How Silicon Valley enshittified the internet
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this compelling Decoder episode, Sarah Jeong interviews prolific author and internet activist Cory Doctorow about his new book on "enshittification" - a term he coined to describe the systematic degradation of digital platforms. (04:13) The conversation explores how platforms like Facebook, Google, and others start by being good to users, then abuse them to attract business customers, and finally abuse everyone to maximize shareholder value. Doctorow argues this isn't inevitable capitalism but the result of weakened regulatory discipline, particularly around antitrust enforcement and tech worker power. (19:57) The discussion delves deep into legal frameworks like DMCA, fair use, and intermediary liability laws, examining how they enabled tech monopolies while also exploring the intersection with AI development and potential solutions through interoperability and labor organizing.
• Core theme: How the removal of competitive, regulatory, and worker-driven discipline allowed tech platforms to systematically degrade user experience while extracting maximum valueSpeakers
Sarah Jeong
Features editor at The Verge, standing in for regular Decoder host Nilay Patel. Jeong has extensive experience in tech journalism and has worked closely with organizations like the Electronic Frontier Foundation on internet policy issues.
Cory Doctorow
Prolific author, internet activist, and one of today's fiercest tech critics. Former European director at the Electronic Frontier Foundation, working across 31 countries on internet policy. Author of the new book "Enshittification: Why Everything Suddenly Got Worse and What to Do About It," among many other works.
Key Takeaways
Platform Lock-in Creates Abuse Opportunities
Doctorow explains how Facebook exemplifies the enshittification process by first attracting users away from MySpace with promises of no surveillance, then using the "collective action problem" to keep users trapped. (12:31) As he notes, "if the six people in your group chat can't agree on what movie you're gonna see this Friday, there's there's no way you're gonna agree on when it's time to leave Facebook." This psychological lock-in allows platforms to gradually degrade service while users remain because leaving means losing connections with people they care about more than they hate the platform.
Tech Lost Four Critical Sources of Discipline
Unlike traditional businesses that only face market and regulatory discipline, tech companies historically had four constraints: competition, regulation, interoperability, and scarce tech workers. (19:57) The erosion of antitrust enforcement allowed monopolistic mergers, while IP laws like DMCA Section 1201 made it illegal to modify software to serve users better. Most critically, tech worker scarcity disappeared after massive layoffs, removing the final check on corporate behavior. When workers could "walk across the street and get a better job," they refused to build harmful features - but that leverage is gone.
Interoperability as a Competitive Force
Doctorow argues that the universal nature of computers - their ability to run any program - historically served as a check on corporate power. (20:21) If a company created a "10 foot pile of shit," programmers could build an "11 foot code ladder to go over top of it." However, anti-circumvention laws now make it a felony to install ad blockers in apps, even though the same functionality is legal for websites. This legal framework allows companies to abuse users without fear of technical workarounds.
AI Bubble Economics Don't Add Up
Despite the AI hype, Doctorow presents stark numbers: seven companies representing 30% of the S&P 500 have spent $700 billion in capital expenditure while generating only $45 billion in gross revenue annually. (54:53) The business model relies on displacing high-wage jobs, but the total wage bill of illustrators or other creative workers "rounds to zero" against AI training budgets. For AI to be economically viable, it needs to replace high-wage sectors like software engineering or radiology, but fundamental context window limitations make this unlikely.
Antitrust Enforcement is Experiencing Global Revival
Against all political science predictions that billionaire preferences determine policy outcomes, countries worldwide are pursuing aggressive antitrust action. (71:15) Canada's Competition Bureau, which had pursued only three merger cases in its history and never won, received expansive new powers in 2024. The UK, EU, Japan, South Korea, and others are taking action against the same tech companies using similar legal frameworks established by the post-WWII Marshall Plan, creating opportunities for coordinated international enforcement.
Statistics & Facts
- Seven companies now represent 30% of the S&P 500, with $700 billion in capital expenditure but only $45 billion in gross revenue annually, including questionable accounting practices like Microsoft's circular Azure credit arrangements with OpenAI. (54:53)
- The National Bureau of Economic Research estimated that the average tech worker was adding $1 million per year to their boss's bottom line during the era of tech worker scarcity, explaining the lavish perks and high compensation. (22:24)
- Amazon engaged in predatory pricing by selling goods for newborns below cost to the tune of $200 million in just a couple months to destroy diapers.com, then acquired the company for pennies on the dollar. (36:28)
