You feel it every time an app updates and suddenly the feature you loved costs $14.99 a month. You are being squeezed. But to the architects of Silicon Valley, your frustration isn’t a bug—it’s the exact point of the exercise.
Paul Graham, the godfather of the startup machine, recently published a manifesto on ‘Making Startups Powerful.’ The premise is seductively simple: startups gain power by doing the hardest, most indispensable work. You don’t just build a tool; you build the layer that everything else depends on. You eat your way through the stack until you are the stack.
Power without a stated purpose doesn’t just become extraction; it becomes a religion where the user is the sacrifice.
On paper, this sounds like pure engineering brilliance. It taps into the intoxicating founder belief that you can shape reality. But the twist hits when you realize you aren’t the founder shaping reality—you are the one being shaped by their ambition.
One seasoned operator reading Graham’s essay noted a terrifying variant of ‘going full stack’: eating your way gradually through the customer by doing all their hardest work for them. They mentioned a software vendor that provisions front-office applications for banks. The conversation isn’t about serving the bank anymore; it’s about the software vendor evolving to become the bank.
When you do all the hard work for a customer until they are entirely dependent on you, you haven’t built a partnership. You’ve built a hostage situation.
Graham treats power as a neutral engineering problem. How do we optimize for maximum leverage? But the public reaction reveals the glaring flaw in this amoral ambition. The same do-hard-things discipline that builds indispensable companies also rationalizes dominance and dark patterns.
Look at the music industry. Graham points to record labels as the boogeyman, painting startups as the liberators. But as one commenter brutally pointed out, artists were far better off with EMI and Columbia Records than they are under the streaming monopolies of today. The startup didn’t liberate the artist; it just built a more efficient extraction machine. It optimized the squeeze.
The missing variable in this entire playbook is accountability. ‘Powerful’ is absolutely meaningless without answering the question: powerful for whom? The strongest critique of Silicon Valley today isn’t about the methods used to build these giants. It’s the complete absence of a theory of limits.
You cannot solve the alignment problem in AI while simultaneously publishing a human alignment problem that celebrates maximum leverage without limits.
Founders and operators will read this playbook and copy it religiously. They will optimize for indispensability, rationalizing their maximum greed in thoughtful, measured prose that makes them seem like kindly philosophers rather than digital robber barons. Meanwhile, users and policymakers will have to live with the consequences.
Understanding this mechanism isn’t just business analysis anymore; it’s survival. We are watching the rationalization of digital feudalism in real-time. If we don’t start asking ‘Why’ instead of just ‘How,’ we will wake up living entirely inside someone else’s monopoly. And we will be paying rent on the very reality we used to own.
FAQ
Q: Isn't doing the hard work for customers just good business?
A: It crosses the line from service to monopoly when the customer becomes entirely dependent on you. Good business creates mutual value; digital feudalism creates hostages who have no alternative but to pay your toll.
Q: What's the practical implication for founders?
A: If you optimize purely for maximum leverage without a theory of limits, you will eventually default to dark patterns. You might build a massive company, but you'll build it on extraction rather than loyalty, inviting eventual regulatory or user revolt.
Q: Is Paul Graham actively advocating for bad behavior?
A: Not explicitly, but that's the danger. By framing amoral ambition as a neutral engineering challenge, it gives founders intellectual cover to rationalize maximum greed as a logical progression rather than a moral choice.