I saw a LinkedIn post recently arguing that companies once only needed a CEO and CFO, but over time accumulated layers of executive roles — CIO, CMO, CHRO, CISO, CPO, and so on — because the volume of information became too large for a CEO to process alone. The conclusion was that AI now changes all of that. Since AI can aggregate and synthesize information better than humans, many of those executive roles supposedly become unnecessary again.

It is an interesting theory. And I think it fundamentally misunderstands both leadership and organizations. This entire line of thinking feels like another round of irrational exuberance around AI — confusing information management with accountability, judgment, and organizational leadership.

I actually watched a CEO implement a very sophisticated AI management tool inside a company. Every manager had to submit:

• what they worked on during the prior week,
• what they accomplished,
• what they planned to work on next week,
• and what their goals were.

The AI also had access to the company’s broader strategic objectives. It synthesized updates across roughly twenty managers, identified bottlenecks, highlighted conflicting priorities, and generated recommendations. Technically, it was very impressive. It created visibility quickly. It summarized patterns no executive could manually digest at the same speed. It surfaced coordination problems. It even appeared to “understand” the organization.

And so what? It still did not manage anything. The people did.

More importantly, the managers immediately learned how the AI system worked. They adapted to the tool itself:

• polished updates,
• carefully framed accomplishments,
• goals optimized for appearance,
• activity designed to look productive,
• language engineered to avoid scrutiny.

Classic garbage in, garbage out. The AI never truly had a chance in the supervisory role because management is not simply information synthesis.

Real leadership involves:

• understanding incentives,
• detecting what is intentionally omitted,
• identifying political behavior,
• challenging assumptions,
• resolving competing objectives,
• allocating scarce resources,
• and making judgment calls under uncertainty.

That does not disappear because a machine can summarize status reports faster.

The same misunderstanding exists in the argument about executive specialization. The CFO role did not emerge merely because accounting became “too complicated” for the CEO. Finance required independent stewardship of capital, liquidity, controls, reporting, taxation, risk management, and long-term investment decisions. The CIO emerged because technology became strategically inseparable from operations. The CISO emerged because cybersecurity became existential. The CHRO emerged because scaling organizations introduced cultural, legal, compensation, and talent complexity.

These are not just “information domains.” They are judgment domains with specialized accountability attached to them. And maybe that is the deeper misunderstanding underneath many AI leadership predictions. They assume organizations fail because leaders lack information. But large organizations usually fail because incentives drift, blind spots emerge, assumptions go unchallenged, political behavior goes unchecked, and difficult tradeoffs require competing perspectives.

That is precisely why mature organizations evolved specialized leadership structures in the first place. Not because the CEO was incapable. Because no single viewpoint is sufficient. Ironically, many AI advocates now argue that a CEO armed with sufficiently powerful AI can recentralize decision-making again. A sort of corporate “king” model. I am skeptical.

The more complex the system becomes, the more I want specialized leaders challenging one another’s assumptions — finance challenging operations, security challenging product, legal challenging sales, engineering challenging marketing. The friction is not a flaw in the system. It is often the system working correctly.

And then a funny thought occurred to me while reading all of these predictions about AI replacing leadership. Wait a minute. Don’t we already acknowledge that different AI tools themselves perform better in different areas?

• One model may be stronger in coding.
• Another in legal reasoning.
• Another in data analysis.
• Another in language generation.
• Another in image creation.

Even the AI ecosystem itself is evolving toward specialization because deeper expertise produces better answers in complex domains. In other words, even the machines are reinforcing the argument for specialized “brains.” So why would we assume organizations suddenly no longer require specialized human leadership and accountable decision-makers?

AI absolutely changes the tools. And it is changing them very fast. It may dramatically compress middle management, administrative coordination, reporting overhead, and certain forms of analysis. Entire layers built around information transmission may disappear. But leadership itself?

• Judgment?
• Accountability?
• Influence?
• Coordination across competing interests?
• The political realities of organizations?

Those things are not disappearing. If anything, faster-moving environments may increase the need for experienced leadership capable of integrating specialized perspectives into coherent decisions.

The military analogy may actually fit best.

• Weapons evolve rapidly.
• Communication evolves rapidly.
• Intelligence gathering evolves rapidly.

But wars are still won and lost through leadership, coordination, incentives, logistics, morale, specialization, judgment, and execution. The “troops and weapons” evolve. The leadership challenge largely does not.

The best organizations are usually not monarchies. They are systems of coordinated and accountable specialists aligned toward common objectives.

No kings.