Paul Graham wrote something in 2024 called "Founder Mode." I read it and thought: "This is actually a sophisticated justification for flat organizational structures."
And if he's right, something changes.
What Graham actually argues
Paul Graham, founder of Y Combinator, wrote this on his blog. The core idea:
Most management theory is built on "hire good people and give them autonomy." It sounds nice. It sounds modern. It sounds humane.
But it doesn't work. Not at scale. Not in fast-moving companies.
Graham's argument: Successful founders stay deeply involved in the work. Not because they're micromanagers. Because they need to be.
The founder's job isn't to delegate. It's to ensure the right decisions are made. And that requires being involved.
This is the opposite of what business schools teach. The myth is: "Good leaders delegate and trust their teams."
Graham's reality: "The best founders are the most involved."
Why this matters now
In the AI era, this becomes even more true. Why? Because the decisions are moving faster than ever before.
A manager who "checks in quarterly" gets blindsided. A founder who's hands-on catches changes in real time.
The cost of a wrong decision in an AI-driven company is massive. Retraining workflows. Losing users. Losing momentum.
So Graham's argument isn't just about leadership philosophy. It's about survival.
The hidden assumption
Here's what Graham doesn't say explicitly but implies:
This only works if the founder is actually good at making decisions.
Founder Mode doesn't work for founder mode's sake. It works when the person at the center is sharp enough to make the right call fast.
That's rare. Really rare.
What this means for organizations
If Graham is right, the future isn't flat hierarchies with autonomy. It's small teams with sharp leaders who are deeply involved.
That's the opposite of what we've been building for the last decade.
Most companies tried to scale by creating middle management layers. Layers that "empower" teams to make decisions without escalation.
Graham's world? Those layers become liabilities.
Questions that linger
If Founder Mode requires the founder to be constantly involved, what happens when the company grows beyond what one person can actually think about?
Does the company stay small? Does it hit a ceiling? Or do you eventually need to hire people good enough to make decisions without the founder?
And if you do hire people that good, haven't you already moved past Founder Mode?
(Graham doesn't answer this. But I suspect the answer is: most companies fail before they have to.)
Want more on managing in the AI era?
The Little Book for New Managers has helped 30,000+ managers figure out what actually works.
Get the book ($19) →