Pluralistic: 22 Oct 2020

Today's links

IDing anonymized cops with facial recognition (permalink)

As pandemic and climate emergency force the contradictions of capitalism to the breaking-point, the world's streets have erupted in ceaseless, ferocious protest. In a desperate bid to prolong their rule, elites have fielded increasingly cruel and violent police responses.

The cops are, to varying degrees, complicit. They have chosen to follow orders rather than risk their jobs (or even, in some cases, their safety from state retaliation).

The increasingly obvious injustice of the cause they fight for also increases the risk they bear.

There are three risks for the shock troops of late-stage capitalism:

I. the risk of official sanction by the state they fight for

II. the risk of punishment by a new regime should their cause fail

III. the risk of vigilante justice for the people they brutalize and murder

To reduce this risk, cops are going anonymous: not just wearing covid masks, but also removing their badges, nametags, and (notoriously in Portland), all insignia save generic windbreakers emblazoned POLICE, so even their agency affiliation is anonymized.

For every measure, there is a countermeasure. Networked authoritarianism has driven down the cost of facial recognition tools, and protesters have turned these tools on anonymized cops.

Kashmir Hill's NY Times story on the phenomenon is wild.

During the Hong Kong uprisings, Colin Cheung was arrested after he posted a video showing how he was identifying anonymized cops from online photos:

Artist Paolo Cirio posted an online exhibit called "Capture" with images of 4,000 French cops who participated in the crackdown of the Gilet Jaunes protests as a step toward automated identification (the photos were removed after government threats).

And in Portland, a self-taught programmer named Christopher Howell responded to police leadership's exhortation for officers to cover their nametags by developing facial recognition to reidentify law enforcement officers who took the advice.

Howell's project came to light when he responded to the city's call for comments on a proposal to ban facial recognition tools, a measure that was meant to curb authoritarian surveillance. He wanted to know if the rule also banned antiauthoritarian surveillance?

The city's lawyer confirmed that Howell's tools would be legal as the rule only banned organizations – not individuals – from using facial recognition.

According to Hill, the authorities are "not pleased."

ENDSARS (permalink)

People of Nigerian descent and human rights activists around the world have taken to the streets under the banner of #EndSARSProtest: a global protest movement over Nigeria's lawless, murdering Special Anti-Robbery Squad.

SARS was founded in 1984 in answer to a wave of property crimes, today, its founder Fulani Kwajafa says that it has "turned into banditry" – Amnesty International has documented 82 cases of torture, brutality and murder by SARS since 2017.

The current wave of protests was ignited by the public murder of a young man by SARS officers on Oct 8. President Muhammadu Buhari has disbanded the unit, but the criminals who served in it have been deployed elsewhere in Nigerian security forces, spreading the contagion.

The End SARS movement has five demands:

I. Release protesters

II. Justice for survivors of police violence and for families of the murdered

III. Independent oversight of police brutality complaints

IV. Retraining and psych evals for former SARS officers before they are allowed to serve again

V. A living wage for cops so they do not need to commit crimes in order to survive

#EndSARS protests have been met with brutality in Nigeria, including lethal police gunfire.

This video from Trevor Noah is an excellent backgrounder on the protests and their demands.

Companies target robots in disclosures (permalink)

To understand financialized snake-oil, you must understand Goodhart's Law: "Any measure can become a target." Goodhart's Law explains why something that works really well at the outset quickly turns into an arms-race with grifters.

For a recent example, think of Pagerank, the original secret sauce of Google Search. Larry Page had a key insight: a link from one web page to another page was an indicator of significance.

The web was written primarily by human hands and making links took work. If a web-writer linked to something on the web, it meant that they thought it was important. Thus, you could assess the relative significance of a web page by counting the number of links pointing to it.

This worked far better than rivals' methods – Altavista's idiotic keyword-counting, for example (a page is relevant to the query "cat" if the word "cat" and its synonyms appear frequently on that page; pages with the most "cat"s go at the top of the listing).

At the outset, inbound links were a great measurement of significance. But once inbound links became a way to game search-rank, they became a target, too: web-writers found ways to garner inbound links, from the relatively benign "webrings" to gross "linkfarms."

Today, Google's search ranking considers hundreds of "signals," locked in an arms race scammers who want to chaff or spoof the system.

The same thing is going on in finance.

Public companies have a regulatory duty to publish financial disclosures, which range from fanciful (Warren Buffet's annual letters) to dry. Finance analysts once carefully pored over these reports looking for clues to a company's fortunes.

As natural language parsing tools and machine learning improved, this process was automated. The robots that digested these reports didn't confine their analysis to the numbers: they also used "sentiment analysis" to try to guess at the mental state of the reports' authors.

Sentiment analysis is a notoriously garbage technology, grounded in low-quality, unreplicable psych research. Even when implemented by the biggest corporate R&D; labs, it is effectively ML graphology, pure pseudoscience.

The finance sector is full of superstitious nonsense. This is the industry whose "smartest investors" hand (literal) trillions to hedge funds that underperform simple index funds. It's not surprising that they were marks for slicksters selling sentiment analysis magic-beans.

Analysts' ratings control share-prices, and corporate executives derive the lion's share of their pay from stock in their own companies, and so corporate governance becomes a giant game of Goodhart's Law in which execs target the metrics that analysts rely on.

Share-based compensation is supposed to align managers' interests with the shareholders: instead, it aligns their interests with the prejudices of analysts.

Think of how Frontier went bankrupt after leaving $800m in profits on the table because the spending needed to get it was dispreferred by the analysts who controlled the company's share price (and thus its execs' take-home pay):

Predictably, then: the modern financial disclosure is optimized for machine readability by analysts' robots, and it uses language that is designed to be interpreted as positive by sentiment analysis systems:

"Firms with high expected machine downloads manage textual sentiment and audio emotion in ways catered to machine and AI readers, such as by differentially avoiding words that are perceived as negative by computational algorithms as compared to those by human readers, and by exhibiting speech emotion favored by machine learning software processors."

-"How to Talk When a Machine is Listening: Corporate Disclosure in the Age of AI," Cao, Jiang, Yang & Zhang, NBER

The most grimly hilarious part of this: it will doubtless be offered as evidence for sentiment analysis, when the real lesson is a tautology: "If you speak in words that algorithms interpret as positive, the algorithms will interpret the speech in a positive light."

It's the finance version of self-driving car grifters who insist that their vehicles will be safe for pedestrians once we teach all the pedestrians to behave in ways that the cars can correctly interpret.

US border cruelty, powered by Google cloud (permalink)

In 2018, 20,000 googlers walked off the job in a protest over the company's tolerance for (and rewards to) sexual predators.

The walkout was the culmination of a long year of moral reckonings for Google workers, amid revelations that the company was secretly building AI tools to help the Pentagon's drone program and search tools to help the Chinese state suppress and spy on dissidence.

And while the walkouts killed these controversial projects, forced some exec resignations, and ended the company's use of binding arbitration clauses in employment contracts, the key organizers were forced out:

Ultimately, the company's commitment to networked authoritarianism and profits over human rights has continued to overpower any fear of worker walkouts. How else to explain yesterday's news that the company is helping CBP police the US-Mexican border?

As Lee Fang and Sam Biddle write for The Intercept, Google's role in providing cloud AI services for the "virtual border" was laundered through Thundercat Technology, a fact revealed by Jack Poulson, an AI scientist who quit Google over its Chinese censorship project.

Google's Cloud and AI systems are being integrated with tools sold by Anduril Industries, the company founded by the neofascist Palmer Luckey, who was made a billionare by Facebook when they bought his Oculus VR startup at a hugely inflated price.

Luckey is a sociopath pariah who has been caught secretly funding white nationalist movements. His company is allied with Palantir, a company that produces tools for automating racial oppression and mass surveillance.

Luckey boasts that his company's recruiting materials court engineers who want to build weapons. He has donated $1.7m to Donald Trump's re-election campagin. He is precisely the kind of unsavory would-be war-criminal that Google leadership promised it wouldn't associate with.

The history of tech is strewn with technologists who turned a blind eye to the use of their inventions to commit hideous war crimes, such as IBM's complicity in Nazi death camps.

Google's use of a cutout to disguise its participation in the ethnic cleansing project on the US sourther border is eerily reminiscent of IBM's systems for laundering its complicity with extermination camps.

Free the law of Wisconsin (permalink)

Wisconsin's judges work with the state university system to produce standard "jury instructions" that are key to how the state's juries interpret the law when they deliberate. Judges then adapt these for each trial based on the facts of the case.

Jury instructions are produced at public expense, by public employees, including members of the judiciary in the course of their normal duties. By any reasonable standard, these should free and in the public domain.

A bedrock of law – dating to the Magna Carta – is that it must be public to be the law.

But neither the standard instructions nor the case-based versions are available to the public. They are sold for $500 a pop, festooned with copyright notices and dire warnings.

This is red meat for rogue archivist Carl Malamud, who has made a career out of publishing copyrighted laws, including a recent victory at the Supreme Court over whether the state of Georgia's copyright claims over its laws were valid.

After being contacted by a WI lawyer – who pointed out that law firms were relying on out-of-date versions of jury instructions because staying current costs $500/yr – Malamud wrote to the judges and the law profs behind the jury instructions.

Malamud put them on notice that they should be publishing this vital part of the state's law, and that if they didn't, he would. The judges never wrote back, and the law profs turned it the matter over to their university's counsel.

After some delays, the university's lawyer wrote back to Malamud and told him that they were getting out of the business of copyrighting the law, and the WI judicial conference would be making these public by February.

The letter claims they're doing this because it's right, but adds Wisconsin is within its rights to continue to claim copyright over the law.

To understand why this is a terrible and wrong idea, check out Public Resource's excellent explainer video:

The video features EFF legal director Corynne McSherry, who has represented Malamud on many occasions. Her explanation (about 25 mins in) is an excellent, <10m primer on the need for the law to be public.

This day in history (permalink)

#5yrsago Fable Comics: anthology of great comics artists telling fables from around the world

#5yrsago Son of Dieselgate: second line of VWs may have used “defeat devices”

#5yrsago DHS admits it uses Stingrays for VIPs, vows to sometimes get warrants, stop lying to judges

#5yrsago Half of Vanuatu’s government is going to jail

#5yrsago Obama administration petitions judge for no mercy in student debt bankruptcy

#1yrago Margaret Atwood’s “The Testaments”: a long-awaited Handmaid’s Tale sequel fulfills its promise

#1yrago Materiality: a new science fiction story for the Oslo Architecture Triennale about sustainable, green abundance

#1yrago Ernst and Young subjected women employees to “training” about keeping the company’s men happy

#1yrago NJ school district bans indebted students from prom and field trips, refuses offer to pay off lunch debt

Colophon (permalink)

Today's top sources: Four Short Links (, Slashdot (

Currently writing: My next novel, "The Lost Cause," a post-GND novel about truth and reconciliation. Yesterday's progress: 530 words (75542 total).

Currently reading: Harrow the Ninth, Tamsyn Muir

Latest podcast: Someone Comes to Town, Someone Leaves Town (part 17)

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