Pluralistic: 12 Dec 2020

Today's links

Youtube, fair use, competition, and the death of the artist (permalink)

Today we have three major record labels, the meanest, most voracious gobblers-up of the competition, companies that clawed their way to global dominance through absolute ruthlessness, particularly to the musicians who did the work that brought in the money.

For example: the Beatles earned $0.01/record.

Split four ways.

But 15% of that was creamed off by the label to account for "promotional copies" that they later admitted they sold at full price without giving the Beatles their one penny of royalties.

When Napster and other digital distribution platforms appeared, many musicians were angry about them, but many were hopeful: was this – finally – a way to be free of the labels?

It was!

For a while.

The subsequent record industry lawsuits combined with runaway mergers and acquisitions in both tech (5 giants) and entertainment (4 studios, 4ish publishers, 3 labels) killed the dream of a pluralistic, fair alternative to the content oligarchy.

This was confirmed when Youtube launched its music service in the early part of the past decade. After hammering out a deal with the Big Four (now three) labels, YT told all the independent artists and labels that this would be their deal too – or they could leave Youtube.

In other words, Youtube was now an honorary member of the Big Four (now three) and every musician in the world who wanted access to the monopoly audio-video distribution platform would have to toe the Big Four's (now three) line…or else.

The fusion of Youtube with the big entertainment companies includes a complete disregard for fair use – the limitations and exceptions to copyright that are so necessary to free speech, criticism, and new artistic expression.

It's hard to square copyright with the First Amendment ("Congress shall make no law.. abridging the freedom of speech"). I wrote this, it attracted a copyright at the moment of fixation, and now the US government gives me to the power to bar you from repeating it.

How can a government that is constitutionally barred from abridging your speech also have a law telling you what you're allowed to say?

The Supreme Court explored this in Eldred v Ashcroft, with RBG calling fair use copyright's "safety valve."

Fair use is how you can ban people from saying stuff without abridging their free speech: it allows for transformation, criticism, commentary, education, parody, archiving and a host of other uses. Without fair use, copyright is unconstitutional censorship.

In 2007, Viacom sued Youtube for $1b; the court released emails between Viacom execs admitting that their goal was to "steal" Youtube and take it over (the emails were very sweary and included vicious fights about which exec would get to run YT)

As the suit was making its way through the courts, Youtube launched Content ID, its automated takedown system. Content ID – which cost $100m to build and run – allowed select rightsholders to "claim" certain audio and video elements as their copyrights.

Content ID scours uploaded videos for matches to claimed works, then (depending on rightsholders' preferences), Content ID either removes the matching video, takes the money it is generating, or (if it's ad-free) puts ads all over it and gives the money to the rightsholder.

And Content ID can't figure out if a use is fair or not. No algorithm can. How software tell if something is parody? If it's commentary? If it's sufficiently transformative to constitute a fair use? It can't, and so Content ID doesn't.

What do you get when you mix automated takedown, entertainment and tech monopolies and the absolute annihilation of fair use?

The contemporary arts world.

In a superb white paper, my EFF colleague Katharine Trendacosta painstakingly documents how Youtube transfers gigantic sums from working artists to rent-collecting rightsholders in a system that incentivizes false copyright claims and punishes those who appeal automated verdicts.

The paper starts by documenting how Content ID really works, and shows how the pretty, streamlined, easy to follow flowcharts that YT publishes to prove that it has a fair and transparent systems are a sham.

The reality is a giant, gnarly hairball of rules, threats, ultimatums and dire punishments, where videos that pass muster one day can be taken down the next, and where challenging the robot (or a rightsholder) can annihilate your Youtube account – and thus your artistic career.

It's a system where creators producing major video essays can upload them in five minute chunks to YT to check for copyright claims, find none, piece the whole feature together and upload it – only to get five copyright claims.

It's a system where a ten-hour video of white noise attracts six copyright claims – one for an "infringement" that allegedly lasted for less than one second.

Trendacosta's paper revolves around deep interviews with three very different Youtubers.

The first is Hbomberguy, a long-form video essayist with 600k subscribers, who spends weeks playing blind man's bluff with Content ID, slicing more and more sections out of his video to get rid of copyright claims (and doing it over again when the algorithm changes).

The second is Todd Nathanson, a musicologist who produces videos that dive deep into one-hit wonders and other musical oddities. He has given up on getting any of the money his wildly popular videos generate through YT ads; it's impossible to fend off the Content ID robots.

Instead, Nathanson relies on Patreon subscribers to pay his bills, and allows the rightholders who lay claim to the money from his legal, canonical-fair-use musical excerpts to simply steal the money he is entitled to.

Finally, there's Lindsay Ellis, a bestselling novelist with 1m subscribers who has been shunted from one service to the next as all of Youtube's competitors were driven out of business, landing on Youtube after she ran out of alternatives.

Ellis's high profile means she can sometimes actually reach a human being at Youtube to discuss copyright claims against her work, but they are monumentally unhelpful. Ellis knows an awful lot about copyright, but that's actually a detriment when it comes to Content ID.

That's because defending yourself to the Content ID system means that rightsholders can deny your defense and generate "copystrikes" against your account. Rack up three strikes and your account gets suspended or even deleted.

Ellis wonders "why I bothered playing by the rules all these years because fair use doesn’t matter. Content ID is all that matters."

Youtube's Content ID statistics are opaque and out of date, but Trendacosta estimates that the system rendered automated judgements in 122,500,000 copyright claims in 2017, a number that has surely climbed since.

Once, creators dreamed of the internet as an escape hatch from the airtight cell that the Big Four (now three) labels had crammed all the world's music into. Today, most creators have been corralled into a single online platform.

That platform colludes with the Big Three to misappropriate the incomes of artists who never signed their one-sided record deals, and the ripoff continues with a long tail of penny-ante grifters who claim copyrights over silence, white noise and birdsong.

Carbon offsets are bullshit (permalink)

The third stage of grief is bargaining. When the climate emergency hove into sight, we started with denial (Exxon's criminal suppression of its own research on the looming crisis #ExxonKnew).

From there, it was onto anger: slurs like "treehugger" and performative Drill-Baby-Drill-style incursions into protected lands.

But now we're into bargaining: can't we please keep the rapacious, immoral system that jeopardized our planet and species?

In other words: carbon offsets.

Rather than punishing firms and executives who sacrificed our planet and species for their short-term gains, we'll just give them a path to realizing even handsomer rewards for not murdering us all.

We decide how much carbon each country can emit, and then we let them trade those allotments between one another. If you forgo your carbon allowance – by not chopping down a forest, say – you get a credit that you can sell to a planet-destroying cruise-ship operator.

This is supposed to provide the incentives needed to get the market to reduce its own carbon footprint, by putting a price on carbon and then turning capitalism loose to reduce its obligation to pay that price with all the ingenuity that the human mind can muster.

"Incentives matter" is the rallying cry of the right. They're correct. And the incentive here is to reduce the amount paid for carbon by any means necessary, which is exactly what the market did.

To understand what happened next, consider a labor law that causes companies to pay fines for hurting their workers. Depending on how the fines are structured, they might incentivize employers to make their workplaces safer.

But there are other ways to respond. If the fine is low enough, it might be cheaper to maim workers than to upgrade your physical plant ("a fine is a price"). It might be cheaper to buy off the regulators and get the fines reduced.

It might be cheaper to hire Pinkertons to spy on maimed workers and find ways to discredit and smear them so the fines are nullified. Regulation works best in the presence of personal liability for corporate execs, steep fines, and weakened corporate power.

Carbon offsets have none of the above. What happened next is exactly what was predicted to happen: polluters were incentivized to find the cheapest possible carbon offsets, which turn out to be the most meaningless ones.

To do so, they only had to perform the relatively straightforward chore of corrupting the largest environmental charity in America: The Nature Conservancy, a 69-year-old nonprofit that buys up endangered lands and prevents them from being logged or despoiled.

Buying and preserving endangered lands generates carbon credits, and those carbon credits can be sold onto the world's worst polluters, which generates more capital to put into buying and protecting endangered lands. Sounds good, right?

But you can generate even more carbon credit profits by cheating: taking land that's not endangered and falsely claiming carbon credits for it, then flogging them to the oil, travel, and finance industry as imaginary offsets for very real planet-destroying carbon emissions.

Last year, the Nature Conservancy generated $932m in revenues. As Ben Elgin writes for Bloomberg, that's more than the next three largest US environmental charities combined.

The Conservancy insists that the credits it generates come from real preservation efforts.

But that's not supported by the data. Take the half-million tons' ($2m) worth of carbon offsets the Conservancy generated and sold to JP Morgan and others for not logging 2380 acres of Pennsylvania forest.

That forest is already a conservancy protected by the Hawk Mountain Sanctuary Association, which does not permit logging. The Association was misled into believing that its share of the credits would come from planting saplings.

The credits came from the absurd assumption that the land would be logged The American Carbon Registry – which oversees the credits – says that you can't tell if the Association mightn't have changed its mind and logged the lands it's protected for 90 years.

The Conservancy has a long history of cozying up to large polluters and allowing them to buy their way to respectability ("The only problem with tainted money is there tain't enough of it" -Patrick Noonan, Conservancy President, 1973-80).

But procuring and selling phony carbon offsets doesn't just change how people feel about Conservancy partners like Exxon and Dow – it changes how they perceive the climate impact of their everyday activities.

Maybe you've heard that cruise ships are destroying the planet, but if Disney Cruises promises you that the offsets it bought from the Conservancy more than make up for it, the cruise starts to look pretty good.

Disney and the Conservancy know the offsets are bogus, but you don't. Market trufans call this "information asymmetry" and understand that over time, it leads to a "market for lemons" in which only defective goods are bought and sold.

It's not a new idea. Gresham's Law (1860) says that "bad money drives out good" – the credits generated by "preserving" land that no one was going to log anyway are cheap to produce.

Meanwhile, good credits – like those that Stripe gets from Climeworks AG for carbon sequestration at $775/ton – cost more than fake ones for not logging a forest that no one was ever going to log.

Incentives matter. Gresham's Law produces a market for lemons.

All of this is firmly within market orthodoxy. That's the core of the bargaining stage, after all: "Can't we just keep doing exactly what we've always done and hope for a better outcome?"


Outgoing Facebookers blast the company (permalink)

There's two reasons that Facebook is such a cesspit of disinformation, unproductive flamewars and trolling:

First, it's too fucking big. That is, not only is Zuck monumentally unqualified to control the social lives of 2.6b people – no one is.

Second, they don't really want to shut down the bad stuff. Paid disinformation is revenue, after all. That's why FB (which insists it should gobble up its competitors because scale is "efficient") outsources the bulk of its "user safety" function.

Those outsource subcontractors don't work for Facebook because they believe in its mission – they take the jobs because they don't have any leverage, which is why they're paid low wages and denied the benefits that FB lavishes on its in-house tech employees.

But high wages and free kombucha in the campus mini-kitchen only go so far. Those high-paid workers are running up against the limits of their ability to rationalize their role in one of the world's most toxic enterprises.

The ones that joined to fix Facebook from the inside have overwhelming evidence that Facebook doesn't actually want to fix its problems, particularly disinformation.

Reality has a leftist bias, so any crackdown on disinformation will disproportionately affect conservatives.

When that happens, Ted Cruz gets angry at Zuck and drags him into the Senate. Plus, Zuck really enjoys the company of far right assholes, and his version of "listening to both sides" boils down to "I meet with Stormfront and the RNC."

There's a tradition for departing Facebookers: a "badge post" – a farewell note to the company summarizing the lessons they've learned in their tenure. These days, badge posts are apt to be full-throated denunciations of the company:

These employees want their colleagues to know that Facebook's toxic stew isn't just down to the enormity of its task: there are plenty of instances in which Facebook CAN stop harmful (even lethal) disinfo, and chooses not to.

The company has disbanded its civic integrity team (mission: "protect the democratic process…stopthe spread of viral misinformation and fake accounts") after spiking the team's plan to block provable disinformation posted by Trump's account.

The company failed to act on its team's warnings about armed right-wing militias that were planning murder-sprees, including the plans that led to the murders of protesters in Kenosha, WI.

It ignores the data generated by its own researchers that attribute the bulk of disinformation to powerful, politically connected far-right outlets and figures like Breitbart, Fox, the Daily Caller, Ben Shapiro and Donald Trump.

That blind eye doesn't just mean that these accounts' disinformation spreads – it also generates income for them by helping them solicit donations and payments.

Facebook's internal research consistently shows that fighting disinformation would have the largest impact on right-wing causes, and the departing scientists and techies say that the fear of political retaliation prevents FB from acting on its own data.

The departing employees' badge posts say that working for Facebook is "embarrassing" and call the company's vision "a dystopia." They dismiss claims that automation will solve FB's problems as self-serving delusions.

One described FB as "a net negative effect on the quality of political discourse."

All of this reminds me of Robert Oppenheimer, the brilliant physicist and manager who led the Manhattan Project and created the first atom bomb, an act he regretted instantly and forever.

Oppenheimer turned away from that first test blast and quoted the Bhagavad Gita: "Now I am become Death, the destroyer of worlds." He spent the rest of his career trying to end nuclear proliferation.

Imagine if Oppenheimer had yielded to his moral sensibilities before he helped usher in the atomic age.

Moral reckonings are definitely in season in 2020. People are taking stock of the trade-offs they made to get by – or get rich – and wondering if they were worth it.

Our ability to rationalize our way into terrible situations – and the difficulty we face in reasoning our way out of them – is the theme of my novel ATTACK SURFACE, the third Little Brother novel, which just came out.

It's an urgent question: what does it take to get people to confront their rationalizations…and what do we expect them to do then?

This day in history (permalink)

#15yrsago Homeland Security: Mini-golf courses are terrorist targets

#5yrsago Britons will need copyright licenses to post photos of their own furniture

#1yrago Facebook promised to provide academics data to study disinformation, but their foot-dragging has endangered the whole project

#1yrago Family puts Ring camera in children’s room, discovers that hacker is watching their kids 24/7, taunting them through the speaker

Colophon (permalink)

Today's top sources: Slashdot (

Currently writing: My next novel, "The Lost Cause," a post-GND novel about truth and reconciliation. Friday's progress: 536 words (93097 total).

Currently reading: The City We Became, NK Jemisin

Latest podcast: Daddy-Daughter Podcast, 2020 Edition

Upcoming appearances:

Recent appearances:

Latest book:

This work licensed under a Creative Commons Attribution 4.0 license. That means you can use it any way you like, including commercially, provided that you attribute it to me, Cory Doctorow, and include a link to

Quotations and images are not included in this license; they are included either under a limitation or exception to copyright, or on the basis of a separate license. Please exercise caution.

How to get Pluralistic:

Blog (no ads, tracking, or data-collection):

Newsletter (no ads, tracking, or data-collection):

Mastodon (no ads, tracking, or data-collection):

Twitter (mass-scale, unrestricted, third-party surveillance and advertising):

Tumblr (mass-scale, unrestricted, third-party surveillance and advertising):

When life gives you SARS, you make sarsaparilla -Joey "Accordion Guy" DeVilla