Businesses become critically dependent on platforms for even a small fraction of their revenue (e.g., 20%). This 'monopsony power' creates a stronger lock-in than user network effects, as losing that customer base can bankrupt the business.
Platforms grew dominant by acquiring competitors, a direct result of failed antitrust enforcement. Cory Doctorow argues debates over intermediary liability (e.g., Section 230) are a distraction from the core issue: a decades-long drawdown of anti-monopoly law.
The massive investment in AI isn't justified by displacing illustrators, whose total wages are negligible. The economic model is predicated on replacing high-cost professions like radiologists or software engineers, which is a far more challenging task.
Content moderation laws are difficult and slow to administer. A better solution is requiring platforms to provide users with a simple file of their data and social graph, allowing them to switch services easily and creating real competitive pressure.
Laws intended for copyright, like the DMCA's anti-circumvention clause, are weaponized by platforms. They make it a felony to create software that modifies an app's behavior (e.g., an ad-blocker), preventing competition and user choice.
Previously, scarce and mission-driven tech workers could refuse to build features that harmed users. Mass layoffs created a labor surplus, removing workers' leverage and allowing companies to push through user-hostile changes without internal resistance.
Platforms first attract users with good service, then lock them in. Next, they worsen the user experience to benefit business customers. Finally, they squeeze business customers, extracting all value for shareholders, leaving behind a dysfunctional service.
AI struggles with tasks requiring long and wide context, like software engineering. Because adding a linear amount of context requires an exponential increase in compute power, it cannot effectively manage the complex interdependencies of large projects.