Pluralistic: 14 May 2020

Today's links

Restaurants won't let gig drivers pee (permalink)

The "gig economy" companies are classic rentiers: in the guise of operating a "two-sided market" (restaurants on one side, delivery people on the other, say), they're actually just siphoning off everyone else's margin, slowly draining them dry.

But there's one thing the gig economy can't drain, and that's drivers' bladders.

Restaurateurs, legitimately angry at the third-party delivery apps that are destroying them through dirty tricks and rent-seeking, are denying drivers bathroom access.

It's a classic bit of misdirection, a game of "let's you and he fight." The delivery companies maintain the pretense that they're not employers and thus not required to provide bathrooms for their employees.

The restaurateurs can't strike back at distant corporate execs, nor can they do anything to the apps themselves, so they grab hold of the only tangible element of the system that's destroying them – the drivers, who are also being destroyed by the same system.

Execs have long made up unreasonable policies and then turned frontline staff out to absorb the anger and frustration of customers, while remaining insulated from any of that anger. Now it's happening to employees who deal with suppliers, too.

Meanwhile, everybody poops. And pees. Drivers can't make ends meet unless they pull 12-hour shifts. As the City of San Francisco has stubbornly refused to realize for 30 years, the fact that you don't have anywhere for people to poop doesn't mean they stop pooping.

Which is why the drivers are crouching in alleys, pissing in cups in their cars, etc.

And then handling your dinner.

Facebook's "backfire effect" junk science (permalink)

There are good reasons to worry about Facebook's "fact checking" efforts, primarily that the company is a raging garbage fire that destroys everything it touches and only has two approaches to solving any problem:

  1. Make it fully automated in a way that guarantees that there will be innumerable false positives in which legitimate material is erroneously censored, with no effective means of appeal, even as dedicated trolls exploit the system's blind spot to carry on as normal;

  2. Hire boiler-rooms full of low-waged workers to review horrific materials and make judgment calls that require context they don't have and can't get, until they're so traumatized they literally develop PTSD and sue the company for psychiatric care.

But there's one reason NOT to worry about Facebook factchecking, and that's the "Backfire Effect," a discredited psychological principle that holds that when learn facts that challenge their worldview, they double down on their false beliefs.

The original experiments that established the Backfire Effect as a bedrock of social psychology have spectacularly, repeatedly failed to replicate:

Nevertheless, a Facebook exec told Stat that the reason the company is holding back on thorough factchecks of covid conspiracy theories is they're worried about the Backfire Effect.

Again, I don't trust Facebook to do anything well, let alone factchecking. But among the things Facebook does badly, apparently, is "understanding how factchecking works."

This is well-put in an op ed by Ethan Porter and ThomasWood, authors of "False Alarm: The Truth about Political Mistruths in the Trump Era," a peer-reviewed book from Cambridge University Press.

"By our count, across experiments involving more than 10,000 Americans, fact-checks increase the proportion of correct responses in follow-up testing by more than 28 percentage points."

And they did a new study to show that this would work on FB, too: "Across all issues, people who had seen misinformation and then a related fact-check were substantially more factually accurate than people who had only seen the misinformation."

"Prior research has found that, on social media, fake news is disproportionately shared by older, more conservative Americans. In our study this group did not show any special vulnerability to backfire effects. When presented with fact-checks they became more accurate too."

Learn magic (permalink)

I'm a longstanding fan of The Jerx, a contrarian, anonymous "social magic" blog that proposes a way of performing conjuring and mentalist effects for your friends or single individuals that cuts through much of the pretense in typical magic acts.

Andy, The Jerx's creator, is pretty relentless (and often very funny) in his dunks on other magicians, and especially on businesses that purport to teach you magic, or that sell you gimmicks to perform with.

So I was excited to see a rare endorsement of an online (free!) magic class, Eric Hu's Friends and Astronauts, whose mission is to "teach magic tricks for free (previously unpublished) in support of each other, the magic community, and covid relief."

The tricks will come in time-limited collections from different performers and creators. Once one is done, it'll be removed, so you have to follow along if you want to learn all the effects.

The focus is "stuff you can tinker with while we’re quarantined. Sleights, moves, DIY gimmicks, camera tricks, self-working tricks."

Here's collection #1: "3Fast, 1 Furious," with work from Tatanka Tan, Ryan Plunkett, Joel Greenwich, and more.

Pythonesque small claims (permalink)

In Davy v. Kidwai, a small claim made at British Columbia’s Civil Resolution Tribunal, a fellow who bought a parrot diagnosed with a deadly illness argued that the merchant who sold it to him owed him a refund.

This may strike you as a familiar tale.

The parrot Tiberius sickened with Psittacine Beak & Feather Disease less than a year after Kidwai sold him to Davy.

The court found that Davy had not committed fraud when he averred that the parrot was healthy (Tiberius tested positive for PBFD three weeks after the sale).

But the court did find in Davy's favor on the question of whether Tiberius's quick demise violated the implied warranty of merchantability – the idea that a product is "fit for ordinary use" at time of sale and "durable for a reasonable period of time."

The parties agreed that a healthy Eclectus parrot should live 30-40 years, and the court held that Tiberius's illness meant he was likely to live for less than one more year. Relying on an earlier case about a "defective puppy," the court awarded Davy $2500.

(Image: Eduardo Unda-Sanzana, CC BY)

Modern monetary theory's moment has arrived (permalink)

Modern monetary theory is an economic lens for understanding how money works. It starts from the commonsense assertion that national governments are money-issuers, and the rest of us are money users, and that money works differently when you are in charge of creating it.

But this challenges accepted wisdom, like the idea that national deficits cause inflation and saddle our descendants with debt. If governments are the source of money, then they don't tax us in order to spend. They spend (which puts money into the economy) and then tax back.

If the government runs a "balanced budget" that means it's taxing as much money of out existence as it is spending into existence, leaving behind no money for the rest of us to save or spend. Balanced budgets starve the private sector of the money it needs to operate.

When that happens, banks create private money (lending money that they don't have on deposit, something that they are allowed to do as part of their deal with the national government), and then they get to charge interest for those loans.

So the finance sector hates public money – which benefits everyone – and loves private money – which benefits them. But of course, banks inevitably overspend and since they aren't national governments, they risk defaulting, so they need public money to bail them out.

Governments can't default on debts owed in currencies they issue. They just type more zeroes into a spreadsheet at the national bank and they can pay off any debt they want. They are not "monetarily constrained."

The constraint on governments is resources, not money. If the government tries to buy things the private sector is using, it gets into a bidding war and prices go up. That's inflation.

Likewise, if the private sector has too much money, it bids against itself and again, prices go up – inflation (think of the 2008 crisis, when banks created too much money and we got a speculative housing bubble, leaving governments to create money to bail them out).

Governments don't tax to balance budgets: they tax to remove money from the system to fight inflation. Progressive taxation – higher taxes on the rich – minimize the privation of having some of your money annihilated, and it also weakens the 1%'s power to influence politics.

(Governments have done other things to fight inflation when they felt the need to buy things the private sector wanted, like during WWII: war bonds that locked up the money pumped into economy with war spending, and rationing, which limited bids on war materiel).

MMT has seen many vindications since the 2008 crisis, when the bailouts didn't produce the inflation we'd been warned of. Then, when the GOP tax plan passed and left the fed budget trillions in deficit, we were warned that the USG had no more resources to fight a future crisis.

But when the crisis hit, Congress was able to spend trillions into existence with the stroke of a pen, without inflation, because the private sector no longer wanted the labor of a large fraction of the workforce. The existence of the tax deficit was irrelevant.

The reason to be alarmed by the deficit left behind by the tax bill is that it will make rich people richer, not that it will deprive the federal government of money. That's like saying giving free Itunes gift cards to plutes starves Apple of cards for the rest of us.

All of this is beautifully explained in a new NPR Marketplace interview with Stephanie Kelton, MMT's most articulate, thoughtful and knowledgable spokesperson.

Kelton's interview is tied to her upcoming (June) release, "The Deficit Myth." The book couldn't be more timely. With the private sector no longer interested in the labor of so many of us, nor the goods we can produce, MMT is the lens we need.

Pandemics shatter AI's intrinsic conservativism (permalink)

One of my favorite insights on Machine Learning comes form Molly Sauter's 2017 essay Instant Recall, which describes ML's intrinsic conservativism. Having been trained on historical data, ML can only make good predictions if the future's like the past.

That's why your phone's predictive text nudges you to follow up "My" with "darling" if you habitually text your spouse with that salutation. When you type a word it's never seen, it autocompletes with a statistically normal following word.

I took a stab at talking about some of the implications of this in an LA Review of Books essay last January:

One interesting side-effect of the pandemic is that it magnifies this conservative bias in ML: because we're acting in unprecedented ways, the ML-based predictive systems are misfiring badly ("automation is in a tailspin").

Supply-chain forecasting tools are ordering both too much of things, and too little. Recommendation systems haven't figured out that "in stock" is the most important criteria to sort by. Shoppers aren't interested in big ticket items.

Some of these behaviors are so unusual that they're tripping fraud-detection systems that apply the brakes to an already broken system ("You never bought gardening or bread-making stuff before, why are you ordering so much now?").

And when big players like Amazon alter their algorithms to correct for this, it adds even more chaos to the downstream systems that are trying to predict what Amazon will do.

People who claim to know the future are always either con-artists or delusional. The addition of abstruse stats to the prediction provides the discredited profession of fortune-telling with a veneer of empirical facewash, but it doesn't bring us closer to precognition.

So much of what we call "accurate prediction" is just "forcing/nudging people to act the way they used to" – I can predict with 100% accuracy that you will hold the same posture for the next five minutes, if I'm allowed to shoot you through the heart before the clock starts.

As our conservative AI overlords encounter the reality of an unprecedented situation, the difference between coercion and prediction is becoming a lot clearer.

(Image: Johnny Jet, CC BY)

This day in history (permalink)

#15yrsago Why writers should stop worrying about "ebook piracy"

#15yrsago Pro file-sharing seal for CDs×

#10yrsago The People's Manifesto: Mark Thomas and friends' suggestions for UK political reform

#10yrsago Barbie-themed hotel rooms for three year olds that cost €1,600/night

#5yrsago US Passport Agency contractors harvested Americans' data for identity theft

#1yrago Lawyer involved in suits against Israel's most notorious cyber-arms dealer targeted by its weapons, delivered through a terrifying Whatsapp vulnerability

#1yrago Three years after the Umbrella Revolution, Hong Kong has its own Extinction Rebellion chapter

#1yrago Collecting user data is a competitive disadvantage

#1yrago DOJ accuses Verizon and AT&T; employees of participating in SIM-swap identity theft crimes

#1yrago AT&T; promised it would create 7,000 jobs if Trump went through with its $3B tax-cut, but they cut 23,000 jobs instead

#1yrago Jury awards $2b to California couple who say Bayer's Roundup weedkiller gave them cancer

#1yrago A year after Meltdown and Spectre, security researchers are still announcing new serious risks from low-level chip operations

Colophon (permalink)

Today's top sources: John Weeks, Naked Capitalism (

Currently writing: My next novel, "The Lost Cause," a post-GND novel about truth and reconciliation. Yesterday's progress: 524 words (15319 total).

Currently reading: Facebook: The Inside Story, by Steven Levy.

Latest podcast: Rules for Writers (

Upcoming books: "Poesy the Monster Slayer" (Jul 2020), a picture book about monsters, bedtime, gender, and kicking ass. Pre-order here:

"Attack Surface": The third Little Brother book, Oct 20, 2020.

"Little Brother/Homeland": A reissue omnibus edition with a new introduction by Edward Snowden:

This work licensed under a Creative Commons Attribution 4.0 license. That means you can use it any way you like, including commerically, 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

One thought on “Pluralistic: 14 May 2020”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.