Hacker News Reader

Comment by gumby | Hacker News Reader

This guy Vasallo sounds like an entitled, self-righteous prick. YT offered him a fix but he didn’t want it.

There is sort of a weird point buried in the story. YouTube have Content-ID, so why don't they just use it to ID the same videos across the platform when a DMCA take down notice is issued? Why only remove the one specific video?

Surely a rights holder would be able to tell YouTube: Hey, these videos are mine. I don't want them on your platform.

Why does he have to excuse YouTube from any past copyright violation in order to have them use Content-ID to remove his films? I don't understand YouTube logic here. Sure a waiver would be nice, but it's not required for Content-ID to work.

As I read it, Vasallo want YouTube to use a tool they already have to police their platform.

Copyright is probably YouTube's biggest legal risk. ContentID was developed in response to lawsuits over copyright. The DMCA has major problems, but I think the safe harbor provision is extremely important.

So does Google have liability for past copyright infringement when they quickly responded to his takedown requests? I don't think so. I'm not a lawyer, but if that's correct, signing a waiver won't cost anything. I understand why Google doesn't want to do more for someone who wants to sue them over their biggest legal risk.

I am not a lawyer, but i think there is a substantial difference. specifically, if you issue a takedown and they don't take every copy down, you can't sue them for that.

Basically, to me it reads as if any failures of the Content-ID system to protect your copyright are no longer youtube's problem, but yours. I can see why someone would not be okay with such a waiver.

Mind, I could very well be wrong about the legal interpretation - but if the copyright owner made the same reasoning, he'd still end up at this conclusion. Even if it turns out to be legally incorrect.

>but if that's correct, signing a waiver won't cost anything.

If that's correct, it also doesn't cost YouTube anything to remove it from the contract. Obviously youtube thinks those terms do something, otherwise they wouldn't mind taking them out of the agreement.

I see this all the time when it comes to contract negotiation. If one party has significantly less negotiation power, they're expected to immediately cave on ancillary terms under tortuous logic that they're unimportant.

It's fine to accept terms on a "I'm not losing anything, because I don't care about losing it" basis, but never on a "I'm not losing anything, because it doesn't change anything" basis.

In a lot of ways, I think he’s smarter than a lot of people and companies. He’s not giving up his legal rights.

I feel like we’ve all given YouTube permission to sidestep the law and do whatever the heck they want. We’ve empowered them at the loss of our own state-secured rights.

It looks like he's the "too smart for his own good" kind of smart.

The only thing this lawsuit will do is instigate a careful review of every DMCA complaint he ever filed to make sure they were handled correctly. Those are the only "state-secured rights" he actually has. The whole point of the DMCA is the safe-harbor clause that indemnifies Google/Youtube from the claimed liabilities as long as they followed the DMCA process correctly.

From a legal perspective, it doesn't matter one whit that the DMCA process is so broken that it's basically useless.

Some relevant discussion in the piece:

> “Once a pirated movie was found, Mr. Vasallo and Plaintiff would send YouTube a takedown notice. YouTube would then remove the pirated video movie in its entirety. However, YouTube would not remove all matching videos, as YouTube would specifically only remove the one video from the one infringer related to the single infringing upload identified in the takedown notice.

> “The same infringer would be free to upload the video again until three takedown notices were filed against him within a ninety-day period. Then and only then, would YouTube cancel the infringer’s username. The same infringer could then create a new username and begin the process of posting the pirated videos all over again,” the complaint notes.

> Over the past several years, the law firm sent over 10,000 takedown notices for pirated films that were viewed more than 500 million times on YouTube. The video service maintains that this is how the DMCA works, but the movie tycoon argued that the company should do more.

This (that is, uselessly) is also how YouTube used to handle domestic copyright claims. It is the process the law specifies. But they moved to their ContentID system despite not having to. Why?

Whatever reasons applied there may also apply here.

Assuming YouTube does dedup on their backend then this is the same issue that megaupload has to deal with before they were taken offline. Namely, if an infringing file has multiple links pointing to it, is the platform obliged to take down all of them when presented with a DMCA notice? If so, is the platform then obliged to take down every slightly different version of the same infringing content? Where do their duties end?

As much as we malign content it is still more workable than having YouTube manually purging all videos that appears to be infringing based on very little information. And most content creators who sample copyrighted media under fair use have learned to cope. The more egregious issues are more concentrated around misattributed licenses and musical performance.

> Assuming YouTube does dedup on their backend then this is the same issue that megaupload has to deal with before they were taken offline.

This is not an assumption I would make. MegaUpload was a file transfer site. Videos on YouTube do not tend to be duplicates of other videos on YouTube.

I agree. I'll occasionally come across full episodes of TV shows on YouTube. Uploaders will try to get around copyright detection by doing things like flipping video or inserting black frames. I'm assuming a simple dedupe based block could easily be bypassed.

Publishers sometime elect to leave full episodes on YouTube either to collect passive ad revenue or just to entice people into watching more on a paid platform. For example, the 1st season of Man vs. Food can be found on YouTube in its entirety. However all the later seasons are heavily cracked down by contentID to the point that the handful of surviving episodes all had serious visual or audio distortions rendering them unwatchable - some episodes were literally uploaded with a badly tuned anaglyph 3D filter that will probably need its own algorithm (or a pair of specially purposed 3D glasses) to unscramble.

If so, is the platform then obliged to take down every slightly different version of the same infringing content?

Well, no, because the platform doesn't know which of those links were published with permission of the copyright holder and which weren't.

Why not? Why isn't there a box for "copyrighy owner" on the upload form, and YT can check when the same content has different claimed owners?

If you don't want to sign away those rights, you can manually track content and file DMCA requests. Per the article, YouTube is responding to those promptly. Vasallo isn't entitled to have his cake and eat it too - he can either opt into the automated system with all that implies, or take on the Sisyphean task of identifying infringing videos himself.

It's only sisyphean because Google actively hides their catalog from copyright enforcers.

Responsibility flows from their self-imposed exclusivity.

There's an adversarial relationship there. We have ample evidence that rightsholders often overstep their legal bounds, demanding takedowns for fair use or even stuff they literally don't own. Giving them an easier way to browse their catalog could well increase the number or scale of bad-faith removal requests.

I could frankly see some firms that imagine if they got the right tools, they could legal-DoS YouTube to death, and obviously all the former viewers would come running to their $15-per-month walled-gardens.

He's asking for Google to go into the censorship area of proactively not allowing to publish content. Google will fight nail and tooth against it.

This is the opposite what US gov is calling for - to provide more transparency in the existing takedown practices.

It seems that content platforms are cought between a rock and hard place here. And the innovation needed is not a technical one - we have implemented the current laws to their full extent and they don't seem to fix the problem. We need to move the law forward now.

It seems that content platforms are cought between a rock and hard place here

That’s because they chose to use technology to push growth above all other concerns. They’ve reached such a scale that it’s not feasible for humans to moderate the content users are posting. This is a problem of each platform’s own making.

We need to move the law forward now.

That’s one position. Another one is to argue that not every business model has a right to exist. These days I tend to find myself leaning towards the latter. I think we walk a dangerous road when we give too much power (and too little responsibility) to big tech companies like Google.

> Another one is to argue that not every business model has a right to exist.

That what "move the law forward" means.

Yet you often hear in this forum about how content-id is bad and is too easily (ab)used to take down content. The reality is without content-id most rights holders would be as screwed as Vasallo.

I don’t think Content-ID is bad, it’s just a tool and I can see how useful it is. My problem is with the process it is a part of.

Google seems to treat Content-ID as basically the whole answer, when the process clearly needs more informed discretionary moderation. The recent upheld infringement action against someone playing Moonlight Sonata is a classic example of a failure.

I’ve got no problem with Content-ID matching the piano piece with the scammy upload that claimed copyright. I do have a problem with Google sticking to to guns in such an egregiously obvious case of a spurious claim of copyright. Fortunately that specific issue has now been resolved.

Google seems to treat [their automated processes] as basically the whole answer, when the process clearly needs more informed [human effort]

Isn't this basically their entire company policy?

You’re exactly right. Google resists hiring human moderators at every turn. I think they (reasonably) believe that it’s a slippery slope. As soon as they agree to moderate in some narrow part of their business then people will use that as a precedent to argue they should moderate elsewhere. At the end of this road is a place with so much moderation that their business model is no longer viable.

We as a society have some difficult questions to face over the coming years. Do we want to surrender control over our lives to algorithms owned entirely by big tech companies?

If their business model is not viable without human moderation, then perhaps it shouldn't exist.

Surrendering control to "algorithms" sufficiently deeply is indistinguishable from Kafka's The Trial.

Yeah I more-or-less made this point in my other reply to this thread. I yearn for the web of my childhood (mid 90’s) when web pages were small and highly personal. People formed communities via web rings. Search was very basic but largely unnecessary because the web was so small.

The Indie Web and Small Web movements (I count them as distinct) have done a good job of sparking the conversation so far, plus I guess Mastodon/ActivityPub helped to some extent. I think we'll see a mini-revival of the kind of content you're looking for.

Why shouldn’t it? Youtube has hundreds of millions of users and most of them don’t care the slightest bit about human moderation. If you only want to use platforms that have extensive human moderation systems, they exist (HN is one, I believe dang reads every comment).

Human-less moderation is regularly wrecking the livelihoods of many Youtube streamers, who provide valuable (often irreplacable) educational and entertainment videos to the community. Getting banned, or getting innocent videos demonetized/removed by "the algorithm" is very common. This, if nothing else, should be a motivation to provide some level of human arbitration/appeals court.

I'm sure that a good argument can be made that lack of human moderation makes Youtube an unwitting vector for videos that erode political and scientific discourse by spreading harmful nonsense, but I'm sure someone else already has written about it and I don't have the energy nor will to retread that path.

Why "yet"? Those two ideas don't conflict at all. It's not like it's a binary choice between the current version of content id or no content detection at all. It is easily abused, and biased toward claimers.

The fact that contentid would go too far in solving his problem wouldn't be a reason for him to complain!

Well content ID is bad for some cases like classical music where if you perform a piece that is in the public domain it will get claimed because the system apparently finds it really hard to distinguish different performances of the same piece.

IMO the content ID stuff is not the problem but the entire system, like before you take any drastic measure have a person check, also put as input in your stupid algorithm the channel information , a 5 years old channel with many subscribers would cause a lot of noise if your stupid algorithm causes issue so even from a lazy,selfish, stupid Google dev you would put a if(channelReputition> x) ...be more careful.

After working a lot with angular1 I have a really low opinion about Google devs, maybe they have some stars in their team but this does not mean that the products are good OR we can argue the devs are great but they were required intentional to build crap.

It's not reassuring to read that our only choices are two different ways for small creators to constantly be robbed

Yes

If he was slightly smarter he would have put his content on YT, content-id'd it and have it monetized by the ads by either what he or somebody else posts.

Instead he wants to fight it "the old way" which will not make him more money.

Why do you think the artists are "selling their catalogs" now? Most of them are too old-fashioned to know how to make money in the new platforms.

Reply on news.ycombinator.com