will (and should) never let you call them up and tell at them after your video is demonetized.

Indeed. A friend of mine used to work for a company that shared an office with them. There's far more human work involved than YouTube wants people to think.

I don't really disagree with you, but I can't even imagine what "hire a whole bunch of people." looks like in this case, the numbers would be enormous I would think.

See https://news.ycombinator.com/item?id=20556643 for a rough estimate.

Facebook has 30000 moderators.

Facebook’s 30 000 moderators is problematic as it’s a pit of human misery, with basically fresh souls sent down the mine as canaries to detected hurtful content.

Even looking away from the human aspect a second, it means these 30 000 are not stable employees but contractors needed to be replaced very frequently, burning through the pool of potential workers. And people are not even happy with the current number and clamour for way more moderation on facebook, so the 30 000 number is by cheaping out.

Then Youtube need people both for moderating and copyright/monetarization support so multiply by 2, and for facebook and youtube we’d need close to a roster of 100 000 people rotating every year or less, burning through a million people in a decade.

All of that just for two video platforms on the internet.

I can’t stop myself from wondering if it’s a good use of people.

Agreed, not good for mental health. Story:

https://www.theverge.com/2019/2/25/18229714/cognizant-facebo...

Aren't they all contractors through Cognizant though? It's not like Facebook staffed up an extra 30K FTEs.

Lots of sites are driven by volunteer moderation - an admittedly risky prospect, but in most of those cases where it goes really wrong it seems like the site is functionally 100% volunteer driven.

A big corp could take a mixed approach. For every paid admin you can have dozens of volunteers whose authority and actions are supervised by the paid admins, and the admins and moderators would also have access to the AI tools to see flags on the content.

This could be a good model for Youtube Kids - whitelisting channels and individual videos as age-appropriate. If the admin sees a moderator abusing the privilege, let them revoke it and revert their entire history.

And I don't disagree with you but the parent company is ginormous and the monetary sums almost unfathomable. Its revenues may comfortably support an enormous number of well-paying jobs.

Alphabet's revenues are already supporting those jobs, though. People don't go to Google to do useful work, they go because the owners are hoarders and like collecting engineers to keep them in a zoo and stop them from starting competitors.

Thus Google will hire Vint Cerf because they think he's cool, even though he's not going to contribute anything, but what forum moderator is cool enough for a PhD hoarder to keep around?

The cost of the number of humans required to exclusively curate youtube would render the platform unprofitable in very short order.

Then maybe they go out of business. Capitalism doesn't guarantee that a firm's business plan will work forever.

As long as the advertisers don’t complain, their AI-based business model is perfectly viable.

On youtube, both the creators and the consumers are the product. The advertisers are the customers.

Well yeah, the point is that YouTube would not engage in a business plan that they believe would put them out of business.

AI flags anything that could be suspicious, human moderator reviews the exact circumstances of anything flagged. Google is one of the largest tech companies on the planet, if anyone can solve "impossible" problems its them. But not if they are lacking internal impetus to change anything.

The copyright problem isn't just on Google's end - copyright complaints should be required to be human generated as well. Otherwise anyone can be de-platformed/DDOS someone with bogus automated copyright complaints.

> they just need to bite the bullet and hire a whole bunch of people.

This issue is not specific to Youtube but to all platforms where users can post. I don't see how Facebook or Twitter or Reddit can hire enough people to review their communities.

No. Adding humans is a terrible idea.

"More Than 500 Hours Of Content Are Now Being Uploaded To YouTube Every Minute".

Do that math. Filing copyright claims can be completely automated. Humans can't keep up with that.

In my opinion, YouTube should enjoy Safe Harbor laws. When someone wants to make a Copyright claim, they should have to post a bounty. If their claim is found to be without merit, they forfeit the bounty to the person who created the video in question. And this should be resolved by a governmental judicial group, who also gets a cut of the bounty.

Your math is flawed. This is about reviewing videos that face copyright strikes. Every single minute of uploaded content does not need to be reviewed. And on top of that there are several technology aids that can be used to shorten review times - while still having some human involvement.

> If their claim is found to be without merit, they forfeit the bounty to the person who created the video in question.

That would mean there needs to be an appeals process and humans who do the reviews, right?

My math is not flawed. I highlighted that filing copyright claims can be completely automated. The punishment for doing so is nothing. I'm proposing a way to limit the false copyright claims.

It also shouldn't be Google's responsibility to resolve copyright claims.

If it's Google's responsibility, then you're just raising the barrier to entry for any YouTube competitor.

The people who benefit from Copyright need to themselves pay for fairly enforcing copyright. I can't imagine another way around it.

Google has designed a system where a power dynamic where the copyright owner largely gets the ultimate say because of a combination of artificial "intelligence" algorithms, the ability to perform bulk claims, and googles indifference. To insert a human element in terms of an appeals process or deciding which claims have merit, etc, has nothing to do with reviewing every single minute of content that is uploaded on YouTube.

Actually, it's Safe Harbor laws in the DMCA that have designed that power dynamic. And if there is no negative consequence for making a false claim, then yes, that's exactly what would happen - human reviewers looking at a huge amount of content.

Google could have designed the system where each claim must be specific to a video url, time, and duration. (I think they have made changes recently to do that, but I'm not sure if they're active) That way the copyright owners are left having to review this giant database of video content. Google helpfully runs audio finger-printing on every video and then lets copyright holders issue bulk claims.

>And if there is no negative consequence for making a false claim, then yes, that's exactly what would happen

Google can protect their creators and counter-sue on their behalf or set aside monies for legal defense funds.

>human reviewers looking at a huge amount of content.

More importantly, humans can be part of the appeals process and will be able to determine if someone was illegitimately targeted. As you can imagine its possible to design various review workflows where humans get involved after a certain level of escalation, or after other automated means have been exhausted etc. Also whats your definition of huge? You seem to now agree with me that this is not every single minute of YouTube content as you originally indicated.

> Google could have designed the system where each claim must be specific to a video url, time, and duration. (I think they have made changes recently to do that, but I'm not sure if they're active) That way the copyright owners are left having to review this giant database of video content. Google helpfully runs audio finger-printing on every video and then lets copyright holders issue bulk claims.

No, what happens instead is this: https://en.wikipedia.org/wiki/Viacom_International_Inc._v._Y....

Copyright holders don't want to do it themselves, and they have the muscle to hurt Google over it. This is going to happen to every video site when it becomes big enough to be a target. It's a bit absurd to suggest that Google could have designed a system that gave themselves no responsibility but simply chose not to; do you really think Google wants to be involved in these disputes?

But most of this content will get a negligible amount of views, which means that most of the conplaints come from a much lower number of videos.

But you just proposed the solution. A penalty for a failed copyright claim. Penalty+human intervention would cut the claims dramatically.

YouTube is already covered by safe harbor, and in theory making a false copyright claim is perjury! It's just that the law isn't going to be enforced in that direction.

I think Facebook proved pretty clearly that humans are not a solution to this problem. It doesn't matter if it's an algorithm or a human, at the end of the day the creators believe YouTube wants to censor their content so they will ascribe that motive to the execs, the engineers writing the algorithms, or the humans manually curating content.

Even as a non-youtuber, I think it's been made quite clear that the "motive" is selling ads, and advertisers don't want to be associated with particular topics or tones. It's no longer "Broadcast yourself", but "Deliver us specific content we can monetize, and don't dare touch major IP holders' content", which is a completely different creator audience.

> Deliver us specific content we can monetize, and don't dare touch major IP holders' content

The creators are complicit in this agenda, that's the reason for the backlash, they have an expectation of income from YouTube and when they run afoul of the algorithm it hurts their paychecks.

The Youtube Kids is another good example - algorithms gave us ElsaGate.

Wasn't that the time when little kids chose to look at super weird content?

I think that was utterly fascinating. A long hard look at the dark side of the psyche.

> That system is abused by large companies who know how to exploit their algorithms

Not even this, but the algorithms are designed in way that is bias towards the person making the copyright claim. It isn't like neutral algorithm and the companies just figured out how to leverage it better than creators, it is literally built in their favor.

>YouTube's problem is their allergy towards using humans. AI technology isn't advanced enough to replace empathy yet.

Not really a Youtube problem, but Google Problem. Google and Apple seems to operate at completely opposite end of spectrum. There is nothing Human about Google's Product and Services, even if they did try, it is as if it was designed by Data Mining and Algorithm. But it was fast, efficient, and instantaneous. Apple on the other hand is far too human, literally having a hand on every step of the process, which means it is very slow, and sometimes bureaucratic. But they do seems to care way beyond what a normal persona and other companies would do.

WRT copyright claims:

“Abused” in a similar way to how YouTube existed for (something like) 10 years without paying a single iota of money to anyone for copyrighted material? I’d call that abuse of copyright law myself (neglect can be a form of abuse).

Copyright claims is YouTube finally existing within the scope of the law. There are implementation issues - rerouting all money on a first come, first served basis is wrong - but Rome wasn’t built in a day.

And, from what I know of large record labels, they don’t know how to exploit the algorithms themselves. Most of them work in Microsoft Excel. Beating the claims system is beyond them.

They have software that they’ve bought to do that for them!

>If YouTube wants to remain viable, they just need to bite the bullet and hire a whole bunch of people.

I completely agree, but that's against everything Google wishes for, which is a world where everyone is 100% online 100% of the time with their data 100% in the cloud and decisions being made by AI 100% of the time.

Pushing for human intensive operation inside Google might as well be career suicide, this is why they would much rather keeping spending large amount of resource on "improving the algorithm" instead of just hire more humans and make more progress with a fraction of the cost.

> They simply need to add more humans into their processes

As usual, how many ?

In particular if we are talking copyright claims abuse, could you give a ballpark number of the number of humans you think youtube needs to hire ?

"In particular if we are talking copyright claims abuse, could you give a ballpark number of the number of humans you think youtube needs to hire ?"

That's a really rough one. Not only do I not have an algorithm for a computer to determine if something is copyrighted, I don't have an algorithm for a human to determine if something is either.

Since you're probably going to be stuck with an adversarial system based on claims and counterclaims anyhow, I'd submit the problem is probably more related to a lack of symmetry. The content creators put stuff up, then there are thousands of entities that can analyze their stuff automatically and make automated claims with no apparent consequences for being wrong. The individual creator then gets notified of a claim, but they know nothing about the claimant, whether they make routine fraudulent claims, whether they have legit ownership, etc., and have to address them relatively individually, by appealing to an opaque system that tells them very little about why the ruling is what it is, and where all the penalties appear to fall on the content creator rather than the claimant.

Something needs to be a bit more symmetrical there.

This is something where a "YouTube union" might start really being dangerous to YouTube; suppose the union puts up a page where they ask all the members to record exactly what claims are made on what content, with all the (meager as it may be) metadata YouTube gives them on the claims. I betcha some patterns would emerge. (I believe there's already some word-of-mouth about certain claimants but I bet this would make it even worse for YouTube.)

Enough so that a human can review every claim. I don't know how many claims they process in a day so I can't give a hard number.

But I suspect the number of copyright claims would go down significantly once they have a bunch of people reviewing them, because the erroneous claims would become unprofitable, assuming they punish entities that make too many erroneous claims.

> enough

That could be a tenth of the planet’s working population, and we all know it. It’s not like youtube’s scale is some well kept secret.

If we are thinking about any serious solution and not some philosophical one, just throwing bodies at the issue is a no go (and would be inhumane in my opinion, I don’t want even millions of people spending their life dealing with copyright claims)

Either change the rules about who can emit claim, change how claims are emitted, change the effect of the claims so they’re not so drastic. There’s many realistic directions the issue can be taken.

If there would be a button for claiming copyright that has a checkbox above saying "I accept liability for falsely claiming copyright." the problem might fix itself after enough lawsuits?

Most youtube creators have no money to spend on lawsuits with big corporate entities.

A similar power imbalance exists between workers and employers. If the worker doesn't get paid, they face the prospect of having no money for rent, let alone suing their employer.

Happily, the Department of Labor exists and merely mentioning their name is frequently enough to scare an employer into compliance. I think a similar arrangement could probably be found to even the playing field in copyright, but finding the political willpower to get this done will probably prove difficult.

This would ideally be policed by Google since it ruins their reputation and steals time from them too.

It'd likely also require proof of liability insurance.

They could improve the AI to figure out when it doesn't know something. As for a cheap way to have humans in the loop, some kind of voting by trusted creators might work.

Humans are not scaleable.

Don't think youtube wants the liability that comes with active decisions.

Imagine making a mistake that favors a pirate, lawsuits. Make a mistake that favors a copyright troll, endless hate?

A bunch of countries are going to force them to do it anyway.

I think this is absolutely wrong. Youtube algos offer plausible deniability, something that was existentially threatening to any future profitability for youtube in the case of copyright claims.

I'm fairly sure that most of the content watched on youtube was uploaded by someone other than the copyright holder, at least historically. Having an algo people have to actively thwart, can be unaware that the particular content is copyrighted, and/or generates a significant enough number of false positives that google can claim that they're trying hard enough to damage a portion of their legal business is a tool in and of itself.

Not using an algo makes it clear you're doing editorial, and once you're doing editorial, oversight follows. Anyway, even with the algo you can do as much editorial you want through training and dispute resolution. By slow-walking some disputes and fast-tracking others, you can change the shape of your business to hopefully be more independent of legally fragile stolen content.

The Trump/Russia conspiracy was a boon to all of these algorithmically filtering DMCA carriers, now they can actually explicitly filter points of view. I decided that point was inevitable anyway after I heard of the first facebook censoring of an anti-wingsuit post. IMO humans + that degree of editorial = worst case scenario, at least laundering it through the AI means that editorial requires some cost and degree of engineering skill.

https://www.nationalgeographic.com/adventure/features/why-ar...

> “The post got deleted by Facebook, and I pissed off the entire community,” says Lewis. “My message was followed up by the most fatal month in BASE. I just painfully sat back and watched my friends die one after the other.”

Your point seems to be hiring humans is wrong, because YouTube incorporated more AI to pretend like they're doing something, remain in the good books of copyright folks and make themselves more money. Is that correct? So, what if a person does not sympathize with Google or care about them making more money? Ultimately your perspective is shaped by whose side you're on.

My point is that hiring humans is wrong for Google.

> So, what if a person does not sympathize with Google or care about them making more money?

I have no advice for them. They should do what makes them happy. My point is that Google does not sympathize with them (mostly because it is not a person) and makes decisions based on how the people who control it will benefit.

Okay fair enough.. :)