The Internet Didn't Break By Accident
What a conversation with Cory Doctorow taught me about why everything online feels worse and why that matters for all of us
This entire piece is a result of me listening to Cory Doctorow on this podcast while walking my dog, Mabel, in the cold February Iowa weather. The episode is here https://www.sixpixels.com/articles/archives/making-the-internet-suck-less-with-cory-doctorow-this-weeks-thinking-with-mitch-joel-conversation/amp/
I simply needed a place to gather my thoughts so here they are and no doubt so much more to come as I ordered his book, reading his newsletter again in great detail, and created my own learning course from all the terms he says that I don’t quite understand.
Anyhoo, here is my current headspace
You’ve felt it. The apps you loved five years ago feel like a chore now or simply don’t resonate with you. It is not because of you but because of the platform changing because they don’t care about you. Search results look like ads wearing a costume. Your social media feed shows you everything except what you actually want to see. Customer service is a chatbot maze that ends in frustration.
You probably blamed yourself and thought that maybe you’re just too old, too set in your ways. Or you blamed the companies that they got greedy, they got lazy.
Both of those explanations are wrong. And understanding why they’re wrong might be the most important mental model you pick up this year. Or maybe this is me projecting my own thoughts in hopes you join me.
The Word That Named What We All Feel
Cory Doctorow for those of you who don’t know is science fiction author(I reviewed one of many of his great books back in 2010), digital rights activist, and one of the sharper minds thinking about the internet who coined a term a few years back that went unexpectedly viral:
“Enshittification”
The pattern goes like this: A platform starts by being genuinely good to users. Then it shifts to extracting value from users to benefit business customers (advertisers, sellers). Then it extracts from those business customers too. Finally, it serves only itself until it collapses or pivots into something unrecognizable.
You watched this happen in real time. Facebook used to show you your friends. Then it needed you to pay to reach your own audience. Then it filled the feed with strangers, bots, and outrage content because that kept you scrolling. It’s not a mystery. It’s a predictable cycle.
But here’s what most of us miss:
Doctorow’s argument is that this isn’t about corporate greed getting worse. It’s about the policy environment that made greed more profitable than quality.
The “Big Short” Moment for AI
The AI conversation right now operates on two competing fantasies:
• AI will be so powerful it will replace most human workers, justify trillions in investment, and reshape the global economy.
• AI is already conscious, on its way to general intelligence, and might either save or destroy us.
Doctorow calls both of these “completely non-credible” and he makes the economic case bluntly.
The AI industry has spent $600–700 billion in capital expenditure already. They’ve committed to $2 trillion more. Their total global revenues? About $60 billion and even that number is “grossly inflated” by accounting that counts money companies loan to each other as revenue.
The math doesn’t work. Not even close.
To justify a $3 trillion valuation, AI companies need to displace $3 trillion in labor. But most of the jobs being disrupted first such as illustrators, content moderators, data entry workers are not highly paid. The economics of the pitch collapse under basic scrutiny.
He compares the current moment to the housing bubble: “Stein’s Law states anything that can’t go on forever, eventually stops.”
The credit default swap markets (the same instruments the “Big Short” investors used) are already swelling against AI companies. Someone has put $1 billion on the line betting that this all falls apart.
When the bubble deflates, what remains will be genuinely useful which is open source models running on commodity hardware, useful as specialized tools and plugins. Not civilization-transforming. Not conscious. Just... useful software. The kind we used to call a “cool plugin” before trillion-dollar valuations got attached to it.
The Fisher Price Steering Wheel
Here’s an image that will stick with you from this podcast convo that really hit home for me.
Most bosses know, somewhere in the back of their mind, that if they don’t show up to work, the business stays open, but if the workers don’t show up, the business shuts down. They’re sitting in the back seat with a Fisher Price steering wheel that isn’t connected to anything.
AI is the promise of hooking up the Fisher Price steering wheel to the actual car.
Not because it will work, it won’t, Doctorow argues, but because the fantasy of a business with no workers, no pushback, no “you’re wrong about this” from people who actually do the job, is intoxicating to a certain kind of leader.
This is why you can convince a boss to fire skilled workers and replace them with AI that can’t actually do the job. The chatbot will never tell you you’re wrong. It will gladly produce whatever you ask for, even if what you get is unusable. That’s the feature, not the bug.
For K–12 leaders: this pattern is showing up in education too. When AI is sold to administrators as a way to reduce the number of teachers or specialists needed, the same dynamic applies. The AI will say yes to everything. The educators pushed back because they had expertise worth listening to.
See Alpha School as an example
Lock-In Is Not Your Fault
One of the most liberating reframes in the conversation: you’re not on Facebook, Instagram, or any other platform you half-hate because of some character flaw. You’re there because of structural forces.
The Collective Action Problem
You can’t leave a platform alone. You need your people to come with you. And your people can’t agree on where to go. The person organizing your carpool is on Facebook. The support group is on Facebook. Your family abroad is on Facebook.
Mark Zuckerberg’s core insight: as long as you love your friends more than you hate him, you’ll stay. So he carefully manages being just barely tolerable.
Monopsony Power
This is monopoly’s less-famous cousin. A monopoly is a powerful seller. A monopsony is a powerful buyer. I had never heard this word or concept until the podcast. The ad platforms are so dominant that if you’re a publisher or advertiser, they represent 20–40% of your revenue. Losing 20% of revenue overnight is existential for most businesses which is why advertisers went back to Meta even after discovering $200 million a year in ad fraud. They had nowhere else to go.
The lock-in is structural, not a personal failure.
The Policy Choices Behind the Mess
Doctorow’s most important argument: enshittification isn’t primarily caused by bad people making bad choices. It’s caused by specific policy decisions made over the past 40 years.
Starting with Reagan and Thatcher, the Chicago School of Economics convinced regulators that monopolies were actually efficient. This gutted antitrust enforcement. Facebook could buy Instagram (Zuckerberg literally emailed his CFO: “it is better to buy than to compete”) and regulators waved it through.
The second key policy: anti-circumvention laws (in the US, Section 1201 of the DMCA). These laws make it a felony to modify products you own without manufacturer permission. This is why:
• Farmers can’t fix their own tractors
• Hospitals couldn’t fix ventilators during COVID shortages
• You can’t install a third-party app store on your phone
• Students and schools can’t modify locked educational software
• You can’t easily export your data from one platform to a competing one
These aren’t natural features of technology. They’re legal structures that lock in corporate control and prevent competition.
What This Means for Your Daily Life
Here’s how to translate all of this into practical thinking:
When AI feels underwhelming: You might be right. The hype is designed to justify investment, not to accurately describe capability. Use AI as the useful tool it is, not the magic it’s sold as.
When a platform gets worse: Recognize it as structural, not random. The company isn’t “losing its way.” It’s responding rationally to an environment that rewards extraction.
When someone says “just leave” a platform: Remember the collective action problem. Leaving alone costs you your community. This requires a policy solution (interoperability, data portability), not just a consumer choice.
When your district is excited about AI replacing staff: The chatbot will say yes to everything your educators pushed back on. The question worth asking: what expertise walks out the door, and what happens when you need to rebuild it?
When you hear “the market will sort it out”: This is a 40-year-old idea from a specific school of thought, not a law of nature. Markets need rules to produce good outcomes. Those rules are a choice.
The Bottom Line
The internet you miss isn’t gone because people got dumber or companies got meaner. It’s gone because specific policy decisions about antitrust enforcement, about the right to modify your own devices, about what mergers to allow created a world where being worse to users became more profitable than being better to them.
That’s bad news and good news simultaneously. It’s bad news because it’s systemic and hard to reverse. It’s good news because systems were designed by people making choices, and different choices are possible.
Don’t blame yourself for being stuck in these systems, and don’t believe that individual consumer choices will fix them. These are structural problems that require structural solutions.
Understanding that distinction between individual behavior and structural conditions might be the most useful thing you take away from this.
Based on Episode 1024 of “Thinking with Mitch Joel” (formerly Six Pixels of Separation), featuring Cory Doctorow. Doctorow’s newsletter publishes daily at pluralistic.net.
Coming next besides my regular newsletter will be my learning guide I am creating from all the things he talks about that I don’t quite understand.