The following is a rough transcript which has not been revised by The Jim Rutt Show or Cory Doctorow. Please check with us before using any quotations from this transcript. Thank you.
Jim: My guest today is Cory Doctorow. He is a prolific writer of science fiction and nonfiction books. The man writes books like—I don’t know how many books he’s written, but it’s a bunch. He’s also a journalist, an activist, a special adviser to the EFF. He’s an A.D. White Professor at Large at Cornell, and he’s proprietor of pluralistic.net, one of the few truly old school blogs that’s still around and still worth looking at. And as always, we’ll have the links on our episode page. Welcome back, Cory.
Cory: Thank you, Jim. It’s nice to be back on.
Jim: Cory is an OG on the Jim Rutt Show. He was on EP four when I didn’t know what the hell I was doing, to talk about Radicalized. Four novellas?
Cory: Yeah, was four novellas. That’s right.
Jim: And I really enjoyed them. They’re all good. That’s really worth reading.
Cory: Thank you.
Jim: And then later we had him on The Internet Con, another excellent nonfiction book. And his next book coming up next month, I believe, is The Bezzle—wait, I’m going to have to check that out. Sounds great. Very good. Today we’ll be digging into his most recent book, Inshitification: Why Everything Suddenly Got Worse and What to Do About It. Cory coined the term “inshitification,” though I’m sure many of us thought about something analogous, back in 2023 in a blog post called “TikTok’s Inshitification,” published on his blog on January 21, 2023. The word went viral, achieving escape velocity, and got named word of the year by the American Dialect Society, and it’s now in most dictionaries. Pretty good. Not bad for a blogger’s profanity.
Before we jump into the argument in the book—and it is an argument, not just a lament, which makes it interesting—could you define a couple of terms that you use which may or may not be familiar to our audience? Those terms are two-sided market, monopsony, and, of course, monopoly.
Cory: Well, a two-sided market, sometimes called a platform, is a firm that provides services both to sellers and to buyers and sits in the middle. You will recall, Jim, because you’re an old Internet person, that there was a time when we were very excited about the potential of the Internet to disintermediate things. A two-sided market is the intermediator. And there’s a reason that despite the fact that we were so all-fired excited about getting rid of intermediaries, we still have them, and they remain very central to what we do. And that’s because doing it all for yourself is hard. And there are a lot of people who would not get involved in all kinds of beneficial activities if they had to handle all the minutiae.
When I was growing up in Toronto, there was a writer I really doted on. He was kind of an outsider artist named Crad Kilodney, and he used to write, typeset, copy edit, print, publish, and sell his own books on a street corner in Toronto with a sign around his neck that said “Very Famous Canadian Author, Buy My Books.” And when he was feeling spicy, he replaced that with a sign that said “Margaret Atwood.” And Crad was amazing. He was a machine. He would even secretly record the drunks who abused him at night and then sell best-of cassettes. Really admirable guy and quite a writer. But there are a lot of writers I hear from and whose work I enjoy who are never going to stand on a street corner with a sign around their neck that says “Very Famous Author, Buy My Books.” And so we have booksellers and typesetters and publishers and copy editors and distributors and all the other pieces of the supply chain of the publishing industry, not because these are parasites with their hands out, but because they provide a real service.
But as my friend Tim Wu says, intermediaries have a tendency to get main character syndrome. They tend to lose sight of the fact that they are there to facilitate interactions between two different parties, and not because those two parties are there to make them richer. And when they are shorn of all constraint, as we’ll get into later, they metastasize, and they cease to be helpmeets who facilitate things, and they become gatekeepers who extract and control and rip off—which is why we were so excited about getting rid of intermediaries in the nineties.
Jim: Yeah. That is so funny. When that was the term of the year, what was that? About 1993, something like that?
Cory: Exactly. We weren’t wrong to want to do something about the fact that intermediaries had usurped the role of the buyers and sellers, the audiences and the performers, the different parties to these transactions, and had reoriented those transactions around the facilitator. It’s as though the most important part of a purchase was the cash register and not the person buying or the person selling. And honestly, the cash register is just a utility and should not be the thing we organize everything else around.
And then you asked about monopsony. So we all know about monopoly, but very few of us have heard of monopsonies, because there’s no board game with that name that you play with your family until you want to kill them and yourself. But a monopsony is the kind of flip side of a monopoly. It’s a market dominated by strong buyers, as opposed to a monopoly, which is a market dominated by strong sellers. And the thing about monopsony—it’s a little counterintuitive, but it’s much easier to attain and maintain a monopsony than a monopoly.
There’s a little thought experiment I like to use to illustrate this. So imagine every morning when you walk to work, you walk past five coffee shops, and one of them is your favorite. So 20 percent of that supply really matters to you. And that one coffee shop, it’s next to an office building, and the workers in the business in that office building are 20 percent of its gross receipts. So they have a buyer that amounts to 20 percent of their gross receipts. If the business in that office building goes under and none of those workers ever come back, that’s the end of that coffee shop. Most firms cannot withstand a 20 percent overnight drop in gross receipts. They fold. They can’t make payroll. They can’t make the mortgage. They can’t roll over their loans or service them, and that’s the end of it. So this firm is very sensitive to a 20 percent reduction in the market. Whereas you, if that coffee shop goes under—well, okay, your favorite coffee shop’s gone under. There’s four other coffee shops on the way to work. You’re barely going to notice.
And so monopsony is a really powerful idea in economics, but for complicated reasons—again, which we can get into if you’d like—ideologically, we’ve decided to pretend that monopsonies don’t exist, whereas we’ve paid at least lip service to the idea that maybe we should do something about monopolies.
Jim: Yeah. An everyday example which many of us have heard about is that Amazon now controls, what, 50 percent of book sales, something like that. It’s essentially impossible to be a mainline publisher today and not have a relationship with Amazon. One of the publishers tried for a while, but they folded pretty quickly. And that 50 percent number is extremely deceptive—
Cory: Because it counts a lot of books that aren’t normally sold through retail channels, that are usually sold direct. If you talk about certain genres—if you talk about science fiction—it’s not 50 percent. It’s, like, 90 percent. It’s much higher. It’s only because it’s being diluted by these other books that are sold in specialist channels.
Jim: Gotcha. Alright. Well, not only did Cory observe this inshitification, which we all noticed in one form or another with one name or another—what I found most interesting was he got analytical. He’s laid out a natural history, a four-stage life cycle, and what we could more or less call a theory of inshitification. Take it away, sir.
Cory: Sure. So there’s a descriptive component here that describes how the platforms go bad, and then there’s an analytic and prescriptive component that theorizes about why they’ve gone bad and what to do about it. And the descriptive component, as you say, sets out this three- or four-stage process, depending on whether you want to count the terminal stage in the list there. And I like to use Facebook as an example, a little case study.
So in 2006, Mark Zuckerberg wanted to expand Facebook beyond just college kids who had a .edu email address. His problem was that everyone who wanted to use social media already had an account on a rival service called Myspace. So he made a pitch to those users, and he offered them something pretty good. He said, look, I know you love hanging out with your friends there on Myspace, but I don’t know if you realize this—Myspace is owned by an evil, crapulent, senescent, immortal Australian vampire billionaire named Rupert Murdoch who spies on you with every hour that God sends. And if there’s one thing I know about social media, it’s that you shouldn’t be using social media owned by a weird creepy billionaire. And if there’s two things I know about social media, it’s that your social media shouldn’t spy on you. So come to Facebook, where we will never spy on you, and we will only show you the things that you ask to see—the feeds from the people you follow.
So this is part of stage one of inshitification: being good to end users. But the other part of stage one is locking those users in. And there are lots of ways that platforms have done this. I’ve been on your show before to talk about digital rights management. We see things like serial acquisition of rivals. Google pays Apple $20,000,000,000 a year not to enter the search market. That’s a pretty high-stakes way to lock users in, make sure there’s nowhere else they can go.
Facebook doesn’t have to do any of that, though. They’ve got the advantage of something economists call the collective action problem, which is how economists describe the fact that as much as you love your friends, you have to admit that they’re a giant pain in the ass. And if you can’t agree on, like, where you’re going to go for drinks this weekend, you’re absolutely not going to agree when it’s time to leave Facebook. And so you end up holding each other hostage.
And once Mark Zuckerberg knows that you’re stuck there because you love your friends more than you could possibly hate him, he can make things worse for you—this is stage two—in order to make things good for business customers. And so there are lots of different business customers Facebook courts, but one of them is advertisers. And he goes to advertisers and he says, do you remember I told these rubes that I wasn’t going to spy on them? Obviously, that was a lie. I spy on them from asshole to appetite. You give me a remarkably small sum of money, and I will target ads to them with exquisite fidelity. And because I’m such an honorable craftsperson, I will fill this building over here with quality assurance, anti-ad-fraud engineers who do nothing but make sure that there is no fraud in the network. So if you give me a dollar to show an ad to a specific kind of user, I will find that user and cram that ad in their face.
And so advertisers pile in, and they get locked in too, thanks to monopsony. They become dependent on the users who are locked to each other. And that’s stage three of inshitification, which is making things worse for businesses in order to make things better for the shareholders and the executives.
And this is, I think, a place where my theory breaks with a lot of other criticisms of tech. There’s a lot of people who say, oh, if you’re not paying for the product, you’re the product. They kind of posit that decent treatment from a firm is a kind of customer loyalty perk and not that it’s something that arises out of market incentives—that firms that are afraid of facing consequences for maltreating you, whether that’s competitors or regulators or losing key personnel or having new firms enter the market to undo the mischief that they’re doing, those firms will abuse you if they can.
And so Facebook starts to abuse its business customers. Advertisers find that ad prices have gone way up while ad targeting fidelity has gone way down, and ad fraud has exploded to levels that can barely be appreciated. You know, in 2017, Procter and Gamble thought that perhaps they weren’t getting good value for money with their surveillance ads—the industry calls them behavioral ads. And so they took their $200,000,000 a year surveillance advertising budget to zero, and they saw a zero percent drop in sales, because to a first approximation, $200,000,000 was just going down the fraud hole.
Meanwhile, publishers, who are another big group of business customers who became locked to the platform, found that rather than getting their stuff recommended to users who hadn’t subscribed to them, even to reach their own subscribers, they had to put longer and longer excerpts from material that had originally appeared on their website onto Facebook. And then eventually, if they put a link back to their website on Facebook, it would be treated as a malicious link and the post would be suppressed. And so they had to effectively replicate their website on Facebook, and they could only monetize it using Facebook’s corrupt advertising system.
So this is getting to the final stage of inshitification, where all the value has been extracted from businesses and end users, leaving behind a kind of void with a homeopathic residue of stuff you’ve asked to see, while advertisers and publishers are charged hundreds of billions of dollars to fill that void with things that they’re getting ripped off for.
You’d think that Facebook would be happy about this, but they’re not, because they can appreciate that this is a very brittle equilibrium. The difference between “I hate Facebook, but I can’t seem to stop coming here” and “I hate Facebook, and I’m never coming back again” is razor thin. You get a livestream mass shooting or a whistleblower, people bolt for the exits. Shareholders get very nervous because they’re wondering when the day will come that Facebook stops growing and therefore should be revalued not as a growth stock that has a very high multiple of its earnings, but rather as a mature stock that trades at a much lower multiple. When users leave, these shareholders stage mass sell-offs. This provokes panic among Facebook executives whose portfolios are very top-heavy with Facebook stock, because they get paid in Facebook stock. And being technical people, they don’t call it panicking. They have a technical term. They call it pivoting.
And this is how it came to be that one day Mark Zuckerberg arose from his sarcophagus and said, hearken to me, brothers and sisters, for I’ve had a vision. I know that I told you that your future was for you to argue with your most racist uncle using a primitive text interface I created in my dorm room to nonconsensually rate the fuckability of Harvard undergraduates, but the actual future will consist of you and everyone you love being transformed into a legless, sexless, low-polygon, heavily surveilled cartoon character so that I can imprison you in a virtual world I stole from a twenty-five-year-old dystopian satirical cyberpunk novel, that I call the metaverse.
And that truly is the final stage of inshitification. That’s the giant pile of shit. Though in this case, it didn’t work—or at least hasn’t worked so far.
Jim: Well, the metaverse didn’t, but we’re still stuck there. Facebook still has billions of users across its platforms.
Cory: While the metaverse didn’t work, it did stop investors from doing that mass sell-off that would have made Facebook very vulnerable. Because, of course, when your firm is a growth firm and trading at a very high multiple of its earnings, other firms are very excited to get your shares, which means that if you want to do an acquisition, you can do stock deals, which is great because you make stock on the premises by typing zeros into a spreadsheet. But once people are less interested in your stock and they expect cash—well, you can’t really make cash on the premises, not if you don’t want a visit from some very angry men from the Treasury Department. So now you find yourself bidding with things you can only get from other people, like creditors and customers, who are a little harder to get stuff out of than your spreadsheet jockeys who issue new shares.
And so you tend to enter a kind of death spiral where you lose bidding wars for personnel and key acquisitions, both of which are an extremely easy way to grow. What could be an easier way to grow than acquiring a rival firm? You don’t even have to grow a market. You just take someone else’s. And so now you have to grow the hard way, by making things people like. And I think Mark Zuckerberg has, like, only done that once in his life. Everything else he’s done that people like, he bought from someone else. He’s not like Willy Wonka. He’s Rich Uncle Pennybags.
Jim: Yeah. Indeed. And now he’s trying it again with AI.
Cory: With AI. In fact, at this point, people are so skeptical of AI that now it’s superintelligence—in the same way that it wasn’t web three, it was metaverse. It just kept going through these iterations. Like, if I call it by a new buzzword, will you finally believe that this is not a bald spot? It’s actually a solar panel for a sex machine.
Jim: Alright. Well, that’s telling, from the Facebook perspective. Another story that you tell, which I think is maybe even a stronger example of the platform-becomes-incentivized dynamic, is what used to be one of my favorites, which is Amazon. I had to admire a guy who started a business in his parents’ garage—and very few people still remember this, but in the beginning, he was being outspent and outgunned by lots of other competitors. He was nobody, basically. There was a company called Value America, one of those footnotes of Internet history, that had hundreds of millions of dollars worth of funding and was selling all kinds of stuff online back when Bezos was still selling books and CDs mostly. But they were selling the wrong stuff at the wrong time and not executing well. They pissed their way through $200,000,000 and they were gone. And Bezos just kept executing. He managed to avoid getting crushed by Barnes and Noble and Borders and all these guys. And I had to admire him. He executed well. He seemed to understand the strategy of what rate to try to sell things online, etcetera, and negotiated on behalf of the customer. Life was good. But then what happened?
Cory: Yeah. I think you raise a really good point, which is that Bezos ran the firm in a very customer-centric way. I mean, he was always very hard on his suppliers. His mantra has been for decades, famously, “your margin is my opportunity.” And so he was always really hard on suppliers, but he was very good to customers. And he used to, like, answer emails from customers who were upset about Amazon’s customer service issues, and he would just solve them.
I sent him an email after our daughter was born. We live up six flights of stairs in London. We’d ordered a new coffee machine, and the doorbell was broken, and they wouldn’t call my phone to tell me that they were downstairs. And it just was not being delivered. And I was at wit’s end. I wasn’t sleeping. I had a newborn. I wanted my coffee. And I emailed him, and he literally sorted it out. Like, I just sent an email to jeff@amazon.com.
So there was this period where he was really focused on customers. But when he became too big to fail—which he got not just by executing, but also by serial acquisitions, and not just any serial acquisitions, serial acquisitions basically at gunpoint—you’re right that he executed really well, but then he started buying companies. And eventually he met a company that didn’t want to sell, a company called diapers.com. No points for guessing what business they were in. Diapers.com had a really good business, and they were actually owned by a group that had a bunch of other successful vertical online retail firms. And they didn’t want to sell. And so Amazon tapped the capital markets. And unlike Value America, which accidentally blew through $200,000,000, Amazon deliberately blew through $200,000,000 selling diapers below cost. And within a couple of months, diapers.com was sold to Amazon for pennies on the dollar. And the holding company that owned it sold to Amazon even though they had a better offer from Walmart, because they were afraid that Amazon would do this again to all their other companies if they didn’t sell to Amazon.
So those $200,000,000—we could call this execution—it was money well spent, because no one ever turned down Jeff Bezos when he came to buy their company again, because they knew what would happen if they did. And so Bezos bought his way to dominance. He did build some very excellent systems—AWS and his logistics systems and so on—but he also found ways to corner markets.
Having done both of those things, by being a competent monopolist, he became very insulated from the consequences of being bad. So one of the things that he did was he instituted a most-favored-nation policy with his retailers. Most-favored-nation is a policy that says a retailer has to offer their best price on your platform, and they can’t beat that price anywhere else. Now Bezos was able to do this because the majority of affluent households in America had Prime. And Prime is a form of lock-in. It’s basically paying for your shipping in advance. So you’d be a mug to pay for your shipping in advance and then pay for shipping somewhere else when you’ve already paid for it on Amazon. So people who have Prime start and end their retail journeys on Amazon.
America has become so unequal through a series of what economists call K-shaped recoveries—where rich people get richer and poor people get poorer—that nearly all the consumption spending is crowded into, like, the top decile of homes. So if you want to sell to those people who buy everything, those are the same people who all have Prime. So you have to be on Amazon.
So what Amazon was then able to do was jack up their junk fees. When I wrote the book, the junk fees for selling on Amazon were 45 to 51 percent for a seller. So if you bought something on Amazon, 45 to 51 percent ended up in Bezos’ pocket. Today it’s 50 to 60 percent and rising. But, of course, firms don’t have 60 percent margins, so what they do is they raise prices. But because of most-favored-nation, they raise prices at Walmart and at Target and at their factory store and at the mom-and-pop shop as well as on Amazon. So Amazon has driven up prices across the economy.
And some of the junk fees that they levy have these second-order effects where they make things worse yet again. So the most profitable line of business that Amazon is in—now worth $80,000,000,000 a year—is what they call an advertising business. And the reason I qualify it with “what they call” is it’s not advertising. It’s payola. It’s money that you pay to Amazon to be listed at the top of the search results, irrespective of whether you’re the best match for that search. And so what you have every time you run a search is an auction in which vendors bid up—bid down each other’s margins, rather—to give free money to Amazon.
This $80,000,000,000—it was $35,000,000,000 when I wrote the book, it was $55,000,000,000 when the book came out, it’s $80,000,000,000 now—is worth as much as all the newspapers in the world’s gross revenue times two and a half. So you could have two and a half world news medias for what they get from this. It makes search worse. It makes products more expensive. It benefits no one except Amazon. It is just a pure force for evil in the world, and it’s the most profitable thing Amazon does. It’s the most profitable line of business among all the lines of business Amazon is in.
Jim: Including some good ones like AWS.
Cory: Yeah, like AWS, like its logistics network, Fulfillment by Amazon, which it charges merchants a fortune for. In fact, it charges so much for access to Fulfillment by Amazon that Amazon’s own fulfillment is free. So Amazon has the cost of its own deliveries fully covered by its competitors—the platform sellers that it competes with on its own platform.
Jim: I could feel there was something in my stomach that told me Amazon was starting to circle the drain when they put those advertisements in. When you did a search for, you know, I’m looking for a new receiver for my stereo system—sponsored, sponsored, sponsored, sponsored, sponsored. And I go, what the fuck? It used to be the algorithm was actually pretty good at knowing my tastes more or less. And then you do a second-order scan: alright, it’s got to have four and a half stars, and then you’ll look at it. But now it’s like, alright, what are these 12 sponsors? Is it all horseshit? Is some of it good? And it turns out it’s a mix.
Cory: Well, on average, the top result on an Amazon search is 29 percent more expensive than the best match. The top row is 25 percent more expensive, and the best match is usually on the second screen, somewhere around position number 17.
Jim: That’s pretty scary. Yeah. When I learned that in the book, I had no idea that the Amazon platform was taking on the order of 50 percent of revenue from vendors. I know they all bitch about Amazon, but I was figuring it’s 15 or 20 or 30 percent, even like those app stores. But when you laid out pretty clearly that you could say it’s 50 percent—that’s a staggering number for a historically low-margin business like retail.
Cory: Yeah. And it’s a real drag on the whole economy. So when Karl Racine, the attorney general of DC, sued them, that was the basis of the lawsuit—that everyone who bought anything anywhere paid a premium because of Amazon’s monopsony and the attending most-favored-nation status. So even if you never bought anything on Amazon, if you only bought things direct from sellers, those sellers probably can’t survive without being on Amazon. And if they are on Amazon, they’ve had to raise prices at their direct sales store. Otherwise, Amazon will kick them off the platform.
Jim: In some ways, that’s worse even than Walmart. Walmart has always been horrible for vendors. If you’re a vendor who sells to Walmart, you say, well, the only reason I sell to Walmart is to get the margin for the rest of my business. But at least Walmart—at least it didn’t used to, when I was still active in business—charge outrageous transaction fees. It just beat the hell out of you on price.
None of this is mysterious. The forces that gave rise to this—
Cory: —are not the mysterious great forces of history or the iron laws of economics. They’re specific policies, and inshitogenic policies made the worst ideas and the worst people the most profitable. So you mentioned Walmart. The reason Walmart is Walmart is twofold, and it’s more or less the same thing that gave us Amazon. It arises out of a decision by first Carter and then Reagan to stop enforcing antitrust laws. So Walmart and the rest of the economy were given a break from enforcement of something called the Robinson-Patman Act, which is an act that bans preferential retail pricing. Actually, Biden’s FTC brought a case against Walmart under the RPA—the first one in a generation—which eventually the Trump administration just sort of let die. But the allegation in that was that Walmart and Pepsi had colluded to raise prices at Walmart on the condition that prices would be raised everywhere else even more. So Walmart would still have the everyday low price, but it would be high, and everyone else would be higher.
That’s where we got Walmart—exemptions from RPA enforcement. And the other place we got Walmart was exemptions from the parts of antitrust law that dealt with monopsonies. So the theory of antitrust that Carter and Reagan embraced, and that became the doctrine of the last forty years, is something called consumer welfare, which basically says monopoly is efficient. And if you observe a circumstance in which we all buy our goods from a single vendor, you should infer that that vendor is the best, and it would be perverse to punish them for making us so happy we switched to buying their products.
And so there was an encouragement of monopoly—an encouragement of monopoly provided that you didn’t raise prices. So you could do anything you wanted to your suppliers, but not to your customers. That’s the consumer welfare part. And so what that meant is that the protections that we’d historically had for suppliers—farmers, manufacturers, small goods makers, writers—all of those protections went out the window. And then once these firms were able to squeeze their suppliers, eventually they ran out of runway to squeeze the suppliers. And because they were able to squeeze the suppliers and violate Robinson-Patman and do preferential pricing, they squeezed out all their competitors too. That’s how Walmart killed Main Street. And once there was nowhere else for you to go, they could raise prices on you.
And this was the thing that the progenitors of our antitrust law understood—that if you allowed monopsonies, you would get monopolies, that they were like peanut butter and chocolate, they just went so great together. And it’s the thing that was denied by these neoclassical economists that Reagan and Thatcher and Mulroney loved so much. I think they were vindicated, because we have arrived at a juncture where all of our monopsonists became monopolists too.
Jim: I was actually involved in one of those monopoly-monopsony games, which is in legal publishing, where there’s a handful—three or four big companies. And I was at Thomson, now Thomson Reuters. And we were number two, and I was part of the team that acquired West, which was number one. And we had to go through the antitrust clearance with the Department of Justice, I believe it was—usually it’s either FTC or Department of Justice. We got thrown to Justice because it was legal stuff. And we made exactly the consumer welfare argument and gave up just the smallest amount of divestiture, etcetera. And we didn’t raise prices for a while. And then—100 percent—once we had 70 percent of the market in certain categories and the five-year window had expired, Katie bar the door, people.
Cory: Sure. I mean, look. If you go ask a physicist to describe a little thought experiment, they’ll say, well, first, let us assume a perfectly spherical cow of uniform density on a frictionless surface. And if you go to a business school professor, they will do the same kind of thought experiment. They’ll say, first, let us assume a firm that can charge infinity for its outputs and pay nothing for its inputs. And obviously that’s not a real business—unless you’re an academic publisher, in which case that’s exactly how you operate.
Jim: That’s pretty close.
Cory: Everyone else has to take some discipline. Everyone else has to worry about rivals, about competitors. In the case of tech, there was a long period where they had to worry about tech workers who often were extremely ardent believers in user welfare, because they themselves had found so many advantages from getting online—expanding who they could be, finding the words to describe what they felt and who they were, escaping the circumstances of their birth. They really wanted to confer those benefits to other people. I sometimes call them Tron-pilled, because they wanted to fight for the user.
We all know those people. There’s the cliché of the tech bro, but really—find someone who loves how technology has made their life better and then ask them about it. Ask a Linux user about Linux, and then try to make them stop talking about it. They want everyone to experience these benefits. And tech workers were in short supply. There were ten bosses at the factory gates who wanted to give your engineers a job, and the NBER says that those tech workers were adding a million dollars a year to your bottom line. And so if they walked out the door, a million dollars walked out with them. And so when those tech workers said, “I won’t inshitify that thing I missed my mother’s funeral to deliver on time,” it stuck—because you couldn’t fire them. And if they quit, you couldn’t replace them. And if they didn’t want to do it, it just didn’t happen.
And then, distinct to tech, you had this extremely powerful form of new market entry that is not like the market entry you get in other sectors. The only kind of digital computer that we know how to make is what computer scientists call the Turing-complete universal von Neumann machine, which is how computer scientists describe a calculating engine that can compute every valid program—which means that every computer can run every program, which means that every program has a program waiting to come into existence and counter your gambit with one that is more attractive to users. So if you lock the printer so it only takes your ink, someone can enter the market not just with cheaper ink, but also with an unlocking program.
But in 1998, Bill Clinton signed a law—the Digital Millennium Copyright Act. Section 1201 of the DMCA established a felony punishable by a five-year prison sentence and a $500,000 fine for modifying digital technology that had been designed to resist modification. And as a result, you ended up with something that looks a lot like felony contempt of business model, where a company that skins its technology in the thinnest layer of IP can criminalize entering the market with a complementary product or service that might tempt your customers away.
And so now we have this circumstance where firms merge to monopoly, and once they merge to monopoly, they capture their regulators—because five companies can capture a regulator, whereas a hundred companies can’t even agree on when they would have the meeting to capture the regulator. They don’t have to worry about their workers, and no one new is going to enter the game because they get to decide who can compete with them. And now they abuse you—and they abuse you for the same reason your dog licks its balls: because they can, and we don’t stop them.
Jim: Yeah. I would add there’s one other thing here, which is some of the social welfare or social change work that I do. Sometime around 1975, the ethos of business stopped being constrained by what we would laughingly now call morality. I remember I started work in 1975, and I happened to work for two good companies, both of which would not do things that would be profitable in their domain because they thought they were wrong. During the next decade, the ethos became, well, whatever is arguably legal and profitable, not only should I do, but I must do. And then we went further, and I would say by the early 2000s, the ethos of business had become—and I think Google and Facebook are great examples of this—not only do I do what’s arguably legal, I even do what’s illegal as long as I know the costs of getting caught are going to be smaller than the benefits I accrued. You know, Google and Facebook—how many times have they been fined hundreds of millions or billions of dollars? You can guarantee that they made two or three times whatever they were fined.
Cory: Yeah. A fine is a price.
Jim: Exactly. It’s part of the price of doing business. So I do believe that in addition to these other structural concerns, there was a moral decay that started sometime in the late sixties or mid-seventies—the Jimmy Carter, Reagan thing.
Cory: Yeah. You know, you’re describing this, and it sounds an awful lot like my friend Ed Zitron’s theory of the rot economy, which he traces back to Jack Welch and shareholder supremacy, Milton Friedman, and so on. And certainly these people were, in a sense, moral philosophers of immorality. There’s this argument: when it’s railroading time, you get railroads. So what was it? Because they didn’t invent immorality. There were cheaters all along. Cheaters could prosper, at least in the short term or in certain ways. There was an argument for cheating at different times. So what stopped the cheaters from taking over the firms?
Some of that, as you say, has structural elements—shareholders who were not buy-and-hold but were shorter-term and more interested in the coming quarter; the rise of institutionalism; CEO compensation in shares, executive compensation in shares, which incentivized them to goose the share price through financial engineering rather than building a sustainable business. All of that stuff. But I think it’s always easier to resist the devil on your left shoulder if the angel on your right shoulder has more to say than “it’s wrong.” If the angel can say “it’s wrong and you’ll get caught,” it’s a lot easier to listen to the angel.
And so there’s an example that Ed Zitron reported out, that I cite in the book, about how Google went bad. It dates from 2019, 2020. So Google had capped out its search revenue growth for the very good reason that they had a 90 percent market share. And firms with 90 percent market shares do not grow. I mean, you can raise a billion humans to maturity—that’s a product called Google Classroom—but it takes a minute, and shareholders are looking at the next quarter.
And so when the DOJ sued Google over monopolization—there were two successful DOJ cases and then a private case, so Google’s now a thrice-adjudicated monopolist—in the search case, they published internal memos from Google that detailed the power struggle within the firm, and you can really see how the external incentives gave the worst people at Google the most power.
So the solution that one faction of Googlers came up with for this revenue stagnation was making search worse. This guy Prabhakar Raghavan, who was an ex-McKinsey guy, ex-Yahoo, thought: if we make search worse, you’ll have to search more than once. And every time you search, we show you ads. So if we show you ads three times for every query, that’s triple the revenue from that query. And he’s in charge of search revenue, so this makes a lot of sense to him.
Now his opponent is a faction led by a guy called Ben Gomes. And Gomes is this old-school Googler. He was in charge of Google search tech, the server build-out. He started with a couple of computers under a desk and built out all the data centers in the world, and then he’s actually hands-on running search tech. And he’s palpably very horrified by this. He’s like, why did I give my life to this company to make search worse? That’s a terrible idea. But that’s all he’s got.
Because at this point, Google is bribing Apple $20,000,000,000 a year not to enter the search market. They’ve bought all the shelf space in the search market. So whether you buy your device from whichever manufacturer, whichever operating system it runs, whichever browser you choose, you are searching with Google. Every search box you find in the wild is wired into Google servers, which chokes off the market oxygen. Why invest in a big, ambitious search tool if no one will ever find it?
And so Raghavan’s argument is: why are we spending $25,000,000,000 a year to buy the search shelf space if we’re not going to take advantage of it to increase our revenue? What are we worried about—that people are going to use Bing? Come on. And so he wins. Google made search worse. That’s why Google sucks. That’s why we’re like, well, maybe this AI search isn’t so bad, because Google really sucks.
Jim: It really sucks. I mean, I actually even wrote a search engine one time, and I was friends with the people at AltaVista. This was an area that I’ve always followed quite closely. Google is now actually a piece of shit. How could this be? The amount of talent they have there, the amount of servers they have. But as you say, it’s obvious that if your job is to sell green-highlighted lines, if you can make somebody get three sets of green-highlighted lines instead of one, money in the bank.
Cory: Absolutely. And there’s a funny coda to this. So my now-retired novel editor, the wonderful Patrick Nielsen Hayden, is the most talented autodidact I know. He read a science fiction novel that he bought on a pharmacy spinner rack at 14 in suburban Phoenix, had his mind blown, dropped out of high school, went across the country publishing science fiction fanzines, ended up in New York, and finished his career as a vice president of Macmillan. And I was at his place—he and his wonderful wife Teresa moved back to Arizona—and I was at their place for the Tucson Writers Festival. We were sitting around using our laptops and drinking coffee in the morning, which is what we do whenever I stay over at his place. And he said, have you tried Kagi yet? And I was like, what’s Kagi? He said, oh, it’s a search engine, but it’s like the first time you use Google. You know, you were trying these long complex queries with AskJeeves and AltaVista, and then you gave three vague keywords to Google, and it told you exactly what you’re looking for. That’s what Kagi is like, but it’s $10 a month.
So I started using it. It was really great. Within—so I paid them the $10, and then by the end of the day, I was giving them $20 so I could get accounts for my wife and daughter, because it was so good. And I thought, wow, this is really amazing that they were able to enter the market, because they’re just a little startup. And then Jason Koebler, who’s a great tech reporter, one of the founders of 404 Media and the former editor in chief of Vice Motherboard, wrote an article about Kagi. And in that article, I learned that Kagi is a front end to Google’s search index. So they are syndicating and remixing Google’s search results with a staff of a couple of dozen.
So Google, using all of its servers, all of its talent—as you say, a large plurality of all the CS and computational linguistics and applied math people to graduate from every major university in the world for the last twenty-five years are working at Google. Kagi’s got, like, a couple of dozen people working on startup salaries, and they can do not just a slightly better job with Google’s search results, but a massively, orders-of-magnitude superior job. Google just doesn’t want to. That’s the only conclusion you can reach. They are not just too big to fail. They are too big to care.
And this is a thing we’ve understood for a long time. You will remember when Lily Tomlin used to do these bits, first on Laugh-In and then on Saturday Night Live, where she’d play a telephone operator, Ernestine, and she’d do fake advertisements for AT&T. And they would always end with her turning to the camera and saying, “We don’t care. We don’t have to. We’re the phone company.”
Jim: And when you’ve got a 90 percent search market share, you don’t care. You don’t have to. You’re the search company. It’s just amazing. I think I don’t even use Google except to find the nearest Mexican restaurant or something. I’ve moved essentially all my business to Perplexity. It has not yet become corrupted. It does a very balanced job of search with AI. I was trying to search for a fact—I was trying to search for the name and date of your TikTok inshitification blog post. And I tried Google three or four searches and didn’t come up with it, or wasn’t clear what the answer was. Hit Perplexity. I go, what was Cory Doctorow’s blog post date and link for the one that introduced the idea of inshitification? Fifteen seconds later, there it all was, plus a pretty nice description of inshitification.
Cory: So I hate to do this to you, but it was wrong. There was a post before that in 2022. So the TikTok post went viral, but it wasn’t the first mention. So Perplexity is doing an okay job, but far from perfect.
Jim: Ah, well, okay. Shame on Jim for believing the goddamn AI. But it seemed—because I had seen 2022 brought up—except that you brought it up. Yep. Well, shame on me. Shit happens. Alright. Let’s—we are just a few minutes here. Let’s get back to: you talked about the fact that bad actors will act less bad if they think there’s someone who’s going to punish them if they get caught. And so let’s go all the way back to your diapers.com example. That is clearly illegal under antitrust law, under the theory of predatory pricing, and yet nothing happened. How do you explain the chain, again and again and again, of nothing happening?
Cory: Oh, well, we decided not to enforce the law. The explanation’s really easy. You just had successive generations of political appointees sitting in the attorney general’s chair who directed the DOJ, and sitting at the top of the FTC, who directed their own staff not to enforce the law. So Tim Wu—again, my old pal, we went to elementary school together—he worked at the FTC during the Obama years and then was Biden’s special adviser on tech and antitrust. And he tells this incredible story about how the Waze merger got through, because Google bought Waze, and Google already had Google Maps. And he could never figure out how this passed muster at the FTC. So he talks about having drinks with someone he used to work with there and saying, how did you get this through? And they said, well, we started from the presumption that we wanted Google to buy Waze, and then we worked backwards until we found a justification. And we decided that Google Maps was what you used to find out where you were, and Waze was what you used to find out how to get there.
And it’s a ridiculous thing. First of all, it’s not how anyone uses Google Maps or Waze. But also, even if it were true, it’s such a dumb distinction without a difference that it requires this act of will. It’s a bit like—I don’t know if you remember when legal cannabis first came in, but you needed a medical pretext. And so you’d go to the doctor to get your cannabis prescription. They’d say, so tell me about your headaches.
Jim: Or your football injury or something.
Cory: What headaches? And they’d go, tell me about your headaches. And you’d say, I get headaches. And they’d go, oh, well, if you get headaches, you need a prescription for cannabis. And this was the basis for all of these mergers going through—these flimsy excuses.
I mean, you want to talk egregious. One of the things about these mergers that made it possible to allow them to continue is that under the consumer welfare theory, at the time of the merger, you had to be able to show that the intention of the merger was anticompetitive. And of course, who knows what evil lurks in the hearts of men? How do you divine someone’s intention? Unless they’re Mark Zuckerberg and you put down every bad thought you have in an email that is then discovered by the Department of Justice, which is exactly what happened.
So when Zuckerberg bought Instagram, he sent an email to his CFO that was like, hey, Bob. You want to know why we’re buying this 12-person company with 25,000,000 users, most of whom don’t log in every month, for a billion dollars—which back then was real money for an acquisition. Well, I’ll tell you: people like Instagram better than they like Facebook. When they go to Instagram, they don’t come back. So this is as close as you get to, like, hey, Bob, that guy we’re thinking of killing—just so you know, it’s a murder, and now that I mention it, I’m premeditating it. And yet the DOJ waived the merger through even with that in hand.
So it wasn’t even a pretext. The actual doctrine was: we will just let every merger through. And the fig leaf was: unless we can prove that it was undertaken for anticompetitive purposes. And the reality was: even if we can prove beyond a shadow of a doubt that this merger is anticompetitive, we will still let it go through.
Jim: Yeah. Very, very bad. In fact, the truth is, from my numerous years in business at every scale, essentially every acquisition is anticompetitive in some sense.
Cory: Right. At the margin or right in the middle. Instagram is certainly right dead center. I might draw a distinction for distressed assets—a firm that’s going bankrupt, that’s purchased at the end of its life for whatever patents or key personnel or plant. I can see that not being exactly anticompetitive in the way that we think of it.
Jim: What can our listeners do? What can citizens in Canada, the United States, the UK—what can we do, if anything, to try to push inshitification back?
Cory: Well, I like that you said citizens and not consumers, because you’re not going to consume your way out of this. Shopping your way out of a monopoly is like recycling your way out of a wildfire. It’s just silly. And for forty years, we’ve been told that you should vote with your wallet. I think it’s a conspiracy from billionaires who have wallets that are thicker than ours, but there aren’t that many of them. And so it’s the kind of vote that they can reliably win—the vote that you take with your wallet instead of the vote that you take as a citizen.
During the Biden years, we had a period when it really looked like the federal government was getting very serious about antitrust. That’s dead now. And that means that internationally, it’s dead too. Like all the best Americans, I’m Canadian. And Canada saw a real surge in antitrust enforcement following the lead of the Americans. We got new powers for our Competition Bureau, which had historically been terribly weak. Canada’s Competition Bureau had only challenged three mergers in its history and had never succeeded in challenging a merger, but it got new powers in 2024. It was really on the bandwagon. But now Trump is saying, if you try to enforce antitrust laws against my tech companies abroad, I will rain down fire upon you. So the European officials who worked on the Digital Services Act—a very important antitrust law—they’ve been sanctioned. They can never enter the United States again. And Trump has just requisitioned the private communications of more European officials so he can expand the sanctions list for anyone else who might have thought about taxing or enforcing laws against tech companies.
But I don’t think that means we’re out of options. At the state level, there’s still a lot of work that can be done. Antitrust law can be carried on at the state level, and state AGs have standing to challenge mergers and anticompetitive conduct. And as we all know, AG stands for aspiring governor, and attacking unpopular mergers—here in California, well, I’m in London right now, but I live in California when I’m not in London—our AG is set to challenge the Paramount-Warner’s merger. You have the states carrying on the case against Ticketmaster after Trump let them buy their way off the hook, and actually securing a victory. And so there’s work to be done at the state level.
Internationally, I think our best bet is for the public to start talking to their policymakers about repealing their laws that prevent reverse engineering. Because, as Jeff Bezos says, your margin is my opportunity. If Apple’s getting 30 cents on every dollar spent by a Canadian in the App Store—so when a Canadian pays a Canadian newspaper a dollar for a news item, 30 cents comes to rest in Cupertino—that’s a market for Canadians to enter by making jailbreaking kits for iPhones that have app stores that take three percent instead of 30 percent.
And just like the $80,000,000,000 a year that Amazon makes on ads is its most profitable line of business, Apple’s payment processing is also its most profitable line of business. It’s $100,000,000,000 a year. That margin is someone’s opportunity. The base cost of processing a transaction is south of three percent. So you give people a 90 percent discount, you can still make $10,000,000,000, and a grateful world will buy that tool from you.
Canadians could go into business jailbreaking tractors and selling every farmer in the world the software that they need so that when they fix their tractor, it just starts working again—as opposed to the situation now, which is you fix your tractor, and then you call John Deere, and they make you wait two to three days while hail wipes out your harvest, and then charge you $200 for a technician to type an unlock code into your console before your tractor starts working again. That’s the thing every farmer in the world would buy.
Mechanics currently spend $10,000 per model per year for diagnostic tools that turn “check engine” into an actual diagnosis. You could sell every mechanic in the world a subscription for $25 a month that diagnoses every car. And that could be Canada, could be Mexico, could be Europe. Anyone could become that nation. And the only thing stopping these countries from doing it historically was that the US had threatened any nation that allowed reverse engineering of American products with tariffs.
So Trump has given us a liberation day. Because if someone threatens to burn your house down unless you do as you’re told, and you do it, and they still burn your house down, you are a sucker if you keep doing what they tell you to do. So you can just stop—especially when Trump is now weaponizing the tech companies to shut down officials who take actions on behalf of sovereign nations that he dislikes.
A high court judge in Brazil who sentenced Jair Bolsonaro lost all of his Outlook accounts, calendar, working files, his email address that he uses to log in to everything else—so he’s locked out of everything—his address book, everything. He’s effectively been bricked. The same thing happened to the chief prosecutor of the International Criminal Court when he swore an arrest warrant for genocide against Benjamin Netanyahu. So Trump is really making the case that he’s not going to have to roll tanks to steal Greenland. He can just brick Denmark. He can Hormuz the world supply of Lego, Ozempic, and black licorice at the stroke of a pen.
And so this is the moment for us to really start thinking about how we replace this American inshitified Internet with open, auditable, transparent, sovereign digital public goods that are maintained internationally and run and controlled locally within each sovereign territory’s borders, under the control of the people who use them. And I think that’s a pitch that everyone should be making to their policymakers. I think whoever gets there first is going to turn America’s trillions into their national billions, and the remainder is going to become a consumer surplus that everyone in the world—including Americans—is going to enjoy. Because if Americans can buy reasonably priced insulin from Canada over the US Postal Service, they can buy software from anywhere in the world. All you need is a browser and a payment method.
Jim: Well, that’s a wonderful vision. I won’t hold my breath, but I will try to help.
Cory: I don’t think we’ve ever been closer. It doesn’t mean it’s going to happen, but I’m full of hope.
Jim: Well, I do agree with you that the right to repair is kind of an easy entry point—so hard to argue against. In fact, on principle, I will not buy John Deere tractors for my farm. I only buy Kubota tractors because I won’t deal with their fucking shit. But anyway, as you pointed out, my little farm ain’t enough to change their policy. Well, I want to thank Cory Doctorow, one of the more interesting and important thinkers of our time, for appearing again on the Jim Rutt Show. And let me say one more thing, which is that the best way to get involved in these issues and stay involved is to join the Electronic Frontier Foundation, where I’ve worked for twenty-five years. And I was member number seven.
Jim: So, yeah. Excellent. Good man. And I was one of the financial backers for quite a while, till they got big enough and rich enough that they didn’t need my money anymore.
Cory: Well, we’re still member supported. We don’t get a lot of grant money. We don’t get corporate money. We run on small-dollar donations from individuals. EFF.org, people. Go give them $50. Come on.
Jim: Thank you, Jim. Alrighty. Nice talking. Bye.