I'm on a 20+ city book tour for my new novel PICKS AND SHOVELS. Catch me in SAN DIEGO at MYSTERIOUS GALAXY on Mar 24, and in CHICAGO with PETER SAGAL on Apr 2. More tour dates here.
Even by Amazon standards, this is extraordinarily sleazy: starting March 28, each Amazon Echo device will cease processing audio on-device and instead upload all the audio it captures to Amazon's cloud for processing, even if you have previously opted out of cloud-based processing:
It's easy to flap your hands at this bit of thievery and say, "surveillance capitalists gonna surveillance capitalism," which would confine this fuckery to the realm of ideology (that is, "Amazon is ripping you off because they have bad ideas"). But that would be wrong. What's going on here is a material phenomenon, grounded in specific policy choices and by unpacking the material basis for this absolutely unforgivable move, we can understand how we got here β and where we should go next.
Start with Amazon's excuse for destroying your privacy: they want to do AI processing on the audio Alexa captures, and that is too computationally intensive for on-device processing. But that only raises another question: why does Amazon want to do this AI processing, even for customers who are happy with their Echo as-is, at the risk of infuriating and alienating millions of customers?
For Big Tech companies, AI is part of a "growth story" β a narrative about how these companies that have already saturated their markets will still continue to grow. It's hard to overstate how dominant Amazon is: they are the leading cloud provider, the most important retailer, and the majority of US households already subscribe to Prime. This may sound like a good place to be, but for Amazon, it's actually very dangerous.
Amazon has a sky-high price/earnings ratio β about triple the ratio of other retailers, like Target. That scorching P/E ratio reflects a belief by investors that Amazon will continue growing. Companies with very high p/e ratios have an unbeatable advantage relative to mature competitors β they can buy things with their stock, rather than paying cash for them. If Amazon wants to hire a key person, or acquire a key company, it can pad its offer with its extremely high-value, growing stock. Being able to buy things with stock instead of money is a powerful advantage, because money is scarce and exogenous (Amazon must acquire money from someone else, like a customer), while new Amazon stock can be conjured into existence by typing zeroes into a spreadsheet:
But the downside here is that every growth stock eventually stops growing. For Amazon to double its US Prime subscriber base, it will have to establish a breeding program to produce tens of millions of new Americans, raising them to maturity, getting them gainful employment, and then getting them to sign up for Prime. Almost by definition, a dominant firm ceases to be a growing firm, and lives with the constant threat of a stock revaluation as investors belief in future growth crumbles and they punch the "sell" button, hoping to liquidate their now-overvalued stock ahead of everyone else.
For Big Tech companies, a growth story isn't an ideological commitment to cancer-like continuous expansion. It's a practical, material phenomenon, driven by the need to maintain investor confidence that there are still worlds for the company to conquer.
That's where "AI" comes in. The hype around AI serves an important material need for tech companies. By lumping an incoherent set of poorly understood technologies together into a hot buzzword, tech companies can bamboozle investors into thinking that there's plenty of growth in their future.
OK, so that's the material need that this asshole tactic satisfies. Next, let's look at the technical dimension of this rug-pull.
How is it possible for Amazon to modify your Echo after you bought it? After all, you own your Echo. It is your property. Every first year law student learns this 18th century definition of property, from Sir William Blackstone:
That sole and despotic dominion which one man claims and exercises over the external things of the world, in total exclusion of the right of any other individual in the universe.
If the Echo is your property, how come Amazon gets to break it? Because we passed a law that lets them. Section 1201 of 1998's Digital Millennium Copyright Act makes it a felony to "bypass an access control" for a copyrighted work:
That means that once Amazon reaches over the air to stir up the guts of your Echo, no one is allowed to give you a tool that will let you get inside your Echo and change the software back. Sure, it's your property, but exercising sole and despotic dominion over it requires breaking the digital lock that controls access to the firmware, and that's a felony punishable by a five-year prison sentence and a $500,000 fine for a first offense.
The Echo is an internet-connected device that treats its owner as an adversary and is designed to facilitate over-the-air updates by the manufacturer that are adverse to the interests of the owner. Giving a manufacturer the power to downgrade a device after you've bought it, in a way you can't roll back or defend against is an invitation to run the playbook of the Darth Vader MBA, in which the manufacturer replies to your outraged squawks with "I am altering the deal. Pray I don't alter it any further":
The ability to remotely, unilaterally alter how a device or service works is called "twiddling" and it is a key factor in enshittification. By "twiddling" the knobs and dials that control the prices, costs, search rankings, recommendations, and core features of products and services, tech firms can play a high-speed shell-game that shifts value away from customers and suppliers and toward the firm and its executives:
https://pluralistic.net/2023/02/19/twiddler/
But how can this be legal? You bought an Echo and explicitly went into its settings to disable remote monitoring of the sounds in your home, and now Amazon β without your permission, against your express wishes β is going to start sending recordings from inside your house to its offices. Isn't that against the law?
Well, you'd think so, but US consumer privacy law is unbelievably backwards. Congress hasn't passed a consumer privacy law since 1988, when the Video Privacy Protection Act banned video store clerks from disclosing which VHS cassettes you brought home. That is the last technological privacy threat that Congress has given any consideration to:
This privacy vacuum has been filled up with surveillance on an unimaginable scale. Scumbag data-brokers you've never heard of openly boast about having dossiers on 91% of adult internet users, detailing who we are, what we watch, what we read, who we live with, who we follow on social media, what we buy online and offline, where we buy, when we buy, and why we buy:
To a first approximation, every kind of privacy violation is legal, because the concentrated commercial surveillance industry spends millions lobbying against privacy laws, and those millions are a bargain, because they make billions off the data they harvest with impunity.
Regulatory capture is a function of monopoly. Highly concentrated sectors don't need to engage in "wasteful competition," which leaves them with gigantic profits to spend on lobbying, which is extraordinarily effective, because a sector that is dominated by a handful of firms can easily arrive at a common negotiating position and speak with one voice to the government:
Starting with the Carter administration, and accelerating through every subsequent administration except Biden's, America has adopted an explicitly pro-monopoly policy, called the "consumer welfare" antitrust theory. 40 years later, our economy is riddled with monopolies:
Every part of this Echo privacy massacre is downstream of that policy choice: "growth stock" narratives about AI, twiddling, DMCA 1201, the Darth Vader MBA, the end of legal privacy protections. These are material things, not ideological ones. They exist to make a very, very small number of people very, very rich.
Your Echo is your property, you paid for it. You paid for the product and you are still the product:
Now, Amazon says that the recordings your Echo will send to its data-centers will be deleted as soon as it's been processed by the AI servers. Amazon's made these claims before, and they were lies. Amazon eventually had to admit that its employees and a menagerie of overseas contractors were secretly given millions of recordings to listen to and make notes on:
https://archive.is/TD90k
And sometimes, Amazon just sent these recordings to random people on the internet:
Fool me once, etc. I will bet you a testicle* that Amazon will eventually have to admit that the recordings it harvests to feed its AI are also being retained and listened to by employees, contractors, and, possibly, randos on the internet.
*Not one of mine
If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
nasty to see some guy twiddling with his dick in a video w/his heartbeat. just stick to showing off your chest or something, not your damn cock..........gross
Checking in on the state of Amazonβs chickenized reverse-centaurs
I'm on a tour with my new book, the international bestseller Enshittification: catch me next in Vancouver, Montreal, Ottawa! Full schedule here (New dates just added in San Diego and Denver!).
Amazon has invented a new kind of labor travesty: the chickenized reverse centaur. That's a worker who has to foot the bill to outfit a work environment where they nevertheless have no autonomy (chickenization) and whose body is conscripted to act as a peripheral for a digital system (reverse centaur):
"Chickenization" is a term out of labor economics, inspired by the brutal state of the poultry industry, where three giant processing companies have divided up the market so that every chicken farmer has just one place where they can sell their birds. To sell your birds to one of these plants, you have to give them total control over your operation. They sell you the baby chicks, they tell you what kind of coop to build and what lightbulbs to install and when they should be off or on. They tell you which vet to use and which medicines can be administered to your birds. They tell you what to feed your birds and when to feed them. They design your coop and tell you who is allowed to maintain it. The one thing they don't tell you is how much you'll be paid for your birds β that's something you only discover when it's time to sell them, and the sum you're offered is based on the packer's region-wide intelligence on how you and all your competitors are faring, and is calculated to be the smallest amount to allow you to roll over your loans and go into more debt to grow more birds for them.
At its root, "chickenization" is about de-risking, cloaked in the language of entrepreneurship. Chicken farmers assume all the risk for the poultry packers, but they're told that they're their own bosses. The only way in which a chicken farmer resembles an entrepreneur is that they have to bear all the risk of failure β without having any upside for success. Packers can (and do) secretly decide to experiment at farmers' expense, ordering some of their farmers to vary their feeding, light and veterinary routines to see if they can eke new efficiencies out of the process. If that works, the surplus is reaped by the packer. If that fails, the losses are borne by the farmer, who is never told that they were funding an experiment.
Amazon makes extensive use of chickenization in its many commercial arrangements, tightly defining the working conditions of many "self-employed" workers, like the clickwork "turkers" who power the Mechanical Turk service. But the most chickenized of all the people in Amazon's network of cutouts and arm's-length arrangements are the "entrepreneurs" who are lured into starting a "Delivery Service Platform" (DSP) business.
To start a DSP, you borrow lots of money to buy vans that you outfit to Amazon's exacting specifications, filling them with interior and exterior sensors and cameras, painting them with Amazon livery, and kitting them out with shelving and other infrastructure to Amazon's exacting specification. Then, you hire workers β giving Amazon a veto over who you hire β and you train them β using Amazon's training materials. You sign them up for Amazon's platforms, which monitor and rank those workers, and then you get paid either $0.10 per parcel, or maybe $0.50 per parcel, or sometimes $0.00 per parcel, all at Amazon's sole discretion.
That's a pretty chickenized arrangement. But what about reverse centaurs?
In automation theory, a "centaur" is someone who is assisted by some automation system (they are a fragile human head being assisted by a tireless machine). Therefore, a reverse centaur is a person who has been conscripted to serve as a peripheral for a machine, a human body surmounted and directed by a brute and uncaring head that not only uses them, but uses them up.
The drivers that DSPs hire are reverse centaurs. Using various forms of automation, Amazon drives these workers to work at a dangerous, humiliating and unsustainable pace, setting and enforcing not just quotas, but also scripting where drivers' eyes must be pointed, how they must accelerate and decelerate, what routes they take, and more. These edicts are enforced by the in-van and on-body automation systems that direct and discipline workers, tools that labor activists call "electronic whips":
The chickenized owners of DSPs must enforce the edicts Amazon brings down on their reverse centaur workers β Amazon can terminate any DSP, at any time, for any reason or no reason, stranding an "independent entrepreneur" with heavily mortgaged rolling stock that can only be used to deliver Amazon packages, long term leases on garages and parking lots, liability for driver accidents caused by automation systems that punish drivers for e.g. braking suddenly if someone steps into the road, and massive loans.
So when Amazon directs a DSP to fire or discipline a worker, that worker is in trouble. Amazon has hybridized chickenization and reverse centaurism, creating a chickenized reverse centaur, a new kind of labor travesty never seen before.
In "Driven Down," a new report from the DAIR Institute, authors Adrienne Williams, Alex Hanna and Sandra Barcenas draw on interviews with DSP drivers and Williams's own experience driving for Amazon to document the state of the Chickenized Reverse Centaur. It's not good:
"Driven Down" vividly describes β often in drivers' own words β how the life of a chickenized reverse centaur is one of wage theft, privacy invasions, humilation and on-the-job physical risks, for drivers and the communities they drive in.
DSP drivers interact with multiple automation systems β at least nine apps that monitor, score and discipline them. These apps are supposed to run on employer-supplied phones, but these phones are frequently broken, and drivers face severe punishment if these apps aren't all running during their shifts. As a result, drivers routinely install these apps on their own phones, and must give them broad, far-reaching permissions, such that drivers' own phones are surveilling them for Amazon 24/7, whether or not they're on the clock. It's not just DSP owners who are chickenized β it's also drivers, footing the bill for their own electronic whips.
First and foremost, these apps tell the drivers where to go and how to get there. Drivers are dispatched to hundreds of stops per day, on a computer-generated route that is not vetted or sanity-checked by a human before it is non-negotiably handed to a driver. Famously, plotting an efficient route among many points is one of the most insoluble computing problems, the so-called "traveling salesman" problem:
But it turns out that there is an optimal solution to the traveling salesman problem: get a computer to make a bizarre and dangerous approximation of the optimal route, and then blame and fine workers when it doesn't work. This doesn't optimize the route, but it does shift all the costs of a suboptimal route to workers.
Crucially, Amazon trusts its computer-generated routes, based on map data, over the word of drivers. For example, drivers are often directed to make "group stops" β where the driver parks the van and then delivers to multiple addresses at once (for example, at an apartment complex or office block). Amazon's mapping service assumes that addresses that are in the same complex or development are close together, even when they are very distant. If a driver dares to move and re-park their van to deliver parcels to distant addresses, the app punishes them for making an unauthorized positional adjustment. If a driver attempts to deliver all the parcels without moving the van, they are penalized for taking too long. Even if drivers report the mapping error, it persists, resulting in strings of infractions, day after day.
When drivers fail to make quota, the DSP's per-parcel payout is reduced. DSPs whose drivers perfectly obey the (irrational, impossible) orders of Amazon's apps get $0.50 per parcel delivered. If drivers fall short of the apps' expectations, the per parcel-rate can fall to $0.10, or, in some cases, zero.
This provides a powerful incentive to DSPs to pressure drivers to engage in unsafe practices if the alternative would displease the app. Drivers are penalized for sudden braking and swerving, for example, but are also penalized for missing quota, which puts drivers in the impossible position of having to drive as quickly as possible but also not to swerve or brake if a sudden traffic hazard pops up. In one absurd tale, a driver describes how they were shifted to an electric van that did regenerative braking when they released the accelerator. The app expected drivers to slow down by releasing the accelerator, not by touching the brakes, but this meant that the van's brake lights never switched on. When a driver slowed at a yellow light, they were badly rear-ended by a following UPS truck, whose driver had assumed the Amazon DSP driver was going to rush the light (because the van's brake lights didn't light up).
Meeting quota means that drivers are also not able to stop for bathroom breaks or to take car of other personal hygiene matters. This is bad enough when it means peeing in a bottle, but it's even worse when the only way to take care of period-related matters is to go into the back of the van β where cameras record everything you do β and manage things there.
Drivers are told many inconsistent things about those cameras. Some drivers have been told that the footage is only reviewed after an accident or complaint, but when drivers do get into accidents or have complaints lodged against them, they are often fired or disciplined without anyone reviewing the footage. Meanwhile, drivers are sometimes punished for things the cameras have recorded even when there was no complaint or accident.
The existence of all that empirical evidence of things happening in and outside an Amazon DSP van makes little to no difference to drivers' employment fairness. When a malfunctioning seatbelt sensor insists that a driver has removed their seatbelt while driving, 80+ times in a single shift, the driver struggled to get their docked wages or lost jobs back. When a driver swerved to avoid an oncoming big rig whose driver had fallen asleep and drifted across the media, the driver was penalized β the driver this happened to had his score in "Mentor" (one of the many apps) docked from 850 to 650. Amazon won't tell drivers what their Mentor scores mean, but many drivers β and DSP owners β believe than anything less than a perfect score will result in punishment or termination.
Attaining and maintaining a perfect score is an impossible task, because Amazon will not disclose what drivers are expected to do β it will only penalize them when they fail to do it. Take the photos that Amazon drivers are expected to snap of parcels after they are delivered. The criteria for these photos is incredibly strict β and also not disclosed. Drivers are penalized for having their hands or shoes or reflections in the image, for capturing customers or their pets, for capturing the house-number. They aren't allowed to photograph shoes that are left on the doormat. Drivers share tips with one another about how to take a picture without losing points, but it's a moving target.
Among drivers, there's a (likely correct) belief that Amazon will not tell them how the apps are generating their scores out of fear that if drivers knew the scoring rubric, they'd start to game it. This is a widespread practice within the world of content moderation and spamfighting, where security practitioners who would normally reject the idea of "security through obscurity" out of hand suddenly embrace secrecy-dependent security measures:
All this isn't just dangerous and dehumanizing, it's also impoverishing. Drivers who get downranked by these imperious and unaccountable and unexplained algorithms have their hours cut or get fired altogether. The apps set a quota that can't possibly be reached if drivers take their mandated (and unpaid) 30 minute lunch and two 15-minute breaks (drivers who miss quota twice are automatically terminated). This time is given over to unpaid labor. As the report explains:
Drivers are not paid for their 30 minute lunch. A full-time employee working an 8 to 10 hour shift would be working either 4 or 5 days out of each week. At $20 an hour, that is two hours a week for four-day employees, resulting in $40 of unpaid labor a week, $160 a month, almost $2,000 a year.
Drivers are also assigned "homework" β videos they are required watch and simulator exercises they are required to complete as remediation for their real or imagined infractions. This, too, is unpaid, mandatory work. Drivers are required to attend "stand up" meetings at the start of their shifts, and this is also often unpaid work.
Amazon makes a big show of "listening to drivers," but they're never heard. A driver who reported being held at gunpoint by literal Nazis who objected to having their parcels delivered by a Jew had his complaints ignored, and those violent, armed Nazi customers continued to get their parcels delivered.
Even modest requests go unanswered. Drivers for one DSP begged for porta-toilets in the parking lot, rather than having to waste time (and miss quota) legging it to a distant bathroom. They were ignored, and all 50 drivers continue to share a single toilet.
But β thanks to chickenization β none of this is Amazon's problem. It's all the problem of a chickenized DSP "entrepreneur" who serves as a useful accountability sink for Amazon and who can be bankrupted at a moment's notice should they fail to do Amazon's precise bidding.
There's one bright spot here, though: the National Labor Relations Board has brought a case in California seeking to have Amazon held to be a "joint employer" of those reverse centaurs behind the wheels of those vans:
This is the very last residue of the NLRB's authority, the rest having been drained away by Trump as part of Project 2025. If they prevail, it will open the door to drivers suing Amazon for unfair labor practices under both federal and state law β and in California and New York, that labor law just got a lot tougher for Amazon:
The chickenized reverse centaur is a new circle of labor hell, a genuinely innovative way of making workers' lives worse in order to extract more billions for one of the most profitable companies in history.
If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
Hey, German-speakers! Through a very weird set of circumstances, I ended up owning the rights to the German audiobook of my bestselling 2022 cryptocurrency heist technothriller Red Team Blues and now I'm selling DRM-free audio and ebooks, along with the paperback (all in German and English) on a Kickstarter that runs until August 11.
Delta airlines has announced a new surveillance pricing plan: they're going to feed an AI the nonconsensually harvested personal data that data-brokers and credit bureaux hold on you to predict the maximum you're willing to pay, and then price their tickets accordingly:
Data-brokers hold all kinds of data on you, from the "legitimate" information about everywhere your car has driven, to everywhere point in space that the Bluetooth radios on your phone and headphones have passed, to everything you've bought, to every website you've visited and every search you've performed. They also buy data that has been straight up stolen from you by spyware implanted on your phone:
All of this can be merged into a single file that you have no right to scrutinize, let alone redact. Biden's Consumer Finance Protection Bureau passed a rule banning all this shit, but Trump illegally killed off that rule:
Capitalism's highest form of creativity is finding ways to rip you off, and the business world's most creative minds have found a million ways to exploit this data, including surveillance pricing. For example, McDonald's has invested in a Kiwi startup called Plexure that offers to help restaurants jack up the price of your usual order on payday, when you can afford to pay more:
And then there's the Big Three "Uber for nurses" apps, who use surveillance data to calculate wages for nurses, offering lower hourly rates to nurses who are carrying a lot of credit-card debt, on the grounds that they are too desperate to turn down a lowball offer:
And just as these gigwork apps are deciding what your labor is worth, surveillance pricing systems decide what your money is worth, charging you more than another otherwise identical customer, for an identical product, meaning your dollar is worth less than that other customer's dollar:
Now we have Delta, which promises to do the same thing, but for plane tickets. Obviously, the aviation industry has long practiced a form of "price discrimination," charging radically different sums for the same seat, based on when you buy the ticket, or when you plan to return.
But this is different, and to explain why, here's a link to an article by the great Hubert Horan, who may be best known to my readers for his incredible breakdowns of Uber's finances, but whose life's work is as an aviation analyst:
Horan draws a distinction between surveillance pricing and "second degree price discrimination." Surveillance pricing targets you, personally, based on your personal information. "Second degree price discrimination" charges everyone like you the same price: like, everyone who buys a roundtrip ticket without a Saturday night stay is charged extra on the grounds that they are probably a price-insensitive business traveler whose fare is being paid by a corporation.
Surveillance pricing is first-degree price discrimination, with every customer seeing a different price. Horan argues that second-degree discrimination created efficiencies, for example, by offering cheap last-minute seats to people thinking about going away for the weekend, who fill seats that would otherwise go empty. Horan says these efficiencies have tapped out, thanks to the application of straightforward pricing algorithms to tickets.
Now, Delta wants to squeeze more profits out of price discrimination, but by employing first-degree discrimination, they're doing so without any benefit to fliers (unlike second-degree discrimination, which made many fliers better off because they were able to score cheaper tickets). This makes Delta's surveillance pricing a "pure transfer" β shifting wealth from fliers to shareholders with no benefit to those fliers.
Delta is doing this in partnership with an Israeli firm called Fetcherr, whose sales pitch denies that they are using surveillance data to price tickets, despite what Delta has claimed. Horan doesn't know what to make of this, but he speculates that because Fetcherr bills itself as an AI company, Delta thinks it can impress investors by claiming that it will goose prices by combining surveillance (well understood to be a way to benefit corporations at the expense of their customers) and AI, a hype-filled technology that is endlessly impressive to credulous investors.
A bigger mystery is how Fetcherr plans to do surveillance pricing without surveillance. Horan points out that the company's founders come from hedge funds, where automated high-speed AI trader-bots fed on tons of public market data are routinely used. He thinks it's possible that "Fletchrr doesnβt understand airline pricing very well." Also, being finance bros, they thought "airlines were 'outdated' 'undisrupted' and had seen few recent technological advances." But, Horan continues, the reason airlines aren't doing a lot with their algorithmic pricing is that they've already done it all, having pioneered the field.
Horan's favored explanation for the disconnection between what Fetcherr and Delta claim they're doing is that, on the one hand, they want to obscure the fact that they're doing surveillance pricing (to avoid regulatory scrutiny and consumer backlash), but on the other hand, they want to telegraph (to investors) that this is exactly what they're doing.
It's what Uber already does, repricing both the labor of its drivers based on their economic desperation, and the cost of your fare based on what its surveillance dossier suggests you're willing to pay. It's certainly increased Uber's margins β by effecting a pure transfer from riders and drivers to shareholders.
But Uber rides are last-minute, small dollar purchases, which decreases the likelihood that a rider will shop around before booking. By contrast, Horan says, most fliers buy well in advance, from online travel sites that show them lots of competing prices.
One thing Horan doesn't mention here is that British Airways has just done a top-to-bottom rejig of its frequent flier program to severely penalize anyone who buys tickets from one of these sites, effectively requiring its fliers to buy from BA.com. For example, I booked a $300 Alaska Airlines ticket on Alaska's website, using my BA frequent flier ID.
Under the old system, this would have been worth 10 tier points out of the 1500 needed to get Gold status (0.66%). Under the new system, I got 12 points out of the 20,000 needed to get Gold (0.05%) β a 93% reduction in the reward value of this flight.
Which is to say that if you don't book on BA's site, you effectively cannot make status. BA has also announced a surveillance pricing deal with an AI company β and this gambit will block its best fliers from getting a better price from an online travel agency.
One other key difference between Uber and Delta: Uber has gone to great lengths to hide the fact that it's doing surveillance pricing from both drivers and riders. Delta issued a press-release!
There's a certain kind of neoclassical economist who loves surveillance pricing and praises its "efficiencies." These apologists claim that by increasing the amount of "information" in the system, we encourage sellers to discount to customers who can't afford as much, making everyone better off:
This is nonsense. Sellers don't want to "increase the amount of information in the system." They want to spy on you. If you doubt it for an instant, just ask the firms that scrape airline websites for up-to-date pricing information:
Not only will airlines sue you for trying to find out what their fares are, they'll also sue you for figuring out how to get a better deal on their fares:
Companies that do surveillance pricing are violently allergic to sousveillance pricing. When they spy on you, that's progress. When you monitor their behavior, that's piracy.
As an aside, this reminds me of one of the AI industry's most egregious hoaxes-du-jour: the pretense that "agentic AI" is just around the corner, and soon we will be able to ask a chatbot to (e.g.) comparison shop across multiple website for the best airfare and book us a ticket:
This absolutely totally does not work. You should not give your credit-card number to a chatbot and ask it to go out an buy you anything, lest you end up paying $30 for a dozen eggs and buying tickets to a baseball stadium in the middle of the ocean:
AI agent demos are so dismal that AI companies are no longer claiming that "agentic AI" will involve chatbots that nagivate the web as is. Rather, they're claiming that every website will eventually re-tool so that it can be reliably and predictably addressed by an AI agent, with all of its user interface elements well-labeled and/or addressable programatically, via an API.
This is a remarkable sleight of hand! First of all, re-engineering every website to embrace a common set of labels and API fields is a gigantic engineering feat β formally called "the semantic web" β that has been attempted since 1999 without any meaningful progress:
https://en.wikipedia.org/wiki/Semantic_Web
In fact, the first viral article I ever published online was "Metacrap," a critique of semantic web efforts. That essay is now 24 years old:
In that essay, I suggest that there are multiple reasons that companies will not voluntarily retool their sites to make it easier to comparison shop. One important reason is that companies don't believe their products are comparable with competing products (or they don't want you to think so). Coach wants you to think that its $40,000 handbags can't be replaced with a well-made $100 bag or even a $0.10 plastic bag. They are not going to voluntarily categorize their handbag in a way that facilitates these comparisons.
Then there are companies that do want to be compared to rivals, for disingenuous reasons. That's why we saw such a proliferation of junk fees (stupid surcharges tacked on at checkout time): hotels, airlines and car rental agencies knew that the majority of their customers shopped for their offerings on comparison sites. By offering a low sticker price, a company could win on price comparison, even though it was substantially more expensive after its junk fees were factored in.
Finally, there's the fact that companies want to lie to you, and adding "semantics" to the web does nothing to prevent such lies, and indeed, makes them easier to tell. Think of all the Amazon sellers who use deceptive product photos to make you think you're getting (e.g.) a useful kitchen spatula, when they're selling a spatula so small that it appears to be engineered for a dollhouse; or companies that sell powerbanks that look like a useful portable battery but can't even recharge an LED flashlight, etc, etc. AI agents can't tell if metadata is correct or not!
Every complex ecosystem has parasites; that goes triple for the web. We won't fix agentic AI by asking people to accurately label their offerings, not when they stand to benefit by lying:
And if we could rejig the web to make it hospitable to agentic AI, we wouldn't need AI to make this happen. Fetching airfares for several routes and comparing them isn't something you need an AI-style inference engine for β it's a straightforward algorithmic problem that can be easily solved. The part that agentic AI purports to solve isn't figuring out which airfare out of a list is cheapest β it's compiling the list itself, from unstructured data retrieved from heterogeneous websites that are doing everything they can to prevent the compilation of such a list.
This is a well-known AI gambit. First, announce that agentic AI will be able to automate tasks that only humans can manage today; then insist that everything has to be changed to be amenable to the new technology. This is exactly what the self-driving car grifters (who were on the leading edge of the AI grift) did. First, they announced that AIs would be able to pilot cars in spaces filled with human drivers, walkers and cyclists. Then, when it became clear that this would result in slaughtersome robot-on-human violence, they demanded that humans curtail their behavior to avoid upsetting the robot.
They call this "the pogo-stick problem":
βI think many AV teams could handle a pogo stick user in pedestrian crosswalk,β Ng told me. βHaving said that, bouncing on a pogo stick in the middle of a highway would be really dangerous.β
βRather than building AI to solve the pogo stick problem, we should partner with the government to ask people to be lawful and considerate,β he said. βSafety isnβt just about the quality of the AI technology.β
Automation is real and can deliver real benefits to people. Sometimes, automation requires that other systems be adjusted to facilitate its functioning. But this is a gambit. It's a scam. AI agents aren't going to replace human labor. The only way we'll replace human labor with software agents is by redesigning all these heterogeneous, competing systems owned by people who benefit from the status quo and have every motivation to obstruct this project.
Good luck with that.
Support me this summer in the Clarion Write-A-Thon and help raise money for the Clarion Science Fiction and Fantasy Writers' Workshop! This summer, I'm writing The Reverse-Centaur's Guide to AI, a short book for Farrar, Straus and Giroux that explains how to be an effective AI critic.
If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
Even if you think AI search could be good, it wonβt be good
TONIGHT (May 15), I'm in NORTH HOLLYWOOD for a screening of STEPHANIE KELTON'S FINDING THE MONEY; FRIDAY (May 17), I'm at the INTERNET ARCHIVE in SAN FRANCISCO to keynote the 10th anniversary of the AUTHORS ALLIANCE.
The big news in search this week is that Google is continuing its transition to "AI search" β instead of typing in search terms and getting links to websites, you'll ask Google a question and an AI will compose an answer based on things it finds on the web:
Google bills this as "let Google do the googling for you." Rather than searching the web yourself, you'll delegate this task to Google. Hidden in this pitch is a tacit admission that Google is no longer a convenient or reliable way to retrieve information, drowning as it is in AI-generated spam, poorly labeled ads, and SEO garbage:
Googling used to be easy: type in a query, get back a screen of highly relevant results. Today, clicking the top links will take you to sites that paid for placement at the top of the screen (rather than the sites that best match your query). Clicking further down will get you scams, AI slop, or bulk-produced SEO nonsense.
AI-powered search promises to fix this, not by making Google search results better, but by having a bot sort through the search results and discard the nonsense that Google will continue to serve up, and summarize the high quality results.
Now, there are plenty of obvious objections to this plan. For starters, why wouldn't Google just make its search results better? Rather than building a LLM for the sole purpose of sorting through the garbage Google is either paid or tricked into serving up, why not just stop serving up garbage? We know that's possible, because other search engines serve really good results by paying for access to Google's back-end and then filtering the results:
Another obvious objection: why would anyone write the web if the only purpose for doing so is to feed a bot that will summarize what you've written without sending anyone to your webpage? Whether you're a commercial publisher hoping to make money from advertising or subscriptions, or β like me β an open access publisher hoping to change people's minds, why would you invite Google to summarize your work without ever showing it to internet users? Nevermind how unfair that is, think about how implausible it is: if this is the way Google will work in the future, why wouldn't every publisher just block Google's crawler?
A third obvious objection: AI is bad. Not morally bad (though maybe morally bad, too!), but technically bad. It "hallucinates" nonsense answers, including dangerous nonsense. It's a supremely confident liar that can get you killed:
The promises of AI are grossly oversold, including the promises Google makes, like its claim that its AI had discovered millions of useful new materials. In reality, the number of useful new materials Deepmind had discovered was zero:
This is true of all of AI's most impressive demos. Often, "AI" turns out to be low-waged human workers in a distant call-center pretending to be robots:
The AI video demos that represent "an existential threat to Hollywood filmmaking" turn out to be so cumbersome as to be practically useless (and vastly inferior to existing production techniques):
But let's take Google at its word. Let's stipulate that:
a) It can't fix search, only add a slop-filtering AI layer on top of it; and
b) The rest of the world will continue to let Google index its pages even if they derive no benefit from doing so; and
c) Google will shortly fix its AI, and all the lies about AI capabilities will be revealed to be premature truths that are finally realized.
AI search is still a bad idea. Because beyond all the obvious reasons that AI search is a terrible idea, there's a subtle β and incurable β defect in this plan: AI search β even excellent AI search β makes it far too easy for Google to cheat us, and Google can't stop cheating us.
Remember: enshittification isn't the result of worse people running tech companies today than in the years when tech services were good and useful. Rather, enshittification is rooted in the collapse of constraints that used to prevent those same people from making their services worse in service to increasing their profit margins:
These companies always had the capacity to siphon value away from business customers (like publishers) and end-users (like searchers). That comes with the territory: digital businesses can alter their "business logic" from instant to instant, and for each user, allowing them to change payouts, prices and ranking. I call this "twiddling": turning the knobs on the system's back-end to make sure the house always wins:
https://pluralistic.net/2023/02/19/twiddler/
What changed wasn't the character of the leaders of these businesses, nor their capacity to cheat us. What changed was the consequences for cheating. When the tech companies merged to monopoly, they ceased to fear losing your business to a competitor.
Google's 90% search market share was attained by bribing everyone who operates a service or platform where you might encounter a search box to connect that box to Google. Spending tens of billions of dollars every year to make sure no one ever encounters a non-Google search is a cheaper way to retain your business than making sure Google is the very best search engine:
Competition was once a threat to Google; for years, its mantra was "competition is a click away." Today, competition is all but nonexistent.
Then the surveillance business consolidated into a small number of firms. Two companies dominate the commercial surveillance industry: Google and Meta, and they collude to rig the market:
https://en.wikipedia.org/wiki/Jedi_Blue
That consolidation inevitably leads to regulatory capture: shorn of competitive pressure, the companies that dominate the sector can converge on a single message to policymakers and use their monopoly profits to turn that message into policy:
This is why Google doesn't have to worry about privacy laws. They've successfully prevented the passage of a US federal consumer privacy law. The last time the US passed a federal consumer privacy law was in 1988. It's a law that bans video store clerks from telling the newspapers which VHS cassettes you rented:
In Europe, Google's vast profits lets it fly an Irish flag of convenience, thus taking advantage of Ireland's tolerance for tax evasion and violations of European privacy law:
Google doesn't fear competition, it doesn't fear regulation, and it also doesn't fear rival technologies. Google and its fellow Big Tech cartel members have expanded IP law to allow it to prevent third parties from reverse-engineer, hacking, or scraping its services. Google doesn't have to worry about ad-blocking, tracker blocking, or scrapers that filter out Google's lucrative, low-quality results:
https://locusmag.com/2020/09/cory-doctorow-ip/
Google doesn't fear competition, it doesn't fear regulation, it doesn't fear rival technology and it doesn't fear its workers. Google's workforce once enjoyed enormous sway over the company's direction, thanks to their scarcity and market power. But Google has outgrown its dependence on its workers, and lays them off in vast numbers, even as it increases its profits and pisses away tens of billions on stock buybacks:
Google is fearless. It doesn't fear losing your business, or being punished by regulators, or being mired in guerrilla warfare with rival engineers. It certainly doesn't fear its workers.
Making search worse is good for Google. Reducing search quality increases the number of queries, and thus ads, that each user must make to find their answers:
If Google can make things worse for searchers without losing their business, it can make more money for itself. Without the discipline of markets, regulators, tech or workers, it has no impediment to transferring value from searchers and publishers to itself.
Which brings me back to AI search. When Google substitutes its own summaries for links to pages, it creates innumerable opportunities to charge publishers for preferential placement in those summaries.
This is true of any algorithmic feed: while such feeds are important β even vital β for making sense of huge amounts of information, they can also be used to play a high-speed shell-game that makes suckers out of the rest of us:
When you trust someone to summarize the truth for you, you become terribly vulnerable to their self-serving lies. In an ideal world, these intermediaries would be "fiduciaries," with a solemn (and legally binding) duty to put your interests ahead of their own:
But Google is clear that its first duty is to its shareholders: not to publishers, not to searchers, not to "partners" or employees.
AI search makes cheating so easy, and Google cheats so much. Indeed, the defects in AI give Google a readymade excuse for any apparent self-dealing: "we didn't tell you a lie because someone paid us to (for example, to recommend a product, or a hotel room, or a political point of view). Sure, they did pay us, but that was just an AI 'hallucination.'"
The existence of well-known AI hallucinations creates a zone of plausible deniability for even more enshittification of Google search. As Madeleine Clare Elish writes, AI serves as a "moral crumple zone":
That's why, even if you're willing to believe that Google could make a great AI-based search, we can nevertheless be certain that they won't.
If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
Your car spies on you and rats you out to insurance companies
I'm on tour with my new, nationally bestselling novel The Bezzle! Catch me TOMORROW (Mar 13) in SAN FRANCISCO with ROBIN SLOAN, then Toronto, NYC, Anaheim, and more!
Another characteristically brilliant Kashmir Hill story for The New York Times reveals another characteristically terrible fact about modern life: your car secretly records fine-grained telemetry about your driving and sells it to data-brokers, who sell it to insurers, who use it as a pretext to gouge you on premiums:
This is true whether you own or lease the car, and it's separate from the "black box" your insurer might have offered to you in exchange for a discount on your premiums. In other words, even if you say no to the insurer's carrot β a surveillance-based discount β they've got a stick in reserve: buying your nonconsensually harvested data on the open market.
I've always hated that saying, "If you're not paying for the product, you're the product," the reason being that it posits decent treatment as a customer reward program, like the little ramekin warm nuts first class passengers get before takeoff. Companies don't treat you well when you pay them. Companies treat you well when they fear the consequences of treating you badly.
Take Apple. The company offers Ios users a one-tap opt-out from commercial surveillance, and more than 96% of users opted out. Presumably, the other 4% were either confused or on Facebook's payroll. Apple β and its army of cultists β insist that this proves that our world's woes can be traced to cheapskate "consumers" who expected to get something for nothing by using advertising-supported products.
But here's the kicker: right after Apple blocked all its rivals from spying on its customers, it began secretly spying on those customers! Apple has a rival surveillance ad network, and even if you opt out of commercial surveillance on your Iphone, Apple still secretly spies on you and uses the data to target you for ads:
Even if you're paying for the product, you're still the product β provided the company can get away with treating you as the product. Apple can absolutely get away with treating you as the product, because it lacks the historical constraints that prevented Apple β and other companies β from treating you as the product.
As I described in my McLuhan lecture on enshittification, tech firms can be constrained by four forces:
When companies have real competitors β when a sector is composed of dozens or hundreds of roughly evenly matched firms β they have to worry that a maltreated customer might move to a rival. 40 years of antitrust neglect means that corporations were able to buy their way to dominance with predatory mergers and pricing, producing today's inbred, Habsburg capitalism. Apple and Google are a mobile duopoly, Google is a search monopoly, etc. It's not just tech! Every sector looks like this:
Eliminating competition doesn't just deprive customers of alternatives, it also empowers corporations. Liberated from "wasteful competition," companies in concentrated industries can extract massive profits. Think of how both Apple and Google have "competitively" arrived at the same 30% app tax on app sales and transactions, a rate that's more than 1,000% higher than the transaction fees extracted by the (bloated, price-gouging) credit-card sector:
But cartels' power goes beyond the size of their warchest. The real source of a cartel's power is the ease with which a small number of companies can arrive at β and stick to β a common lobbying position. That's where "regulatory capture" comes in: the mobile duopoly has an easier time of capturing its regulators because two companies have an easy time agreeing on how to spend their app-tax billions:
Apple β and Google, and Facebook, and your car company β can violate your privacy because they aren't constrained regulation, just as Uber can violate its drivers' labor rights and Amazon can violate your consumer rights. The tech cartels have captured their regulators and convinced them that the law doesn't apply if it's being broken via an app:
In other words, Apple can spy on you because it's allowed to spy on you. America's last consumer privacy law was passed in 1988, and it bans video-store clerks from leaking your VHS rental history. Congress has taken no action on consumer privacy since the Reagan years:
But tech has some special enshittification-resistant characteristics. The most important of these is interoperability: the fact that computers are universal digital machines that can run any program. HP can design a printer that rejects third-party ink and charge $10,000/gallon for its own colored water, but someone else can write a program that lets you jailbreak your printer so that it accepts any ink cartridge:
Tech companies that contemplated enshittifying their products always had to watch over their shoulders for a rival that might offer a disenshittification tool and use that as a wedge between the company and its customers. If you make your website's ads 20% more obnoxious in anticipation of a 2% increase in gross margins, you have to consider the possibility that 40% of your users will google "how do I block ads?" Because the revenue from a user who blocks ads doesn't stay at 100% of the current levels β it drops to zero, forever (no user ever googles "how do I stop blocking ads?").
The majority of web users are running an ad-blocker:
Web operators made them an offer ("free website in exchange for unlimited surveillance and unfettered intrusions") and they made a counteroffer ("how about 'nah'?"):
Here's the thing: reverse-engineering an app β or any other IP-encumbered technology β is a legal minefield. Just decompiling an app exposes you to felony prosecution: a five year sentence and a $500k fine for violating Section 1201 of the DMCA. But it's not just the DMCA β modern products are surrounded with high-tech tripwires that allow companies to invoke IP law to prevent competitors from augmenting, recongifuring or adapting their products. When a business says it has "IP," it means that it has arranged its legal affairs to allow it to invoke the power of the state to control its customers, critics and competitors:
https://locusmag.com/2020/09/cory-doctorow-ip/
An "app" is just a web-page skinned in enough IP to make it a crime to add an ad-blocker to it. This is what Jay Freeman calls "felony contempt of business model" and it's everywhere. When companies don't have to worry about users deploying self-help measures to disenshittify their products, they are freed from the constraint that prevents them indulging the impulse to shift value from their customers to themselves.
Apple owes its existence to interoperability β its ability to clone Microsoft Office's file formats for Pages, Numbers and Keynote, which saved the company in the early 2000s β and ever since, it has devoted its existence to making sure no one ever does to Apple what Apple did to Microsoft:
Regulatory capture cuts both ways: it's not just about powerful corporations being free to flout the law, it's also about their ability to enlist the law to punish competitors that might constrain their plans for exploiting their workers, customers, suppliers or other stakeholders.
The final historical constraint on tech companies was their own workers. Tech has very low union-density, but that's in part because individual tech workers enjoyed so much bargaining power due to their scarcity. This is why their bosses pampered them with whimsical campuses filled with gourmet cafeterias, fancy gyms and free massages: it allowed tech companies to convince tech workers to work like government mules by flattering them that they were partners on a mission to bring the world to its digital future:
For tech bosses, this gambit worked well, but failed badly. On the one hand, they were able to get otherwise powerful workers to consent to being "extremely hardcore" by invoking Fobazi Ettarh's spirit of "vocational awe":
On the other hand, when you motivate your workers by appealing to their sense of mission, the downside is that they feel a sense of mission. That means that when you demand that a tech worker enshittifies something they missed their mother's funeral to deliver, they will experience a profound sense of moral injury and refuse, and that worker's bargaining power means that they can make it stick.
Or at least, it did. In this era of mass tech layoffs, when Google can fire 12,000 workers after a $80b stock buyback that would have paid their wages for the next 27 years, tech workers are learning that the answer to "I won't do this and you can't make me" is "don't let the door hit you in the ass on the way out" (AKA "sharpen your blades boys"):
With competition, regulation, self-help and labor cleared away, tech firms β and firms that have wrapped their products around the pluripotently malleable core of digital tech, including automotive makers β are no longer constrained from enshittifying their products.
And that's why your car manufacturer has chosen to spy on you and sell your private information to data-brokers and anyone else who wants it. Not because you didn't pay for the product, so you're the product. It's because they can get away with it.
Cars are enshittified. The dozens of chips that auto makers have shoveled into their car design are only incidentally related to delivering a better product. The primary use for those chips is autoenshittification β access to legal strictures ("IP") that allows them to block modifications and repairs that would interfere with the unfettered abuse of their own customers:
The fact that it's a felony to reverse-engineer and modify a car's software opens the floodgates to all kinds of shitty scams. Remember when Bay Staters were voting on a ballot measure to impose right-to-repair obligations on automakers in Massachusetts? The only reason they needed to have the law intervene to make right-to-repair viable is that Big Car has figured out that if it encrypts its diagnostic messages, it can felonize third-party diagnosis of a car, because decrypting the messages violates the DMCA:
Big Car figured out that VIN locking β DRM for engine components and subassemblies β can felonize the production and the installation of third-party spare parts:
The fact that you can't legally modify your car means that automakers can go back to their pre-2008 ways, when they transformed themselves into unregulated banks that incidentally manufactured the cars they sold subprime loans for. Subprime auto loans β over $1t worth! β absolutely relies on the fact that borrowers' cars can be remotely controlled by lenders. Miss a payment and your car's stereo turns itself on and blares threatening messages at top volume, which you can't turn off. Break the lease agreement that says you won't drive your car over the county line and it will immobilize itself. Try to change any of this software and you'll commit a felony under Section 1201 of the DMCA:
Tesla, naturally, has the most advanced anti-features. Long before BMW tried to rent you your seat-heater and Mercedes tried to sell you a monthly subscription to your accelerator pedal, Teslas were demon-haunted nightmare cars. Miss a Tesla payment and the car will immobilize itself and lock you out until the repo man arrives, then it will blare its horn and back itself out of its parking spot. If you "buy" the right to fully charge your car's battery or use the features it came with, you don't own them β they're repossessed when your car changes hands, meaning you get less money on the used market because your car's next owner has to buy these features all over again:
And all this DRM allows your car maker to install spyware that you're not allowed to remove. They really tipped their hand on this when the R2R ballot measure was steaming towards an 80% victory, with wall-to-wall scare ads that revealed that your car collects so much information about you that allowing third parties to access it could lead to your murder (no, really!):
That's why your car spies on you. Because it can. Because the company that made it lacks constraint, be it market-based, legal, technological or its own workforce's ethics.
One common critique of my enshittification hypothesis is that this is "kind of sensible and normal" because "thereβs something off in the consumer mindset that weβve come to believe that the internet should provide us with amazing products, which bring us joy and happiness and we spend hours of the day on, and should ask nothing back in return":
What this criticism misses is that this isn't the companies bargaining to shift some value from us to them. Enshittification happens when a company can seize all that value, without having to bargain, exploiting law and technology and market power over buyers and sellers to unilaterally alter the way the products and services we rely on work.
A company that doesn't have to fear competitors, regulators, jailbreaking or workers' refusal to enshittify its products doesn't have to bargain, it can take. It's the first lesson they teach you in the Darth Vader MBA: "I am altering the deal. Pray I don't alter it any further":
Your car spying on you isn't down to your belief that your carmaker "should provide you with amazing products, which brings your joy and happiness you spend hours of the day on, and should ask nothing back in return." It's not because you didn't pay for the product, so now you're the product. It's because they can get away with it.
The consequences of this spying go much further than mere insurance premium hikes, too. Car telemetry sits at the top of the funnel that the unbelievably sleazy data broker industry uses to collect and sell our data. These are the same companies that sell the fact that you visited an abortion clinic to marketers, bounty hunters, advertisers, or vengeful family members pretending to be one of those:
Decades of pro-monopoly policy led to widespread regulatory capture. Corporate cartels use the monopoly profits they extract from us to pay for regulatory inaction, allowing them to extract more profits.
But when it comes to privacy, that period of unchecked corporate power might be coming to an end. The lack of privacy regulation is at the root of so many problems that a pro-privacy movement has an unstoppable constituency working in its favor.
At EFF, we call this "privacy first." Whether you're worried about grifters targeting vulnerable people with conspiracy theories, or teens being targeted with media that harms their mental health, or Americans being spied on by foreign governments, or cops using commercial surveillance data to round up protesters, or your car selling your data to insurance companies, passing that long-overdue privacy legislation would turn off the taps for the data powering all these harms:
Traditional economics fails because it thinks about markets without thinking about power. Monopolies lead to more than market power: they produce regulatory capture, power over workers, and state capture, which felonizes competition through IP law. The story that our problems stem from the fact that we just don't spend enough money, or buy the wrong products, only makes sense if you willfully ignore the power that corporations exert over our lives. It's nice to think that you can shop your way out of a monopoly, because that's a lot easier than voting your way out of a monopoly, but no matter how many times you vote with your wallet, the cartels that control the market will always win:
Name your price for 18 of my DRM-free ebooks and support the Electronic Frontier Foundation with the Humble Cory Doctorow Bundle.
If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
THIS WEEKEND (June 7β9), I'm in AMHERST, NEW YORK to keynote the 25th Annual Media Ecology Association Convention and accept the Neil Postman Award for Career Achievement in Public Intellectual Activity.
Correction, 7 June 2024: The initial version of this article erroneously described Jeffrey Roper as the founder of ATPCO. He benefited from ATPCO, but did not co-found it. The initial version of this article called ATPCO "an illegal airline price-fixing service"; while ATPCO provides information that the airlines use to set prices, it does not set prices itself, and while the DOJ investigated the company, they did not pursue a judgment declaring the service to be illegal. I regret the error.
Noted anti-capitalist agitator Adam Smith had it right: "People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices."
Despite being a raving commie loon, Smith's observation was so undeniably true that regulators, policymakers, and economists couldn't help but acknowledge that it was true. The trustbusting era was defined by this idea: if we let the number of companies in a sector get too small, or if we let one or a few companies get too big, they'll eventually start to rig prices.
What's more, once an industry contracts corporate gigantism, it will become too big to jail, able to outspend and overpower the regulators charged with reining in its cheating. Anyone who believes Smith's self-evident maxim had to accept its conclusion: that companies had to be kept smaller than the state that regulated them. This wasn't about "punishing bigness" β it was the necessary precondition for a functioning market economy.
We kept companies small for the same reason that we limited the height of skyscrapers: not because we opposed height, or failed to appreciate the value of a really good penthouse view β rather, to keep the building from falling over and wrecking all the adjacent buildings and the lives of the people inside them.
Starting in the neoliberal era β Carter, then Reagan β we changed our tune. We liked big business. A business that got big was doing something right. It was perverse to shut down our best companies. Instead, we'd simply ban big companies from rigging prices. This was called the "consumer welfare" theory of antitrust. It was a total failure.
40 years later, nearly every industry is dominated by a handful of companies, and these companies price-gouge us with abandon. Worse, they use their gigantic ripoff winnings to fill war-chests that fund the corruption of democracy, capturing regulators so that they can rip us off even more, while ignoring labor, privacy and environmental law and ducking taxes.
It turns out that keeping gigantic, opaque, complex corporations honest is really hard. They have so many ways to shuffle money around that it's nearly impossible to figure out what they're doing. Digitalization makes things a million times worse, because computers allow businesses to alter their processes so they operate differently for every customer, and even for every interaction.
This is Dieselgate times a billion: VW rigged its cars to detect when they were undergoing emissions testing and switch to a less polluting, more compliant mode. But when they were on the open road, they spewed lethal quantities of toxic gas, killing people by the thousands. Computers don't make corporate leaders more evil, but they let evil corporate leaders execute far more complex and nefarious plans. Digitalization is a corporate moral hazard, making it just too easy and tempting to rig the game.
That's why Toyota, the largest car-maker in the world, just did Dieselgate again, more than a decade later. Digitalization is a temptation no giant company can resist:
https://www.bbc.com/news/articles/c1wwj1p2wdyo
For forty years, pro-monopoly cheerleaders insisted that we could allow companies to grow to unimaginable scale and still prevent cheating. They passed rules banning companies from explicitly forming agreements to rig prices. About ten seconds later, new middlemen popped up offering "information brokerages" that helped companies rig prices without talking to one another.
Take Agri Stats: the country's hyperconcentrated meatpacking industry pays Agri Stats to "consult on prices." They provide Agri Stats with a list of their prices, and then Agri Stats suggests changes based on its analysis. What does that analysis consist of? Comparing the company's prices to its competitors, who are also Agri Stats customers:
In other words, Agri Stats finds the highest price for each product in the sector, then "advises" all the companies with lower prices to raise their prices to the "competitive" level, creating a one-way ratchet that sends the price of food higher and higher.
More and more sectors have an Agri Stats, and digitalization has made this price-gouging system faster, more efficient, and accessible to sectors with less concentration. Landlords, for example, have tapped into Realpage, a "data broker" that the same thing to your rent that Agri Stats does to meat prices. Realpage requires the landlords who sign up for its service to accept its "recommendations" on minimum rents, ensuring that prices only go up:
Writing for The American Prospect, Luke Goldstein lays out the many ways in which these digital intermediaries have supercharged the business of price-rigging:
Goldstein identifies a kind of patient zero for this ripoff epidemic: Jeffrey Roper, a former Alaska Air exec who benefited from a service that helps airlines set prices. ATPCO was investigated by the DOJ in the 1990s, but the enforcers lost their nerve and settled with the company, which agreed to apply some ornamental fig-leafs to its collusion-machine. Even those cosmetic changes were seemingly a bridge too far Roper, who left the US.
But he came back to serve as Realpage's "principal scientist" β the architect of a nationwide scheme to make rental housing vastly more expensive. For Roper, the barrier to low rents was empathy: landlords felt stirrings of shame when they made shelter unaffordable to working people. Roper called these people "idiots" who sentimentality "costs the whole system."
Sticking a rent-gouging computer between landlords and the people whose lives they ruin is a classic "accountability sink," as described in Dan Davies' new book "The Unaccountability Machine: Why Big Systems Make Terrible Decisions β and How The World Lost its Mind":
It's a form of "empiricism washing": if computers are working in the abstract realm of pure numbers, they're just moving the objective facts of the quantitative realm into the squishy, imperfect qualitative world. Davies' interview on Trashfuture is excellent:
To rig prices, an industry has to solve three problems: the problem of coming to an agreement to fix prices (economists call this "the collective action problem"); the problem of coming up with a price; and the problem of actually changing prices from moment to moment. This is the ripoff triangle, and like a triangle, it has many stable configurations.
The more concentrated an industry is, the easier it is to decide to rig prices. But if the industry has the benefit of digitalization, it can swap the flexibility and speed of computers for the low collective action costs from concentration. For example, grocers that switch to e-ink shelf tags can make instantaneous price-changes, meaning that every price change is less consequential β if sales fall off after a price-hike, the company can lower them again at the press of a button. That means they can collude less explicitly but still raise prices:
My name for this digital flexibility is "twiddling." Businesses with digital back-ends can alter their "business logic" from second to second, and present different prices, payouts, rankings and other key parts of the deal to every supplier or customer they interact with:
https://pluralistic.net/2023/02/19/twiddler/
Not only does twiddling make it easier to rip off suppliers, workers and customers, it also makes these crimes harder to detect. Twiddling made Dieselgate possible, and it also underpinned "Greyball," Uber's secret strategy of refusing to send cars to pick up transportation regulators who would then be able to see firsthand how many laws the company was violating:
Twiddling is so easy that it has brought price-fixing to smaller companies and less concentrated sectors, though the biggest companies still commit crimes on a scale that put these bit-players to shame. In The Prospect, David Dayen investigates the "personalized pricing" ripoff that has turned every transaction into a potential crime-scene:
"Personalized pricing" is the idea that everything you buy should be priced based on analysis of commercial surveillance data that predicts the maximum amount you are willing to pay.
Proponents of this idea β like Harvard's Pricing Lab with its "Billion Prices Project" β insist that this isn't a way to rip you off. Instead, it lets companies lower prices for people who have less ability to pay:
https://thebillionpricesproject.com/
This kind of weaponized credulity is totally on-brand for the pro-monopoly revolution. It's the same wishful thinking that led regulators to encourage monopolies while insisting that it would be possible to prevent "bad" monopolies from raising prices. And, as with monopolies, "personalized pricing" leads to an overall increase in prices. In econspeak, it is a "transfer of wealth from consumer to the seller."
"Personalized pricing" is one of those cuddly euphemisms that should make the hair on the back of your neck stand up. A more apt name for this practice is surveillance pricing, because the "personalization" depends on the vast underground empire of nonconsensual data-harvesting, a gnarly hairball of ad-tech companies, data-brokers, and digital devices with built-in surveillance, from smart speakers to cars:
Much of this surveillance would be impractical, because no one wants their car, printer, speaker, watch, phone, or insulin-pump to spy on them. The flexibility of digital computers means that users always have the technical ability to change how these gadgets work, so they no longer spy on their users. But an explosion of IP law has made this kind of modification illegal:
https://locusmag.com/2020/09/cory-doctorow-ip/
This is why apps are ground zero for surveillance pricing. The web is an open platform, and web-browsers are legal to modify. The majority of web users have installed ad-blockers that interfere with the surveillance that makes surveillance pricing possible:
But apps are a closed platform, and reverse-engineering and modifying an app is a literal felony β several felonies, in fact. An app is just a web-page skinned with enough IP to make it a felony to modify it to protect your consumer, privacy or labor rights:
(Google is leading a charge to turn the web into the kind of enshittifier's paradise that apps represent, blocking the use of privacy plugins and proposing changes to browser architecture that would allow them to felonize modifying a browser without permission:)
Apps are a twiddler's playground. Not only can they "customize" every interaction you have with them, but they can block you (or researchers seeking to help you) from recording and analyzing the app's activities. Worse: digital transactions are intimate, contained to the palm of your hand. The grocer whose e-ink shelf-tags flicker and reprice their offerings every few seconds can be collectively observed by people who are in the same place and can start a conversation about, say, whether to come back that night a throw a brick through the store's window to express their displeasure. A digital transaction is a lonely thing, atomized and intrinsically shielded from a public response.
That shielding is hugely important. The public hates surveillance pricing. Time and again, through all of American history, there have been massive and consequential revolts against the idea that every price should be different for every buyer. The Interstate Commerce Commission was founded after Grangers rose up against the rail companies' use of "personalized pricing" to gouge farmers.
Companies know this, which is why surveillance pricing happens in secret. Over and over, every day, you are being gouged through surveillance pricing. The sellers you interact with won't tell you about it, so to root out this practice, we have to look at the B2B sales-pitches from the companies that sell twiddling tools.
One of these companies is Plexure, partly owned by McDonald's, which provides the surveillance-pricing back-ends for McD's, Ikea, 7-Eleven, White Castle and others β basically, any time a company gives you a hard-sell to order via its apps rather than its storefronts or its website, you should assume you're getting twiddled, hard.
These companies use the enshittification playbook to trap you into using their apps. First, they offer discounts to customers who order through their apps β then, once the customers are fully committed to shopping via app, they introduce surveillance pricing and start to jack up the prices.
For example, Plexure boasts that it can predict what day a given customer is getting paid on and use that information to raise prices on all the goods the customer shops for on that day, on the assumption that you're willing to pay more when you've got a healthy bank balance.
The surveillance pricing industry represents another reason for everything you use to spy on you β any data your "smart" TV or Nest thermostat or Ring doorbell can steal from you can be readily monetized β just sell it to a surveillance pricing company, which will use it to figure out how to charge you more for everything you buy, from rent to Happy Meals.
But the vast market for surveillance data is also a potential weakness for the industry. Put frankly: the commercial surveillance industry has a lot of enemies. The only thing it has going for it is that so many of these enemies don't know that what's they're really upset about is surveillance.
Some people are upset because they think Facebook made Grampy into a Qanon. Others, because they think Insta gave their kid anorexia. Some think Tiktok is brainwashing millennials into quoting Osama bin Laden. Some are upset because the cops use Google location data to round up Black Lives Matter protesters, or Jan 6 insurrectionists. Some are angry about deepfake porn. Some are angry because Black people are targeted with ads for overpriced loans or colleges:
And some people are angry because surveillance feeds surveillance pricing. The thing is, whatever else all these people are angry about, they're all angry about surveillance. Are you angry that ad-tech is stealing a 51% share of news revenue? You're actually angry about surveillance. Are you angry that "AI" is being used to automatically reject resumes on racial, age or gender grounds? You're actually angry about surveillance.
There's a very useful analogy here to the history of the ecology movement. As James Boyle has long said, before the term "ecology" came along, there were people who cared about a lot of issues that seemed unconnected. You care about owls, I care about the ozone layer. What's the connection between charismatic nocturnal avians and the gaseous composition of the upper atmosphere? The term ecology took a thousand issues and welded them together into one movement.
That's what's on the horizon for privacy. The US hasn't had a new federal consumer privacy law since 1988, when Congress acted to ban video-store clerks from telling the newspapers what VHS cassettes you were renting:
We are desperately overdue for a new consumer privacy law, but every time this comes up, the pro-surveillance coalition defeats the effort. but as people who care about conspiratorialism, kids' mental health, spying by foreign adversaries, phishing and fraud, and surveillance pricing all come together, they will be an unbeatable coalition:
Not every federal agency has gotten the message, though. Trump's Fed Chairman, Jerome Powell β whom Biden kept on the job β has been hiking interest rates in a bid to reduce our purchasing power by making millions of Americans poorer and/or unemployed. He's doing this to fight inflation, on the theory that inflation is being cause by us being too well-off, and therefore trying to buy more goods than are for sale.
But of course, interest rates are inflationary: when interest rates go up, it gets more expensive to pay your credit card bills, lease your car, and pay a mortgage. And where we see the price of goods shooting up, there's abundant evidence that this is the result of greedflation β companies jacking up their prices and blaming inflation. Interest rate hawks say that greedflation is impossible: if one company raises its prices, its competitors will swoop in and steal their customers with lower prices.
Maybe they would do that β if they didn't have a toolbox full of algorithmic twiddling options and a deep trove of surveillance data that let them all raise prices together:
Someone needs to read some Adam Smith to Chairman Powell: "People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices."
If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
