Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Exactly -- and the situation arises because "fair use" determinations aren't scalable. It literally ultimately takes a court and a jury to determine if something is fair use or not, which is incredibly expensive.

Now in 99.999+% of individual human situations, it never gets that far because it's pretty clear to a person in advance whether something is fair use or not, or they ask a lawyer and assess the risks, etc.

But when you're dealing with user media at scale, that system falls apart. YouTube can't have a human lawyer manually review each clip for fair use, YouTube certainly can't trust uploaders because there are tons of people trying to upload entire Hollywood movies, and asking copyright owners to manually review each flagged match is similarly not scalable.

So YouTube simply implements a content matching system that disallows more than a certain amount of copyrighted material period. There's no other scalable solution given existing copyright law.

And honestly, this new eraser tool is a really good solution if it works. If it can simply and effectively remove unwanted background music that was never wanted/intended to begin with, then it works for everyone -- pirated content uploads are still blocked scalably as well as unlicensed sountracks on webisodes etc., but people can still upload personal clips and instead of them being blocked, they can just remove the music instead.



This is what the DMCA safe harbor policy was meant to deal with. Hosts are not held liable for user content if they respond to a DMCA takedown notice in a timely manner. The other half of the system is that users are supposed to be able to submit a counter notice to have content immediately restored without judicial intervention. At that point the host is immune from litigation and the original party can seek remediation in the court system.

YouTube doesn't provide an implementation of the second half of the DMCA process because they have a side deal with big media to manage royalties in exchange for a system biased in their favor.


They tried the safe harbour argument against Viacom. It didn't go that smoothly.

https://en.wikipedia.org/wiki/Viacom_International_Inc._v._Y....

That whole debacle spawned Content ID and yes, now they're happily in bed together. They probably prefer it like that, while we'd prefer safe harbour actually works as intended.


> supposed to be able to submit a counter notice to have content immediately restored

Not immediately AFAIK - the provider has to wait ~2 weeks (in which case the original issuer of the notice can sue).

> YouTube doesn't provide an implementation of the second half of the DMCA

AFAIK YouTube does implement this: https://support.google.com/youtube/answer/2807684?hl=en

However, most copyright disputes on YouTube happen within YouTube's own process (where usually the claimant gets the revenue from the video but the content usually stays up), not the DMCA process (takedown and copyright strike)


> There's no other scalable solution given existing copyright law.

(Informercial hands slip on screwdriver) Call Today for our $29.99 special grip that solves all those slippery screwdriver problems you definitely have! There is no other solution to the screwdriver grip conundrum!

No, practicality does not demand "binding shitty algorithmic decisions for thee, extreme latitude for egregious errors from me." Determinations don't need to be scalable to backstop a system of back-and-forth escalating claims that keeps the incentives correct for everyone at all stages: human beats algorithm, identified human beats unidentified human (note that at this point and all subsequent points rights holders have an enormous, automatic scalability advantage), identified human with legal commitment to consequences for being incorrect beats uncommitted human, and finally bump it to the legal system if all else fails, but by now everyone has skin in the game committed to their claims so none of the disagreements will be spammy.

This is all possible, it's not even particularly difficult, but it wouldn't create a cozy relationship with big rights holders which is what youtube actually wants, so instead we get "binding shitty decisions for thee, extreme latitude for egregious errors from me."


> Determinations don't need to be scalable to backstop a system of back-and-forth escalating claims that keeps the incentives correct for everyone at all stages

That's yesterday's game. It might have been possible to do this in the 90s, but today's copyright claims are automatic, authoritative and legally legitimate enough to scare a platform owner. This is entirely legal, too; nothing stops Sony from dumping 800,000 alleged infringements on YouTube's lap and giving them a 2 week notice to figure it out. If Google doesn't respond to every claimed abuse, then Sony can force them to arbitrate or sue them in court for willful copyright violation.

> This is all possible, it's not even particularly difficult

But it's not automatic, it creates unnecessary liability, and it's more expensive than their current solution. It's not overly generous to Google to assume that they also hate the rights-holders, but literally can't be assed to do anything about it because the situation is stacked against them. Even assuming the overwhelming majority of copyright-striked content is Fair Use, the losses incurred by the 0.1% that isn't could make defending YouTube a net-negative. Record labels and movie studios keep IP-specific lawyers on-payroll for this exact purpose, and fighting it out is a losing battle any way you cut it.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: