Welcome to the Utopia Forums! Register a new account
The current time is Thu Oct 18 21:54:05 2018

Utopia Talk / Politics / Fuck Europe: Copyright Edition
hood
Member
Tue Jun 12 09:55:00
http://ars...roposed-eu-copyright-overhaul/

Lots of links to sources, so I suggest reading the source.
------------

"Modern copyright rules fit for the digital age" is how the European Commission describes its proposals for the first major overhaul of EU copyright law since 2001. But a wide range of startups and industry, academic, digital activist and human rights groups believe that key elements of the proposals will cause serious harm to the functioning of the Internet in the EU and beyond. A vote taking place next week in the key European Parliament JURI committee will determine the likely shape of the law.

The most contentious element is Article 13 of the proposed directive (EU-speak for law). It seeks to make Internet services that host large amounts of user-uploaded material responsible for policing their holdings to prevent copyright infringement. Until now, companies have been able to draw on the safe harbor protection in the EU's e-commerce law, which online services enjoy when they are "mere conduits." The new copyright directive would withdraw that protection for any service that "optimizes" content, which includes things like promoting, tagging, curating, or sequencing a site's contents—most major online services, in other words.
Legal and technical problems

In the future, sites would have two options. They could enter into a licensing agreements for all the content uploaded by their users, although the proposed law does not explain how that could be done for fragmented markets where there is no single licensing body. Alternatively, online services must "prevent the availability on their services of works or other subject-matter identified by rightsholders."

The original text from the European Commission speaks of using "content recognition technologies" to achieve that. But the other key legislative bodies of the EU—the European Parliament and the Council of the EU, each of which produces their own versions of the proposed text, which are then reconciled in the so-called "trilogue" negotiations—avoid using this description. A recent version of the text from the European Parliament JURI committee says, "The implementation of measures by service providers should not consist in a general monitoring obligation."

That change is because critics pointed out that the EU's e-commerce law specifically forbids EU countries from imposing "a general obligation on providers... to monitor the information which they transmit or store." However, there is no other way of proactively blocking any unauthorized copies before they are posted without checking every single upload against a list of copyright works, and that requires general monitoring.

Alongside these legal issues, there are serious technical problems with the approach. The volume of user uploads is such that the filters will have to be automated. But navigating the subtleties of copyright law—things like fair use—is hard enough for courts, let alone algorithms. Automated systems inevitably make mistakes, and the risk of financial penalties means that they will be written to err on the side of caution by over-blocking material. This is likely to have a chilling effect on free speech in the region, and on the EU public's use of copyright materials in the production and sharing of memes, parodies and criticism. Public domain material is also likely to be blocked.

The EU recognizes that general upload filters will be onerous, and it intends to allow smaller companies to take down material retrospectively instead. EU politicians have also started adding explicit carve-outs for things like "online encyclopedias"—Wikipedia, in other words—and "open source software developing platforms." But it is not clear whether the latter exemption would apply to GitHub, say, since it hosts some software that is not open source. Adding in such exceptions makes the directive's text increasingly complex and hard to interpret, which is likely to lead to years of court cases trying to clarify its meaning.

An important question is whether the chilling effects, implementation costs, and legal uncertainty that Article 13 brings with it are worth it. Unpublished research carried out on behalf of the European Commission at a cost of $400,000 suggests that unauthorized uploads are not a pressing problem: "In general, the results do not show robust statistical evidence of displacement of sales by online copyright infringements."
Article 11

The other hotly contested section of the proposed law is Article 11. This is the "snippet tax," which seeks to make companies pay when they use even short extracts of text from news publications, as is common today on social media and elsewhere. A "Mythbuster" put out by European media companies speaks of "unauthorized large-scale reuse of publishers' content" as if copyright does not apply to newspapers. It does, of course, which means that publishers already have the legal means to pursue such large-scale unauthorized use. Extra protection for even the shortest snippets will impoverish the public sphere by discouraging the sharing of news.

A technical solution to stop search engines like Google from indexing content on news sites exists in the form of the robots.txt file. The fact that publishers don't adopt this simple means to block search engines is because publications derive considerable benefit from snippets of their stories being listed in search results. That was demonstrated when both Germany and Spain brought in national versions of the proposed snippet tax. Germany's law proved so bad for publishers that they gave Google a free license to index their publications. In Spain, the situation was even worse. Since payment was mandatory, Google closed down the Spanish version of Google News completely (Google search remained). Traffic to Spanish newspapers declined as a result, with studies suggesting falls of between 6 percent and 30 percent.

Although there is little risk that hypertext links themselves will be taxed—the original draft law has been clarified on this point—there is likely to be a reduction in sites linking to newspaper articles, and a loss of traffic for publishers. Revenue from the snippet tax will probably be small, as the German experience shows.

Many have come out against the snippet tax, including over 100 EU politicians, 200 European legal and academic experts, small publishers, and one of the European Parliament's own research departments.

---------------------------------

Tldr: Eu wants to essentially write over their version of Safe Harbor (services are not held responsible for user submitted data) as a handy to IP holders. They also want to try and charge for "snippets" (small pieces of articles being included in links for services like Google news) without directly making it cost money. You know, like how Germany attempted to do, until they gave Google a free pass to get those precious links and subsequent viewers back.
Seb
Member
Tue Jun 12 10:05:24
Equally, it's not exactly tenable to allow Google and Facebook to free ride on the rest of the content creating world and capture the bulk of the value.

Nor is it right that companies like Facebook and Google are able to dodge various rules that other publishes are expected to abide by simply because their explicit business model is based on being able to ignore them.

So while EU legislation on media tends to be daft on the specifics, the general principle here isn't wrong.

And I think the lesson of the German action is that action needs to be at scale. As predicted, Facebook did subsequently become the main channel for news and has had a market moving impact - extracting rents from media companies.

But in doing so it's arguable that it has itself become a publisher given it curates the news feed.

hood
Member
Tue Jun 12 10:37:56
Facebook != Google.

Facebook takes entire articles for their website and hosts then on Facebook - that is vastly different from simply including snippets of an article to preview next to a link to the actual article (at the website of the article's publisher). The law in question doesn't address what Facebook does, it addresses what Google does. Google inevitably links to the publishing website for anyone who wants to see the full article.
Seb
Member
Tue Jun 12 15:50:12
hood:

Then perhaps the author of the original article shouldn't be using google as the exemplar of how the law should apply?

I also think you are wrong: there is a huge issue on facebook where third parties are ripping off content owners, and then publishing on facebook. It's akin to the fake news (where the content has been generated but is bogus).

So this law definitely does capture Facebook stuff.

The law is seeking to make it harder for the big platforms to insert themselves, obligation free, into the market place and exploit their intermediary position to capture the bulk of the value of the market - and in their rush for profit provide a safe haven for copyright theft which also happens to help entrench them as the place to go for content.

The argument that they "can't police the content" is bogus: we are talking about companies with enormous cash piles - not just market capitalisation and cash flow but literally hoards of cash they can't spend - they don't want to police the content because it would cost, but also because copyright theft helps their expansion: it makes them the place to go to get all your news.

I think that's not an unreasonable position for lawmakers then to consider this to be a bad business model, particularly given the spectacular disregard and refusal for the platforms to take responsibility for what they host or publish, be that terrorist videos, copyright theft, fraud, or outright foreign propoganda aimed at subverting elections in clear violation of election laws.

They need to be required to either be a publisher, with the obligations that entails, or not be a publisher.

But as it stands, systemically their business models (whether via snippets or walled garden approach) are undermining traditional press and advertising regulations and the overall media ecosystem.

Perhaps if they voluntarily behaved a little more responsibly rather than defaulting to "not our problem we are just the pipes and have no responsibility for content", then law makers wouldn't be prioritising this issue.
hood
Member
Tue Jun 12 15:57:06
"I also think you are wrong: there is a huge issue on facebook where third parties are ripping off content owners, and then publishing on facebook."

Please explain how "publishing in Facebook" is remotely equatable to "snippets," which is the practice of publishing a single line or paragraph of a larger article?

I am not going to bother reading the rest of your post if we cannot come to an understanding of what snippets are. Whole articles published on Facebook are not snippets.

Unless you're talking about article 13? That is much more aimed at the likes of YouTube (or Facebook).
Seb
Member
Tue Jun 12 17:11:16
hood:


"The most contentious element is Article 13 of the proposed directive (EU-speak for law). It seeks to make Internet services that host large amounts of user-uploaded material responsible for policing their holdings to prevent copyright infringement."

This law would make Facebook liable for pirated news content.

Which would then mean that your claim

"The law in question doesn't address what Facebook does,"

is not correct.

I think the focus on snippets is a weird focus of the author of the article in the first post. There is no law being proposed here that singles out snippets.

But snippets is a great tactic for inserting yourself as an intermediary.

It lets google cream the bulk of the value of targeted advertising. If 80% of the market for news accesses it via a snippet, then why would an advertiser pay for targeted advertising with the news media when if they placed the adds with google they could reach the market earlier and get "people like X" spillover?

Seb
Member
Tue Jun 12 17:11:48
So I can see merit in arguing that snippets are parasitical in an attention based economy.
hood
Member
Tue Jun 12 17:41:31
"I think the focus on snippets is a weird focus of the author of the article in the first post. There is no law being proposed here that singles out snippets."

Uh...

"The other hotly contested section of the proposed law is Article 11. This is the "snippet tax," which seeks to make companies pay when they use even short extracts of text from news publications,"

So yes. There is a direct reference in the law proposal. Essentially, section 11 gives publishers the right to publish (which makes sense), but it refers to section 2, which defines something like google news as a publisher. Google news is hardly a publisher.


"This law would make Facebook liable for pirated news content.

Which would then mean that your claim

"The law in question doesn't address what Facebook does,"

is not correct."

I read your comment regarding Germans and took it to mean you were referencing the relevant law to the german example, article 11. What you described doesn't address what Facebook does, which is in relation to article 11, the snippets law.

Article 13 certainly does, but it is completely asinine to hold a company accountable for the actions of random users. Requiring facebook to potentially remove any preview code they have for links that ends up displaying an article would be reasonable, but suggesting that Facebook is responsible is ridiculous. It may be feasible with a text article since comparing text with a database of known texts is relatively easy; but with other content? How about scanning any and all pictures uploaded for potential violating content? Every video (or worse, live stream!?) uploaded must be audited for potential rights infractions, both audio and video in nature? That's beyond ludicrous.
Seb
Member
Wed Jun 13 03:27:48
Hood:

Missed 11 but it seems to me a sensible move for reasons outlined above.

Google news likely is a publisher: if it curated what you see then from a legal point of view it has many of the key attributes. But for the sake of argument lets say it is merely a non filtering conduit.

From a market perspective, it's leveraging it's position as search engine and operating system to capture the bulk of the revenue.

If one accepts that the new business models enabled by the internet mean something that would have been considered a publisher in print should now not abe consideted a publisher, and able to disown responsibility for content; then it's no great leap to say that new forms of regulation are needed for these business models.

Copyright and patents were created to provide systemic benefits though they infringe natural rights. I think there is a very strong case here that content owners should get this automatic funding rihht. Without it, you make add funded news a non viable business model which is very much against the public good; and it seems strange to exempt platforms from "analogue" rules on copyright as being too onerous and antithetical to the digital consumer but then insist content creators have no right to invoke the same argument when it's clear pursuing individual claims in court over months against transnational corporate giants that can absorb the court costs is clearly non sustainable.

Yes, the consumer may love the status quo: hi quality content served up via google; but that's not sustainable for content producers.

So it's up to govt to look at how to regulate this market in the public interest.
Seb
Member
Wed Jun 13 03:31:59
Hood:

're Facebook being responsible:

Picture recognition is as good as text. Much can be automated, the remainder handled by a well staffed complaints channel.

And then simply arrange a refund system where a flat fee based on average profitability of click through is given to the copyright holder.

If FB can price a page impression, then it I see no barrier to providing compensation.
jergul
large member
Wed Jun 13 06:44:48
Fair usage can be combined with copyright obligations to publishers.

The devil is in the details, but I see nothing wrong with the intent of the directives.
Seb
Member
Wed Jun 13 06:54:30
Facebook already scans each image uploaded for facial recognition.
hood
Member
Wed Jun 13 07:59:57
"From a market perspective, it's leveraging it's position as search engine and operating system to capture the bulk of the revenue."

By offering a naked page of news links? Or do news outlets think that they're somehow going to get people searching headlines on their own sites instead of through rss feeds or aggregators like Google news? We've tried this. It doesn't work how publishers think it will. People aren't going to miraculously start bookmarking the LA Times and checking them for news updates.


"If one accepts that the new business models enabled by the internet mean something that would have been considered a publisher in print"

It is an extreme stretch to accept that argument. Are we now calling news stands publishers? That's essentially what a service like Google news is. It's just your (very large) news stand with a bunch of papers and magazines on display for passersby to gaze at and choose what they find interesting. I guess the main difference is that a news stand has to pay upfront for the copy of news and sell at a profit, whereas Google news just sits there doing very little, letting people buy straight from the newspaper of their choice. Yeah, definitely sucks to not have guaranteed revenue like that. Too bad nobody is legally guaranteed to earn money. Fuck y'all for suggesting you should have that guarantee.


"Without it, you make add funded news a non viable business model which is very much against the public good"

Private companies are not in the business of guaranteeing revenue for other random private companies. If Europe wants to spend tax dollars paying for news organizations to keep the lights on, go for it. You'll receive no complaints from me. Trying to pass off that responsibility to a private, completely unrelated company is cowardly and spineless.


"Picture recognition is as good as text."

I disagree. See: every automated DMCA takedown request despite clear fair-use. Picture recognition may be able to identify that 2 pictures of Donald Trump are the same picture. Can they identify that the words attached to one picture makes it a pretty clear meme, while the words to another are just a text caption attempting to claim undo credit for said picture? The a-a identification of picture may be as good as with text, but the context recognition (of both) still sucks massive dick.


"Facebook already scans each image uploaded for facial recognition."

Which is potentially criminal. I know of at least one lawsuit that claims this practice to be illegal bio-data capture of anyone who has not expressly consented.


"The devil is in the details, but I see nothing wrong with the intent of the directives."

As usual, the vehicle of delivery is crewed by technically incompetent people who cannot grasp proper real-life analogies of tech. Claims that Google news is a publisher are laughable. Nobody would ever call a newsstand a publisher. So what you end up with is an attempt to force Google to prepurchase news without any real profit coming back. It isn't like Google gets the ad revenue when the user clicks a link to go to BBC, BBC does. So Google has to prepay BBC for its news, only to not have any recourse for profit from that transaction? There may be an argument here for what Facebook does, but that is because Facebook displays entire articles and as such, function much closer to an actual newsstand taking the risk for a chance to profit.
hood
Member
Wed Jun 13 08:08:47
One secondary thing about image matching:

Matching text to text generally has fairly easy to determine limits on what you're searching for. Matching images is so far beyond the scope in magnitude of text search. Consider that any movie, TV show, comic, cartoon, or video game would need to have millions upon millions of assets compared to a potentially infringing image. It may be easy with a limited dataset to compare image A to image set 1:9999, but what we're talking about is comparing image A to image millions of image sets 1:9999. Facebook facial recognition is easy, they only have many millions of faces to analyze. They don't have the whole of picture history to sort through.
jergul
large member
Wed Jun 13 09:13:00
Hood
The "newsstand" argument justifies a lot harsher legislation than "publisher" does. Newstands transfer a lot more of their revenue to publishers than publishers do to those holding the copyrights.

On reflection, Fair use captions should be exempt (say a paragraph can be published under fair use), More content than that triggers income transfers to the copyright holder.

I don't think viewing FB as a soapbox and appropriate course for the future.

hood
Member
Wed Jun 13 09:21:24
"On reflection, Fair use captions should be exempt (say a paragraph can be published under fair use), More content than that triggers income transfers to the copyright holder."

Right - something like Google news is just a display. Something like Facebook publishing entire articles is a pretty clear copyright issue.
Seb
Member
Wed Jun 13 11:22:03
Hood:

Yes, as I explained, it does allow them to capture the bulk of the ad value.

I don't think we have tried this in a systemic way. Because laws were local, Google could say "no profit share, cut you out, watch your traffic plumet". A legislatively enforced profit share across the whole EU greatly limits Google's options.


"It is an extreme stretch to accept that argument. Are we now calling news stands publishers?"

I know of no news stand that customised it's content for everyone that walks by, has infinite space for ads and is able to customise its ads to whoever is passing by; and is one of only two news stands in town; and is thus able to persuade all the advertisers to stop placing the most lucratice ads in papers but on the stand itself.

It's a cute analogy, but ignores all the pertinent points about the market.

"companies are not in the business of guaranteeing revenue for other random private companies."
If that's a fundamental principle, why should companies be forced to buy in independent auditors?
Why do we require firms offering some services not to offer others?

We impose costs and limits on companies to ensure a diverse market maximising the public good.

Ok, fine, let's say instead the rule is "you cannot profit from selling advertising based on content produced by others unless you have secured rights to do so". I think this reflects your principle that content providers are not in the business of providing revenues to platform aggregators.

"Can they identify that the words attached to one picture makes it a pretty clear meme"

I would say very likely. A project I know used AI to identify islamic YouTube videos with iirc 98% true accuracy for the UK Home office.
Seb
Member
Wed Jun 13 11:25:20
"Which is potentially criminal. I know of at least one lawsuit that claims this practice to be illegal bio-data capture of anyone who has not expressly consented."

Yes, but scanning for copyright infringement is not, and given FB can do face recognition automatically it seems the argument it's not technically viable to do copyright enforcement is unsubstantiated.

I think you over-estimate your technical savviness.

If Google curated or personalises it's news feed, it is a publisher in many definitions.
jergul
large member
Wed Jun 13 11:33:33
Seb
Would it not be simpler to support copy-right with a punitive damages / fine to ensure compliance?

Copyright holder or government takes to court. Government awards punitive damages or a fine.

No need to automate non-compliance if non-compliance has a hefty price.
hood
Member
Wed Jun 13 12:20:44
"Ok, fine, let's say instead the rule is "you cannot profit from selling advertising based on content produced by others unless you have secured rights to do so"."

I'm fine with this. As far as I know, Google news already operates this way. In fact, I already have the page up and see absolutely no ads. If there is promoted content (that news outlets pay for to get premium screen space), I'd gladly support axing that. I support moves to eliminate tracking of Internet users and targeted advertising.

Of course, we're talking about Google news because it is the most prominent. In the case where there are news aggregators that put ads on their news aggregation pages, I would happily support a move to get rid of those ads.


"one of only two news stands in town; and is thus able to persuade all the advertisers to stop placing the most lucratice ads in papers but on the stand itself."

This is anti competitive and any attempt to force out competition through such methods should be struck down quickly.



So, it seems like all of your concerns can be addressed without the absolute nonsense of calling a news stand a publisher. (Not to be confused with the wholesale copying of an article, a la Facebook)


"I would say very likely. A project I know used AI to identify islamic YouTube videos with iirc 98% true accuracy for the UK Home office."

You are aware that training the AI for Islamic videos is quite literally 1 use case, right? You're testing 1 video for specific elements. The proposal is to test 1 video against millions of videos, not just to look for specific elements.

You pretend like this is easy shit. Legal scholars, experts in the field have a difficult time identifying the minutia in fair use. You are hilariously out of your element if you think we could efficiently automate a process we don't actually solidly know ourselves.
Seb
Member
Wed Jun 13 16:51:28
Jergul:

Google and Facebook have very deep pickets, and there is no guarantee the plaintiff would be awarded costs.

Meanwhile, most new orgs can't take the cash flow hit.

This is why bigger media outfits can blatantly abuse copyright by ripping off peoples content on social media after being explicitly refused rights; and then the platforms can do the same to them in turn.

Hoods point re Germany's rules on snippets doesn't show that regulation is impossible, but it does show that the current framework is impractical for achieving the intent of the legislature.
Seb
Member
Wed Jun 13 17:01:44
Hood:

It's not as simple as direct ads - I used that as an illustrative example - the way they make money is more complex but does basically allow them to cream off the bulk of the value.

Facebooks news feed is way more prominent than Google news though.

"This is anti competitive"
Indeed - hence legislative efforts to try and tackle this because while anti competitive it doesn't unambiguously meet the very tight definitions and case law that apply to non-platform business models. Platforms are more like markets.

"You are aware that training the AI for Islamic videos is quite literally 1 use case, right?"
You were saying detecting fair use memes would be impractical. I think it would be much easier than determining if a video is terrorist propoganda. Look for big caption text for a start. It's pretty trivial. Stop pretending it's hard.

As for hard time fair use,has anyone been successfully sued for copyright abuse related to a meme?
jergul
large member
Wed Jun 13 17:50:09
Seb
Some media corporations also have very deep pockets.

The point with fines and punative damages is to correct behaviour.

Add on class action suits to recourses.

Why are you concerning yourself with how FB etc would seek to comply with the directive?

They will figure it out once a directive is in place.
hood
Member
Wed Jun 13 17:55:58
"You were saying detecting fair use memes would be impractical"

That was a ... snippet ... of my argument. Try reading.


"As for hard time fair use,has anyone been successfully sued for copyright abuse related to a meme?"

That just isn't how copyright/fair use works in implementation. You would be asking about takedown notices, and those happen all the fucking time. Consider the Prince/Dancing Baby debacle ( http://en.wikipedia.org/wiki/Lenz_v._Universal_Music_Corp. ).

Yes, this stuff happens. Your focus on a meme (which, by the way, here - http://www.theverge.com/2018/3/7/17089784/pepe-the-frog-creator-suing-infowars-matt-furie-alex-jones ) is just misplaced. This goes well beyond memes.
Seb
Member
Thu Jun 14 03:30:30
Jergul:

Because it's literally my job to think about regulations. You do need to think about how companies will behave in response - shaping the behaviour is why you regulate in the first place.

Copyright holders already have legal recourse through suing for damages, it is not very effective for the reasons I mentioned.

The effect of such a directive would be to greatly remove legal ambiguity and potentially create new processes for recourse less onerous than now.
Seb
Member
Thu Jun 14 03:35:34
Hood:

Yes but it was the specific snippet I was responding to with the video comment.


I'm not sure where you are now going but presumably we agree it is perfectly possible to determine when something is a meme?

Presumably then automatic take down notices etc. driven by a dumb also that spans everything could reasonably fall under the bad faith point in your second link.
Seb
Member
Thu Jun 14 06:21:37
This seems legit actually.

"http://www...ars-matt-furie-alex-jones"

Is it legal or remotely controversial for Disney to sue people over systemic abuse of Micky Mouse in other peoples work (whether charged or not) - they are entitled to protect the association and presentation of their character. Similarly, bands have been successful in suing politicians using their songs in adds and stump speech events.

Why is Matt Jukes not permitted to do the same? Is it because:
a. Matt Jukes isn't a big corporate entity
b. It's being done on the internet by alt-right people so that makes it about free speech?

Content creators have a right to protect IP - that's how the law currently stands.
hood
Member
Thu Jun 14 08:21:42
"This seems legit actually."

I wasn't arguing whether the Pepe suit was legit or not, I was providing evidence that memes (specifically, since you seem to have an unnatural attraction to focus on them) do absolutely get lawsuits.


"I'm not sure where you are now going but presumably we agree it is perfectly possible to determine when something is a meme?"

Do we also agree that it's perfectly possible to start creating legitimate copyright infractions disguised as a member to fool any such automatic analysis, causing the required analysis to be potentially exponentially more intensive to run?

Let's consider selfie monkey. If I take an image of selfie monkey and below it in nice bold letters, put the message "selfie monkey," we can both agree that the final product isn't a meme, correct? Now imagine I do the same, but put the text "has more rights than a black guy in the US" (a joke about the PETA lawsuit surrounding the picture). This second picture is a meme, correct? If we're just searching for image with text in a traditionally meme-formatted design, both are memes. So now we need image to text translation. We need to train an AI to identify the various subjects in a picture. We need to compare if the text (which we translated from an image to text) relates to the picture in a factual way (i.e. caption, not meme). We need to determine if there is attempted humor.

Suddenly looking for 2-3 items is looking for 6-7 items and performing much more intensive analysis. And this is just memes. What about dancing baby? How do we analyze that for fair use when it took several court hearings? What about the next dancing baby?


"Presumably then automatic take down notices etc. driven by a dumb also that spans everything could reasonably fall under the bad faith point in your second link."

That's my point. Your suggestion, the proposed law, is completely fucking asinine. It makes content hosters do all of this work, removing the burden from the IP holder to act in good faith (which they already don't) and perform corrective action as if wielding a sledgehammer. Politicians and yourself think this can be tackled easily with already existing methods to create a robust system to identify clearly infringing works. It can't. If such a thing were already easily possible, we'd be fucking doing it. We wouldn't have arguments of if a video whose main focus is the amusement of a baby dancing is infringing on Prince's song.


"Why is Matt Jukes not permitted to do the same? Is it because:
a. Matt Jukes isn't a big corporate entity
b. It's being done on the internet by alt-right people so that makes it about free speech?"

I've addressed this above, but I'm not saying that. I was simply providing evidence that lawsuits over memes absolutely do exist. Pepe's creator of well within his rights to sue Alex Jones and I believe he should win. But you asked if this stuff even exists. It does, as you can see.
Nimatzo
iChihuaha
Sun Jun 17 04:36:43
Hood
"Do we also agree that it's perfectly possible to start creating legitimate copyright infractions disguised as a member to fool any such automatic analysis, causing the required analysis to be potentially exponentially more intensive to run?"

As we saw with the youtube adpocalypse (interestingly youtube is held up as the benchmark), the limitations both inherent and emergent (open for abuse) in this tech, do not live up to the intent of the regulation.

The Commission may think that this paragraph:

"Member States shall ensure that the service providers referred to in paragraph 1 put in place complaints and redress mechanisms that are available to users in case of disputes over the application of the measures referred to in paragraph 1."

fixes the problem, but in reality by the time such a process is over (days at best) and you are found innocent, whatever your video or article was about, is old news and irrelevant.

Beyond this, there are judicial concerns as to whether article 13 is compatible with article 14 and 15 in the E-commerce directive.

From the ECD.

Article 14
Hosting

1. Where an information society service is provided that consists of the storage of information provided by a recipient of the service, Member States shall ensure that the service provider is not liable for the information stored at the request of a recipient of the service, on condition that:

(a) the provider does not have actual knowledge of illegal activity or information and, as regards claims for damages, is not aware of facts or circumstances from which the illegal activity or information is apparent; or

(b) the provider, upon obtaining such knowledge or awareness, acts expeditiously to remove or to disable access to the information.

2. Paragraph 1 shall not apply when the recipient of the service is acting under the authority or the control of the provider.

3. This Article shall not affect the possibility for a court or administrative authority, in accordance with Member States' legal systems, of requiring the service provider to terminate or prevent an infringement, nor does it affect the possibility for Member States of establishing procedures governing the removal or disabling of access to information.

Article 15
No general obligation to monitor

1. Member States shall not impose a general obligation on providers, when providing the services covered by Articles 12, 13 and 14, to monitor the information which they transmit or store, nor a general obligation actively to seek facts or circumstances indicating illegal activity.

2. Member States may establish obligations for information society service providers promptly to inform the competent public authorities of alleged illegal activities undertaken or information provided by recipients of their service or obligations to communicate to the competent authorities, at their request, information enabling the identification of recipients of their service with whom they have storage agreements.

****

There are more things to be bothered about, like what is the definition of the introduced requirements in the proposal;

"to the public access" and "large amounts of works"?
When do providers fall into article 14 of the ECD? The same providers that have filter out everything bad in article 13 of this proposal.

This is all made worse by the fact that there is a case law that actually tested and defined article 14 in a little more detail.

Fifthly, concerning the liability of the operator of the online marketplace, the Court first
recalls that only ‘intermediary service providers’ may rely on the exemption from
liability provided for by Article 14 of Directive 2000/31/EC on certain legal aspects of
information society services. The operator of an online marketplace is not an
intermediary service provider of this kind if, instead of taking a neutral position
between the customer-seller concerned and potential buyers, confining itself to
technical and automatic processing of the data relating to the offers it stores, it plays
an active role of such a kind as to give it knowledge of, or control over, those data.
http://ec.europa.eu/dgs/legal_service/arrets/09c324_en.pdf

^Re google/youtube and many other social media providers.

***
They need to go back an rethink this.
Seb
Member
Sun Jun 17 09:02:54
hood:

"I wasn't arguing whether the Pepe suit was legit or not, I was providing evidence that memes (specifically, since you seem to have an unnatural attraction to focus on them) do absolutely get lawsuits."

I think this is a relatively bad example as rather than taking a piece of work and using it largely as reference - normally a segment or a snapshot with additional context - here it's taking and appropriating the character where the distinction between the new work and the original is lost.

Banksies use of Micky Mouse in some of his art as a satire on capitalism is one thing. Someone else producing lots and lots of mickey mouse cartoons where Mickey supports the Nazi party isn't going to be covered by fair use anyway.

The examples you should be using should fall into the former, not the latter if the point you are trying to make is infringement on fair use.

"Do we also agree that it's perfectly possible to start creating legitimate copyright infractions disguised as a member to fool any such automatic analysis, causing the required analysis to be potentially exponentially more intensive to run?"

Anything is possible, but to the extent that is a real thing that might happen (it sounds rather unlikely) that sounds very much like the platforms problem to solve rather than something that means we should let them off respecting copyright. Generally, I think it would be hard to orchestrate an attack like this that would beat other style of DoS attacks.

"If we're just searching for image with text in a traditionally meme-formatted design, both are memes."

Are there actually lots of people publishing non-meme unfair uses design to look like memes?

And correct me if I'm wrong, wouldn't this simply result in the first image being wrongly cattegorised as fair use?

At worst, it creates a narrow edge case where people can infringe copyright by trying to look like a meme. There will always be edge cases, for everything. But it doesn't sound like the bulk of the issue to me.

So no, I don't think we need to go into natural language processing etc.

" Politicians and yourself think this can be tackled easily with already existing methods to create a robust system to identify clearly infringing works"

I think you are probably getting confused, as some engineers do, between the balance of outcomes here. We do not need an automatic system that can get to six sigmas with a technical solution here.

What you need is:
1. Good faith efforts by the platform providers that can handle the bulk of the issues and thus make manual processes effective
2. Robust processes for handling the remainder that are cost effective and do not place undue burdens on the copyright holder

"It makes content hosters do all of this work, removing the burden from the IP holder to act in good faith"

Well, I would argue the opposite. Firstly, the current law puts all the burden on the IP holder - which is unreasonable and unusual. We don't normally make it the homeowners responsibility to prove that someone robbed them out of their own pocket do we? Secondly, part of the reason that IP holders end up acting in "bad faith" is because the scale of the problem for them is huge, and the platforms publishers who are in a much better position to tackle the problem refuse to do so.

So legislation to clarify the issue: that platform publishers are publishers with the same obligations as other publishers is necessary.

Much as it has taken the courts and legislation to remove gig-economy business models that are built on regulatory arbitrage based on the bogus idea that directing someone through an app is different from directing them through written or verbal communications; I think the same is clearly necessary here.

Or, provide print publishers with the same freedoms to violate copyright at will on the basis that its unreasonable for them to go to the burden of complying. Otherwise you have an uneven playing field. What you can't do is have a situation where a particular medium is given a particular legal privelege over others. That is truly asinine.

Nim:

"Fifthly, concerning the liability of the operator of the online marketplace, the Court first recalls that only ‘intermediary service providers’ may rely on the exemption from
liability provided for by Article 14 of Directive 2000/31/EC on certain legal aspects of
information society services. The operator of an online marketplace is not an intermediary service provider of this kind if, instead of taking a neutral position between the customer-seller concerned and potential buyers, confining itself to technical and automatic processing of the data relating to the offers it stores, it plays
an active role of such a kind as to give it knowledge of, or control over, those data."

Yeah, the short version of this (and it applies in many jurisdictions not just EU) if you curate content, then you are a publisher.

And this seems reasonable, unless you want to water down other companies operating in other media from having the obligations internet based one should have.

I don't believe different standards of legal liability should apply to people operating on different channels.

This is how you end up with companies like Uber able to wipe out local taxi firms using a mix of what would otherwise be illegal employment practices, failure to comply with local safety legislation and in some cases criminal law, and taxes.

"It would be too expensive for us to pay employ people to do security checks, and to pay our employees wages, our entire business model is based on a huge number of software developers, and not having any management of our workforce, and claming they are not employees"

I.e. you business model is predicated on breaking laws and regulations that are precisely there to stop businesses making money by cutting corners on safety and employment law. Introducing the internet and an app shouldn't make a difference here.

Sam argument applies to copyright.
Seb
Member
Sun Jun 17 09:07:50
tl;dr

If a business model relies on operating at huge scale, but the only way to scale is to cut out existing regulatory compliance, then any company operating that kind of model isn't based on adding value, it's based at least in part on extracting it from public good and existing businesses that are complying with the regulation or were supposed to be protected by regulation.

Inevitably, as a society we will need to work out whether there is added consumer benefit in these business models that exceed losses elsewhere and which business models need to be curtailed or banned.


Nimatzo
iChihuaha
Mon Jun 18 03:42:36
>>Yeah, the short version of this (and it applies in many jurisdictions not just EU) if you curate content, then you are a publisher.

And this seems reasonable, unless you want to water down other companies operating in other media from having the obligations internet based one should have.<<

The current wording of proposal is not coherent, it contains ambiguous language, it introduces new requirements which it does not define. It is in direct violation of article 15 of the ECD, which recital 42 of the proposal mentions as still valid and is further not compatible with article 14 of the ECD. It is safe to assume the proposal was not meant to override the ECD.

A generic agreement that copyright is valid and important is not the debate, really. It is obvious that the intent of the law is not to destroy freedom of speech, but the current version risk many things, specially in light of the technical limitations is says are now an obligation, when again article 15 of the ECD explicitly forbids a general obligation to monitor.

"Yeah, the short version of this (and it applies in many jurisdictions not just EU) if you curate content, then you are a publisher."

From the ECD article 14:

(a) the provider does not have actual knowledge of illegal activity or information and, as regards claims for damages, is not aware of facts or circumstances from which the illegal activity or information is apparent; or

Further reading recital 42 of the ECD.

"I don't believe different standards of legal liability should apply to people operating on different channels"

You may not believe, but the EU does as it's regulation regularly treat entities (for instance of different size) differently. Which this proposal aims to do, but does very poorly, when it says "large amounts of works". EU product regulation always takes into account the perils of SMEs, trying to leave doors open for simpler assessments and not place undue burden on them. Which I believe is a sound approach. One of the concerns is precisely that this will make things more difficult for start ups in this sector.

"operating at huge scale"

Not defined as it is now. It is a poorly written law which leaves too much open for interpretation.
Seb
Member
Mon Jun 18 04:20:20
Nim:

The argument being if you are curating content then you cannot reasonably claim you are acting as an intermediary. If the post office starts deciding what mail it delivers, it's effectively opening and reviewing the content. And if it's reviewing content, then it needs to take responsibility for what it choses to present.

Demanding different levels of adherence based on size is very different from different laws for different media - and if anything only emphasises why we should expect more from the world's largest companies than we expect from e.g. a local newspaper; in practice we are letting them make vastly less effort.

The law could be written better, I agree. But the principle is sound.

Seb
Member
Mon Jun 18 04:24:18
Re making it harder for start-ups, maybe a little. But a tiny barrier compared to network effect lock-in.

I suppose you could argue that cos should be required open source algos used for copyright detection as a consequence of the tendency to natural monopoly


Nimatzo
iChihuaha
Mon Jun 18 08:48:41
"curating content"
The current wording, taken together with the ECD is not coherent as to who is curating and who is not. It was pretty clear prior to this proposal, but not anymore. Article 15 (forbidding monitoring) of the ECD was also tested in a trial and upheld, see below.

http://eur...N/TXT/?uri=CELEX%3A62010CJ0070

"The law could be written better, I agree. But the principle is sound."

SHOULD, it SHOULD be written better like before they vote on it. And while most of us agree with the principle of copyright, this is about the (concrete) soon to be application of the regulation intended to uphold this principle. And they made it easy for themselves those who wrote this! Filter technology is a deep sea trawler that will catch everything. This is not the intent of this proposal, but a consequences of the poor wording and over selling what filter technology can actually do.
Seb
Member
Mon Jun 18 10:45:22
Nim:

I'm not convinced we are actually disagreeing about anything here.
Nimatzo
iChihuaha
Tue Jun 19 11:51:35
It isn't obvious in this thread. The article, hood and me are concerned about the details and technical application and efficacy of said technology, further the ambiguity in wording and incompatibility with other laws. To which your response was, yes but the principle is sound.

I don't really think that means anything other than we should work on it. And presumably if one thought it was sound, one thought so without the help of the proposal. Copyright as a matter of principle is established in the EU and elsewhere, this law was not written to establish the principle yet again, but to clarify and establish a concrete solution(s) and make things better. So it is on those grounds, not the principle, that I am criticizing it.

I have no experience with copyright specifically, but like yourself I am paid to think (interpret and apply) about regulation, mainly from the EU and Sweden. Reading this thread it seems to me you are defending the proposal or at very least downplaying the risks or simply put, not seeing the cluster fuck that your "literally my job to think about regulations" eyes should see.
hood
Member
Tue Jun 19 12:48:58
I think he's focusing too much on my "Google news shouldn't be considered a publisher" opinion.

He also delved into mindspace arguments with it being unfair (or whatever) that a company like Google or Facebook having the main focus of an audience and a publisher of the various referenced articles being an afterthought. This argument is a bit nonsensical though. Google, Facebook, twitch, YouTube, fucking pornhub, or whatever other marginally popular aggregation service you choose to talk about, has put in serious effort to cultivate those userbases they have. Is the LATimes really doing much to earn dedicated userbases, or are they just mercenaries for eyeballs? Forcing these content hosters to pay publishers extra for directing users to their site is nonsensical. The publisher hasn't done anything of note to cultivate such a userbase, there is no loyalty. They're a one time attraction for the specific work they did. I'm all for requiring max lengths to snippets or for considering a certain length of quoted text to be a republication (i.e. 2 paragraphs - or even more direct: 200 characters - of material = republish and thus triggers payment) is a much more feasible and reasonable system than deciding that Google news is a publisher and therefore has to pay up for every news headline they display. I'm fine with profit sharing when it makes sense (i.e. enough of a material is used to trigger the case - similar to fair use). But this proposal doesn't make sense.


The other case, which is YouTube (or imgur, or twitch, or yes pornhub) style user-submitted content, runs afoul of different issues, part of which is how fair use is handled. The other part is the scanning requirement. The first isn't really addressed here, but should be. The second, more on topic issue, is a fundamental misunderstanding of just how daunting the task is to get it right. You know why YouTube has gone through so many rounds of demonetization? Because monitoring the number of $$ channels they used to have was prohibitively expensive. Instead of trying to handle the millions of channels, they've already condensed it down to thousands of the most actively engaged (i.e. profitable) channels and just given up on the rest. Google didn't need government action to signal that there's an identification issue. They cannot reasonably guarantee that their advertisers will have ads displayed on unoffensive content. They cannot reasonably act as police.

Further, Google hasn't even tried to monitor for copyright content. Instead, you have the various publishers scrubbing YouTube for specific things (prince v dancing baby as an example). These implementations are clearly not sophisticated enough to be reasonably workable or we'd never have had to deal with the nonsense of dancing baby. Can they make it sophisticated enough? Doubtful. According to statistics, over 300 hours of video are uploaded per minute. You think YouTube can analyze that against its 50 million copyrighted reference files (according to YouTube itself) and keep up? Without prohibitive investments? Doubtful.



So the idea behind copyright - great, we all agree. The idea that those creators should have methods at their disposal to get redress - great, we all agree. How to go about doing that, however, I think we clearly disagree.
Seb
Member
Tue Jun 19 13:22:41
Nim:

"Eu wants to essentially write over their version of Safe Harbor (services are not held responsible for user submitted data) as a handy to IP holders."

Hoods commentary in the OP. This, in context sounds like an objection more to the principle. I think my position in the second post is very clear.

Seb
Member
Wed Jun 20 08:33:14
I find hoods arguments very strange.

The argument that content aggregators have done work exactly points to their role as a publisher: tailoring their content to match the profile they have built up. This is the core job of a publisher: aggregating news (much of which will be from press agencies) and tailoring it to their understanding of their target audience preferences.

The key difference is that Google doesn't produce original content; and can curate at an individual level.

Google news couldn't exist without content producers.

Now let's look at it from a market perspective: the bulk of the cost in high quality news is reportage, fact checking, and writing and editing copy.

Other than that small matter,of finding out, fact checking, writing up in a compelling enough way to interest a user - what has the LA Tomes done to attract users? Its certainly an interesting question, but more in terms of what it reveals about the person posing it.

The argument that Google news is "sending eyeballs" to news sights is disingenuous. By inserting themselves as the first step they put themselves in the position of harvesting the bulk of the profile information that news media would previously use as a revenue stream.

Hoods strange belief that Google do not do this is sadly wrong. Google news clearly aggregated news for individual users to a profile.

And of course, recall their "first three clicks free" policy. It's just not the case that Google news is some kind of neutral entrant to the market with no serious impact on media business models and resultant to nctioning of the media market.

"Because monitoring the number of $$ channels they used to have was prohibitively expensive."

I really don't think depriving other people of their intellectual property and revenue should be permitted just because it inconveniences Google's secondary business model.

"Instead, you have the various publishers scrubbing YouTube for specific things"

Again a strange argument to make: surely this is the exact point. Google is quite happy if users interact with pirated content - there's spillover to advertisers (and in part that's what the up next feature is designed to do), and user data can be monetised.
Somehow Google has been allowed to shift the cost burden of policing its own platform to content owners. Not only is this an added burden to copyright holders, their limited ability to do so means a harder approach. Google has been able to claim it's merely a neutral intermediary, but increasingly it's acting as a curator showing this argument is somewhat hollow as it's making money from doing things it claims it's not able to reasonably do.

"Without prohibitive investments"
If it takes a lot of money to operate this business model without ruining other peoples business, is it not simply a failed business model?

hood
Member
Wed Jun 20 09:15:08
"This is the core job of a publisher: aggregating news (much of which will be from press agencies) and tailoring it to their understanding of their target audience preferences."

There is an extreme difference between "aggregating" raw bits of promotional material or raw information into a coherent article and just taking any given published article and providing a link to it. Suggesting that they are the same process ("publishing") is both disingenuous and blatantly retarded.


"Other than that small matter,of finding out, fact checking, writing up in a compelling enough way to interest a user - what has the LA Tomes done to attract users?"

I'm not from LA. What does the LA Times do, from an internet perspective, to earn my eyeballs? Exist? That's all well and good, but them existing and posting articles on their webpage doesn't actually notify me of anything. They still need to somehow attract me to their website. Were I in LA or the surrounding area, I might be much more aware of the existence of that company, but I am not. But the LA Times is from a big city and is a big news company within that city. What about the Deerfield Review? Ever heard of that publication? What have they done to earn Seb's eyeballs? A bit of research, typing and editing doesn't actually let you know that this company exists; and yet if big shit does down in Deerfield, IL, you might just find yourself reading an article by them - not because you're familiar with them, but because a separate source informed you of their existence (like say a search engine or news aggregator).

So, I ask again: what has the LA Times, or the Deerfield Review, or the whateverthefuck publication done to extend their reach outside of their local markets to earn the eyeballs of non-locals like you or me? They really haven't done much of anything beyond allowing themselves to be indexed on search engines, a passive non-move at best.


"The argument that Google news is "sending eyeballs" to news sights is disingenuous. By inserting themselves as the first step they put themselves in the position of harvesting the bulk of the profile information that news media would previously use as a revenue stream."

Since when? Since fucking when did any local paper or magazine ever have the network or resources set up to collect the type of profile data that Google has? You are insanely retarded.


"Google news clearly aggregated news for individual users to a profile."

If it's such a big deal, tell Google to stop. Google organizes such things to compel clicks on articles, which is a net positive for the publishers. It may end up picking favorites, but that is more trend analysis than Google trying to cause harm.


"I really don't think depriving other people of their intellectual property and revenue should be permitted just because it inconveniences Google's secondary business model."

You clearly do not grasp what you're arguing. Or the argument at all, it seems.


"Somehow Google has been allowed to shift the cost burden of policing its own platform to content owners."

Because that is how the law works. Because without safe harbor, the internet as we know it wouldn't exist. Goodbye user content, period. YouTube, Reddit, Twitch, Twitter, Facebook; they all cease to exist without safe harbor. Google is happy to let content producers bear the burden of policing the use of their content because it is the only reasonably functioning way to keep the internet intact. It also isn't Google's responsibility. If I upload Dark Side of the Moon to YouTube, Google didn't do anything wrong, I did. And it isn't Google's responsibility to make sure Dark Side isn't abused, it is Pink Floyd's job (or, likely, their label's). We can take Google out of the picture: what if you set up your own radio station and broadcast dark side? Would some random third party, like the electric company providing you power or the manufacturer of your equipment be responsible for making sure you aren't breaking copyright law by broadcasting dark side? Again, no, it is Pink Floyd's. And in this case, law enforcement.

So really, what you should be asking is why does law enforcement shrug off responsibility to enforcing copyright to the content producers? It's literally their job to enforce the law.


"If it takes a lot of money to operate this business model without ruining other peoples business, is it not simply a failed business model?"

Please apply this to content producers. If it is too expensive for them to police their copyright, haven't they simply failed as a business model? If Pink Floyd cannot afford to deal with infringing dark side uploads, they've failed. Right?

If the LA Times cannot attract enough viewers to pay the bills, they've failed. If Bild cannot get enough revenue without having to try and force Google to pay for indexing their website into their search engine, Bild failed. They are a failed business model. So why are you picking which business models are allowed to fail and which are allowed to succeed?


As I've said earlier: if news is an important function of society (which I do agree with) and it cannot sustain its business without more assistance, then just give it some government money. Don't force a completely separate, private entity give life support to the news agencies.
show deleted posts

Your Name:
Your Password:
Your Message:
Bookmark and Share