Everyone is wrong about what a CEO actually does.
It's not product. It's not marketing. It's not vision boards.
It used to be five things:
In the last nine months, AI has been catching up and actually got better than me at two of the five.
Capital allocation. Cash flow.
It watches every line of the P&L every day and flags what I used to catch on a Sunday night two weeks after the month was over.
And hiring and firing? The part I called 80% of the job is shrinking fast.
Every role I would have posted last year, I now ask a different question first: can an agent do this better, and if not, can a person plus an agent do it with one hire instead of three?
So the job compressed. Five things became three:
Aim
One goal. One hill, one direction, and the psychology to not chase every shiny object when everyone around you is chasing six. If you chase two rabbits you catch none. This is the part AI cannot do, because it doesn't want anything. That's the One Hill Doctrine in action.
Army
Humans and agents, designed together. The CEO's job is no longer filling seats. It's deciding what seat is even worth it. For example: PocketFM went from 0 to $500M in ARR (adding $250M last year alone) on the back of an AI Ad Engine. A creative director would only be looped in if the AI couldn't hit benchmarks.
Assets
Cash, systems, IP. The machine that runs without you. AI made this cheaper to build and more expensive to ignore. You've got a 300IQ CFO thinking about your business 24/7.
My CEOs are implementing these systems and growing at a scale never before seen. They're embracing what I call Architect Mode, where the leader designs the system instead of being it.
The old CEO managed a company. The new one architects it.
The CEO role has fundamentally changed. AI handles capital allocation and cash flow better than most humans. Hiring decisions now start with "can an agent do this?" What remains is Aim, Army, and Assets.
This isn't about working less. It's about working differently. The companies winning right now have CEOs who stopped managing and started architecting.
If you're still doing the job the old way, you're competing against people who aren't.
