NVIDIA’s DLSS 5 Demo Video Briefly Taken Down Because YouTube’s Take Down Process Sucks
from the the-italian-job dept
Last month, we discussed NVIDIA’s demo video for its forthcoming DLSS 5 technology and the controversy surrounding it. While I’m going to continue to be of the posture that an injection of nuance is desperately needed in the reaction to AI tools and the like, our comments section largely disagreed with me on that post. That’s cool, that’s what this place is for, and I still love you all.
But this post is not about DLSS 5. Rather, it’s about the video itself and how it was briefly taken down over automated copyright claims thanks to an Italian news channel. Please note that the source material here was written while the video was still down, but it has since been restored.
And now, here we are in April, and NVIDIA’s DLSS 5 announcement trailer is no longer available to watch on YouTube on the company’s official GeForce channel. And no, it’s not because NVIDIA is responding to the feedback and retooling the technology for a re-reveal or re-announcement; it’s now blocked on “copyright grounds.”
A clear mistake, but also one that highlights the limitations of Google’s automated system for YouTube. Apparently, the Italian television channel La7 included footage from the DLSS 5 reveal in a recent broadcast and has since copyrighted it. From there, essentially every video on YouTube with DLSS 5 trailer footage was issued a copyright strike and said to be in violation, with the videos taken down with the following message: “Video unavailable: This video contains content from La7, who has blocked it in your country on copyright grounds.”
Yes, this was clearly a mistake. But it’s a mistake that I’m frankly tired of hearing about, all while Google does absolutely nothing to iterate on its copyright process and systems to mitigate such mistakes. The examples of this very thing are so legion as to be laughable. Whether due to error or due to malicious intent, videos that include content from other videos for the purposes of reporting and commentary, which are then copyrighted and result in takedowns of the source material, happens all the damned time.
This is almost certainly all automated, which means there are no human eyes looking for an error in the flagging of a copyright violation. It just gets tagged as such and taken down. And, no, the irony is not lost on me that we need human eyes to keep an automated copyright takedown on a video about AI from occurring.
What makes this alarming is that the video was taken down with seemingly no human interaction or input, as it’s clear that NVIDIA not only created DLSS 5, for better or worse, but also the trailer that has been a hot topic of discussion this year. We’re assuming this will be resolved fairly quickly. Still, it will be interesting to see whether YouTube responds to this case and claims that false copyright infringement notices like this are prevalent on the platform.
Google hasn’t been terribly interested in commenting on the plethora of cases like this in the past, so I strongly doubt it will now. Which is a damned shame, honestly, because the company really should be advocating for all of the users on its platform, if not especially those that are negatively impacted by this haphazard process.
But, for now, the video is back, so you can go hate-watch it again if you like.
Filed Under: copyright, dlss 5, geforce, takedowns, video games
Companies: la7, nvidia, youtube
Comments on “NVIDIA’s DLSS 5 Demo Video Briefly Taken Down Because YouTube’s Take Down Process Sucks”
Three strikes for making false claims, then your ContentID account is closed.
What’s good for the goose…
Perhaps it’s clear to a small segment of the population who follows this sort of thing, sure. But I doubt that even among the subgroup who voluntarily call themselves “gamers”, more than a tenth are actually aware of this video.
Which is to say, the premise that it’s feasible to correct this with more human intervention isn’t very convincing. Even if it were feasible for Google to hire enough individual people that all aspects of the human cultural zeitgeist would be clear to at least one of them, we’d simply move the problem from “the automated systems don’t recognize the context” to “the automated systems don’t recognize the context, and therefore can’t properly pick the human moderator who would recognize the context.”
Google could do better, somewhat. But not that much better. Until the law actually deigns to address bad actors, it can only really get worse. Have a few companies go bankrupt for their lies, 99% of the problem would clear up overnight.
If anything, this is probably the best case scenario under the current paradigm. The only realistic alternative under the existing legal paradigm is that all the resources are diverted to give special, individualized protection to the rich and powerful so this doesn’t happen to nvidia, and Google bets on everyone else being too poor to object when nvidia does it to them. At least this way, everyone gets to suffer on occassion.
Re:
It’s easy enough to do better by requiring those on the inside of rhe Content ID system to not poat/submit anything containing original content not their own, until they do something more nuanced like mark segments as “Not for Content ID” in editing. It is easy enough to put a signal of some sort in. That it still has yet to be done just shows what a lot of lazy motherfucking bad faith actors they all are.
Leonard French video
Leonard French recerntly posted a video on this very subject.
https://www.youtube.com/watch?v=NoT_dsC3ZX8
Nuance is difficult to achieve when the companies behind AI continue to try to bullrush it into everyone’s lives in the least nuanced fashions possible, and when the energy and water requirements for the technology are so unfathomably astronomical that they actually had to kill Sora and are asking for increases of energy production that have never been met.
With energy prices already through the roof, and data centers threatening to send it even higher, the nuance splashes into the brick wall of “if my bills go up I will have to choose between heat and food.” That becomes a threat to survival and to quality of life. Any nuanced discussion of the potential gains of AI are utterly meaningless within that context.
But of course, that’s not the point of this article. The problem with the copyright takedown methods is that Google has zero incentive to do anything about it. There’s no tangible profit to be had in making their system less ‘broken’ for the accused – only potential liabilities.
Want them to fix it? We need a massive course correction on copyright, which currently operates on “guilty until proven innocent.” Had I my druthers, all copyright allegations would be manual – automated systems for sending out accusations would be completely illegal. Every single one would require the accuser to affirm that they’d actually reviewed the material they’re accusing, with heavy penalties for failing to do so.
There has to be teeth to false accusations – make those scenarios just as ruinous as actual copyright infringement and see how quickly the dynamic shifts.
Re:
See comment prior to yours. This court ruling could and should change the landscape. But, you know how it all rolls.
As I’ve said before, the issue goes way over Youtube’s head. It’s an issue with how governments make copyright infringement takedown laws (eg: DMCA) that incentivise these sorts of automated systems. It’s Masnick’s Impossiblity Theorem combined with copyright maximalism.
Re:
Except the system doesn’t have to be the system it is. It could have always been far better.
Re: Re:
The issues are threefold.
1. Trying to make changes while staying within “Safe Harbor” provisions of DMCA and the like.
2. The global patchwork of copyright laws.
3. The scale of YouTube as a platform (meaning it runs headlong into the Impossibility Theorem).
This means that, realistically, any “reforms” on YouTube’s side will be tinkering around the edges at best because the root causes of the issues we’re seeing here are outside of YouTube’s control.
Perhaps Nvidia should take the stance that turnabout is fair play, and file a copyright complaint against the Italian TV station, and see how much they enjoy being on the receiving end (doubly so considering Italy’s stupidly maximalist takes on IP)
=> For us Dummies:
What’s the bottom line point/significance of this stuff ?
Re:
Bottom line: copyright works on “guilty by default,” Which results in Nvidia’s own video getting taken down because someone else claimed it.
Some people want to expand this regime to other things – which will only result in more abuse. Don’t support these efforts, from anybody. They’ll just let people actually silence you.
Can the output of DLSS 5 be copyrighted?
Yeah, whatever your feelings on Nvidia’s DLSS5, the trailer for it¹, and/or AI in general, the trailer being taken down from YouTube over copyright because of a news channel’s video about the trailer is just completely nonsensical. Yes, moderation is impossible to do well at scale, but I don’t think it’s too much to ask for it to be done better than this, at least for copyright.
¹ Personally, I’m of the opinion that, at the very least, it doesn’t look particularly great in the trailer, the trailer leaves out important info that might have mitigated the reaction to it, they should have at least let the devs of the games showcased know about it, and it shouldn’t have been called “DLSS5” given how drastically different it is from the other DLSSs.
Yeah, the YouTube copyright takedown system is a disaster. People who fraudulently abuse the system can get away with censorship all day long with little to no consequence while it is creators who suffer from it. So, I agree, the system desperately needs to be reformed in some way so that there is some semblance of balance. This highlighted incident shouldn’t have happened.
Re:
The biggest problem is trying to do that while also stayinbg within the Safe Harbor provisions of DMCA, etc.
And that’s a nigh impossible task when you consider that different countries do copyright law differently.
And of course YouTube is a massive platform so the scale of infringement notices would be absolutely huge (see also: Masnick’s Impossibility Theorem).
Simply put, any “reforms” would be tinkering around the edges at best because the root cause of the issues is outside of YouTube’s control in the first place.
This can probably be attributed to laziness and/or neglect on the end of la7. My understanding is that ContentID has the option and ability to add timestamp segments to exclude sections of a video not owned by the ContentID account holder for exactly this type of scenario. The problem is most ContentID users are too lazy, apathetic, don’t understand/think about the nuances of copyright, or are just overworked and underpaid staff, to take the time and actually use the feature for its intended purpose.
Re:
Oh, so they do have a method for this. Ah.
There is an article that I read somewhere about a struggling folk artist who had her vocal and musical style copied by an AI, then all of her music was copyright striked, stripping her of all her meager income. Seems like a business model has developed whereby someone trains their “AI” to copy you, then copyright strikes the original content it learned on to monetize it.
Re:
This is all over Spotify.