I will be extremely surprised if that will be required from small companies, considering the directive expressly says that "special account shall be taken of [...] ensuring that the burden on SMEs remains appropriate and that automated blocking of content is avoided."
(SME = Small and medium-sized enterprises)
There is also no language saying that pre-screening is required, only that "online content sharing service providers and right holders shall cooperate in good faith in order to ensure that unauthorised protected works or other subject matter are not available on their services", which could, IMO, be satisfied by a DMCA-style complaints process.
> The second point has been missed by most criticism of the Copyright Directive, and to me it's the biggest problem of all.
IMO trying to fix the false positive issue is the best part of the directive, as it requires access to human review, reasonably justified decisions, and it requires users to have access to an independent body for the resolution of disputes. This would be a big improvement over the current Youtube system.
(source PDF: http://www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//... )
The current proposed Presidency compromise text has been leaked, and you can read it here: https://www.politico.eu/wp-content/uploads/2018/11/SPOLITICO... , pg 22 onwards. (The EU Presidency is seen as the independent negotiator in the discussions.)
A couple of things to note -- this "compromise" text moves Article 13 into being a blanket liability, which means that intermediaries can be sued unless they get prior authorisation from all rightsholders. Given that anyone is a rightsholder for their own material, that seems... unlikely. The only conceivable alternative under the rest of the proposal is some sort of copyright filter — but even then, it has to be effectively perfect, so that's potentially not enough.
The other point is to note that under this language, being an SME is one of seven factors that may or may not make you liable; thus injecting even more uncertainty into whether you will be liable for the actions of your users.
(Disclosure: I work at EFF, and have been tracking this for some time.)
Will it? That sounds expensive at scale. I can see how some hosting services might decide that the thing to do is clamp down on who gets to publish things to help keep reviewing and appeals costs under control.
I mean, vetted creators of content are where all the money is to be made anyway, so why bother with the extra costs and pseudo-judicial process that comes with letting random people share media? Seems kinda unprofitable.
"The category of micro, small and medium-sized enterprises (SMEs) is made up of enterprises which employ fewer than 250 persons and which have an annual turnover not exceeding 50 million euro, and/or an annual balance sheet total not exceeding 43 million euro."
The EU added restrictions related to the data itself to prevent outsourcing the liability in the case of GDPR, it will be interesting to see how well they can define and prevent structuring in this case.
[1] https://publications.europa.eu/en/publication-detail/-/publi...
There's stuff in every version I've read about the measures required being different due to size so I don't really follow the point here.
And people on the other end (the uploaders) must have recourse, there's a mandated appeals process unlike the current situation.
The DMCA is a shining example of just how well these appeals processes work...
> Any complaint filed under such mechanisms shall be processed without undue delay and be subject to human review. Right holders shall reasonably justify their decisions to avoid arbitrary dismissal of complaints.
It also requires Member States to ensure "that users have access to an independent body for the resolution of disputes as well as to a court or another relevant judicial authority to assert the use of an exception or limitation to copyright rules".
(source PDF http://www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//... )
From the company’s perspective, it’s like a DDoS attack that is illegal to defend yourself against.
About the measures taken, for example:
> The measures should be proportionate in order to avoid imposing disproportionately complicated or costly obligations on certain online content sharing service providers, taking into account notably their small size.
For small/micro enterprises this then means that for the DDoS attack you'd need the rightsholders to invalidly notify of large amounts of content that needs removing (each one specifically) which must be done with the intent of killing off the platform. Would that be legal? To submit false claims deliberately to hurt the company?
Isn't "not wanting to take the legal risk" reasonable justification?
Emphasis mine.
The section also says member states are free to put in place independent authorities to assess these, although I'm sure somewhere else it in this is a requirement rather than simply a suggestion. I'll update if I can find it.
https://www.consilium.europa.eu/media/35373/st09134-en18.pdf
I’m more concerned about the way copyright itself may need to change to work sensibly with this — according to several UK lawyers I’ve talked to, under UK law any work is copyrighted the moment it is put down in a fixed medium. For example, this comment, the moment it is recorded to any database or cache. It is not a sensible rule as-written, and common law saying “now now, don’t be silly” is the only reason things are not already out of hand. Unfortunately most of the EU isn’t common law and one of the few countries which is is the one which is about to leave. It might be fine, but I can’t tell. I only speak English, mediocre German, and tourist/subtourist French and Greek.
It's the same in civil law countries, and the real reason things aren't out of hand is that every site that allows arbitrary users to upload content has some legal boilerplate such as
By uploading any User Content you hereby grant and will grant Y Combinator and its affiliated companies a nonexclusive, worldwide, royalty free, fully paid up, transferable, sublicensable, perpetual, irrevocable license to copy, display, upload, perform, distribute, store, modify and otherwise use your User Content for any Y Combinator-related purpose in any form, medium or technology now known or later developed.
This is likely very naive of me to ask but is it actually O(n*m)?
I don't know how the fingerprinting works but if it's anything like a hash I don't see why it couldn't be closer to O(n+m) since you only have to hash each copyrighted video once and then compare O(m) uploaded videos against the hashes.
I can imagine you can't actually compare against hashes of all copyrighted videos in O(1) since they can't just be kept in memory but I don't see why it can't be done in O(log n).
Of course this is very far out of my expertise so please someone explain to me why this is completely wrong =]
Of course, the checksum would have to be invariant to pitch shift, rotation, horizontal flipping, framing, clipping, and length changes — all of which I know have been used to evade copyright filters on YouTube. (Also must resist mild noise audio or visual noise, mild convolution filters, and probably hue shifting, but I don’t know if those have been used by pirates yet)
That’s not a technology business. It’s more like insurance. The principal balance sheet item would be legal liability. As a result, legal costs would be high. You’d need to capture most of publishers’ revenues, I think, to make the books balance.
For example, Microsoft already owns CBS interactive and would push more news traffic there with Bing results, rather than pay a link tax to third parties.
That burden or identifying infringing content exists no matter what. Right now, it falls on the creators to patrol the internet and play wack-a-mole with infringing uploads.
I can't really come up with an argument why creators should be saddled with that effort, when they get nothing out of it by default. Whereas all the profits accrue with the platforms.
It also seems trivially obvious that it's far easier for a platform to run some algorithm on their content, than it is for every single photographer and/or musician, author, etc. The latter would require an effort on the scale of Googlebot.
If youtube wants to profit from the possibilities new technology creates, they are on the hook to create the concomitant tools to prevent harms. It's no different than, say, an inventor pushing a new method to synthesize insulin being responsible for its safety. If it's impossible to do so, they can't just throw up their hands, point to their best effort, and tell patients to "take adequate precautions".
I just checked, they literally could buy most of the music and movie industry.
Problem solved
YouTube could just block non-local content to users in the EU and charge content providers in the EU an upload fee. The upload fee could be made low enough so as to not be intrusitive (or incubate competition) but high enough to cover a good fraction of the additional costs.
I literally built my own video sharing platform to get around it: http://brother.ly
Fuck YouTube and everyone who works for YouTube.
We don't "need" them, sure, but we live in a society. We're supposed to be able to delegate work to others who specialize in doing it.
No one is doing any amount of video over T1 lines in 2018.
Does that still happen?
I remember the massive boost getting a T1 at work was, in 2001.
there's no reason someone couldn't make a small bit of desktop software which, fed AWS/Azure/GCP/DO/github pages/etc credentials, set up every aspect of a website, and automatically synced some local directory of content to the relevant site.
The gp (notatoad) mentioned technical skill as a barrier and you responded to that point but easy-to-use idiot-proof software isn't really the issue.
The bigger problem(s) mega-sites like Youtube solves is financial.
When one uploads a video to Youtube, it doesn't matter if it gets 0 views or 1 billion views. The cost is the same to the uploader: $0. Hosting it on one's own with AWS/Azure/GCP/DigitalOcean/etc will have variable egress costs and budgeting/forecasting a potential expense of thousands of dollars is not something regular people want to tackle. (The difficulty of this should be apparent from the stories of sophisticated tech geeks getting blindsided by unexpected AWS bandwidth bills.) The other option of placing a bandwidth limit to cap the costs means you create a different problem: a popular video will result in web surfers seeing annoying "bandwidth exceeded" error messages.
Another financial problem Youtube addresses is monetization. Hosting it on a self-managed cloud means less opportunities for ad revenue. Yes, a lot of HN despise ads but many uploaders of desirable video content like the ad payouts.
Peer-to-peer distribution systems like BitTorrent, IPFS, and Dat are a good solution to this; distribution actually gets easier and cheaper as demand for a file increases.
I’d like to see a hybrid solution that combines the best of both worlds: for “base load”, there’s be a decentralized market to pay for storage and seeding, ala Filecoin; for videos that become popular, viewers would also automatically seed while and after they watch them. Actually, I’d be shocked if there wasn’t already an ICO or two with this exact idea. (Googled it; more like five.) We’ll see if anything actually takes off.
Still, it’s going to take a lot to beat YouTube’s offer. $0 to have unlimited videos stored forever, re-encoded in various quality levels, and streamed to anyone on demand via a fast CDN – with no ads, unless you choose to enable them on your video in exchange for a cut of revenue. And it’s not like that’s only possible because Google’s unsustainably throwing money in a hole. While they don’t disclose revenue or cost figures, rumors hold that YouTube is anywhere from breaking even to already profitable, and it’s certainly meant to be a profit source in the future. It’s a classic example of an economy of scale. YouTube is certainly losing money on the long tail of videos, but in exchange they get all the popular videos on their platform, which make enough money to pay for everything. It’s hard for decentralized to compete with that. Less hard as costs go down over time, perhaps – but still hard.
All those things require engineering, sales and/or other technical know-how.
I'm afraid it's mostly inclusive in the sense of "you're included in the menu" rather than "you're included in the dinner guest list". It's easier than ever to be a human battery for someone else's social media machine. Not sure it's a good thing. I don't see it equivalent to running your own website in the "old days" in any but most superficial ways.
Seeing how some of the non-technical people use Facebook is just soul-crushing.
Even when people are fairly technical, many do not seem to realize the psychological costs to them individually, and social costs to everyone around them collectively.
And it's only going to be cheaper in the future :)
With that said, is it perhaps possible that the problems publishers of content struggle with are in the main not those of the nuts and bolts of publishing? Perhaps, even, that these problems (content discovery, gaining access to users, analytics, ad management) might be well-addressed by platforms?
Giving companies direct liability for policing copyright in user-uploaded content is a much bigger disaster, on two separate fronts, both highly damaging:
1) Implementing youtube-style content id will be a huge burden for small companies. Youtube reportedly spent tens of millions on it over several years. This tilts the field in favor of large incumbents even more than it already is by nature. While the amount of content to screen is really small, human-driven filtering is a possibility, but as soon as you need to scale up you're screwed. Any startup in Europe that needs to do this is a no-go from now on, IMO.
2) Back to that multi-milion content ID system: it sucks. It sucks really badly. It over-reports infringement left right and center. Now imagine what it will be like when youtube is liable a priori for every false negative; any sane legal department is going to demand a minimal number of false negatives at any cost. Now imagine what that will do to the false positive rate. Now imagine what the false positive rate will be when every site has to bodge together a similar system, with neither google's money nor google's technical expertise.
The second point has been missed by most criticism of the Copyright Directive, and to me it's the biggest problem of all. I fear the wasteland that will result. Are we going back to the point where only a handful of content producers exist and everybody else should be content just consuming it? Welcome to the 20th century, Europe.