Forget Complexity. Entropy is the Problem.
Why your organization feels broken — and why it's about to get worse
Every executive I work with describes the same sensation. They can’t quite name it, so they reach for “complexity.” The world has become too complex. The market is too volatile. The org is too matrixed, everything is VUCA.
I don’t think that’s what’s happening.
What they’re experiencing isn’t complexity. It’s entropy — the dispersion of organizational power away from the people who are supposed to wield it. They feel disoriented not because the world changed too fast, but because the internal physics of their company changed beneath their feet, and nobody told them.
I’ve written about the Change Tax — the hidden cost that teams pay when structural misalignment forces them to bridge gaps that shouldn’t exist. Readers from banking, healthcare, venture capital, legal tech, and media wrote to tell me they finally had words for what they’d been feeling.
This essay is about where that misalignment comes from.
Three things that used to be one
Power has three elements in my experience: Authority (who decides), Assets (who holds the means of execution), and Access to Information (who knows what’s actually happening).
For most of the twentieth century, these three were fused, so we never talked about them separately. Geography did the fusing.
If you ran a newspaper in 1986, your authority was literal: you sat in the corner office on the same floor as the ad sales team. Your print operation was probably close, too. The assets — the press, the distribution trucks, the subscriber list — were physical objects in buildings you controlled. And information flowed through you because the people who had it walked past your door.
Andy Grove understood this. His High Output Management, published in 1983, is probably the best operations manual ever written — and it assumes that the manager can see the production line. The breakfast factory, his famous metaphor, works because the eggs, the toast, the coffee, and the cook are all in the same room. Authority, assets, and information are co-located. The manager’s job is to optimize the cooking process.
That book deserves a renaissance right now — not because its prescriptions still apply unchanged, but because understanding what it assumed reveals exactly what has been lost. Grove himself saw the limits. In the same book, he wrote: “When the environment changes more rapidly than one can change rules, or when a set of circumstances is so ambiguous and unclear that a contract between the parties that attempted to cover all possibilities would be prohibitively complicated, we need another mode of control.”
He was describing the moment when the org chart stops working. We are living in that moment.
What happened
The digital transition didn’t just change business models. It decoupled the three elements of organizational power.
Authority fragmented. Decision rights that used to be clear became ambiguous as matrix structures, cross-functional teams, and “empowerment” initiatives distributed nominal authority across dozens of people — without establishing who actually gets to say no. Have you ever found yourself in endless, soul-sucking alignment meetings? Probably yes. Did they warn you about these in business school? Probably not. Because something has changed in the meantime.
Assets dispersed. In knowledge-intensive businesses, the critical means of execution — code, data, customer relationships — moved away from the people held formally accountable for outcomes. The VP of Product may own a P&L, but the tech team controls the deployment queue. The Chief Revenue Officer owns the revenue target, but an algorithm decides what customers see.
Access to Information inverted. In the physical era, information flowed upward because the hierarchy was also the information architecture. The person at the top knew the most because reporting structures ensured it. In the digital era, the person closest to the customer, the code, or the data often knows far more than the person making the decision. The information advantage now sits at the edges, not the center.
The German sociologist Niklas Luhmann studied exactly this phenomenon at the societal level. His theory of functional differentiation describes how modern systems evolve by splitting into autonomous subsystems — each operating according to its own logic, no longer coordinated by a single hierarchy. The economy follows the logic of payment; the legal system follows the logic of legality; science follows the logic of truth. They coexist, but they don’t naturally align.
What Luhmann described in societies is now happening inside companies. Authority, Assets, and Access to Information have differentiated into subsystems that each follow their own logic. The org chart says one thing. The power wiring says another. And in between, talented people exhaust themselves trying to reconcile the two.
Cybernetics offers a complementary lens. Ashby’s Law of Requisite Variety — the first law of cybernetics — states that a control system must be at least as complex as the system it’s trying to regulate. An org chart designed for an era when authority, assets, and information were fused simply does not have the requisite variety to regulate an organization where they’ve decoupled. The control mechanism is too simple for the system it governs. The result, as a commenter on one of my LinkedIn posts put it precisely, is that “the level of efforts ‘managing’ overtake actual efforts of performance.” The organization spends more energy regulating itself than doing the work.
That effort — the meetings, the escalations, the back-channels, the alignment rituals — is the Change Tax.
Why media is the sharpest case
I keep coming back to media because it’s the cleanest example of this decoupling — and because it happened there first. OK, also because I’ve worked in media for almost a decade, and because I care about it.
But the media industry was the canary in the coal mine for digital disruption. A highly profitable, structurally simple business model was suddenly and completely turned on its head. Print advertising revenue didn’t decline gradually; it collapsed. Distribution went from a physical monopoly to an open commodity. The reader relationship, once mediated by nothing but the mailbox or a walk to the next newsstand, became mediated by platforms that publishers didn’t control.
And yet, most media companies responded as if the problem was operational. They restructured. They hired digital talent. They launched transformation programs. They held offsite after offsite to “align.”
What they didn’t do is re-examine how authority, assets, and information were wired. The editor-in-chief still sat atop an editorial hierarchy designed for print. The commercial teams still operated in two different worlds, not paying attention to the fact that data fuels both advertising and subscription revenue. The technology team — now the most asset-rich function in the building — reported three levels below the CEO. Meanwhile, the people with the deepest understanding of audience behavior (data analysts, product managers, growth specialists) had virtually no decision rights.
The talent was there. The strategy was often reasonable. What was missing was structural coherence: the right authority held by the people who controlled the right assets, informed by the right information.
I’ve watched this play out from the inside and as an advisor. The pattern is always the same. Competent leaders fail. Boards replace them. The new leaders fail, too. Everyone blames execution or culture. Nobody examines the wiring.
This is not a media problem
As I said, I have been hearing from leaders in different industries since I started this Substack, not just publishing. The specifics of what resonated varied. In some cases it was accountability for outcomes without control over the assets required to deliver them. In others, it was information trapped at the edges of the organization while decisions were made at the center. In one case, a leader had raised the structure problem with their CEO repeatedly — and watched it go unaddressed, not because the CEO disagreed, but because the organization had no vocabulary for a problem that wasn’t “culture” or “alignment.” The same problem, different industries.
Most organizations don’t do the structural work of recoupling authority, assets and information. They do alignment work instead — which is to say, they ask people to compensate for the absence of structure through personal effort, relationship management, and sheer willpower. That’s the tax.
Why AI makes this urgent
There’s a sentence from Alex Lieberman — co-founder of Morning Brew, now running Tenex, an AI transformation firm — that I keep thinking about. He describes AI as a “funhouse mirror”: it makes good performers dramatically better and mediocre ones worse, because it amplifies whatever is already there.
Extend that logic from individuals to organizations, and the implication is stark.
If your authority structure is coherent, AI accelerates decision-making. If it’s incoherent, AI accelerates the confusion — more options generated faster, routed to people who lack the mandate to choose between them.
If your assets are controlled by the people accountable for outcomes, AI multiplies their leverage. If assets are fragmented, AI produces more output that nobody has the structural authority to deploy.
If your information architecture connects the right knowledge to the right decisions, AI enriches it. If information flows are broken, AI generates more noise in a system already drowning in it.
This is the insight that too many AI transformation programs miss. You cannot automate your way out of structural incoherence. Adding AI to a decoupled organization doesn’t solve the decoupling. It accelerates it.
Grove could optimize the breakfast factory because the factory was structurally sound. The elements were coupled. AI is the most powerful optimization tool ever built — but optimization assumes a coherent system to optimize. When the system itself is decoupled, optimization just makes the entropy faster.
What this means
The Great Decoupling changes what you look for. Instead of asking “Why can’t my leaders align?”, you ask: Where is authority separated from the assets it needs? Where is information flowing to people who can’t act on it? Where are decisions made by people who don’t control the means of execution?
These are diagnosable questions. They have answers. And the answers point to structural interventions — not another offsite.
I’ve written about one such intervention: the Mission Operating Model, a dual hierarchy that separates organizational stability from mission-specific authority. Future essays will go deeper. But the diagnostic comes first.
You can’t fix a wiring problem if you think it’s a plumbing problem.


Great article. I'm gonna need some more time to digest this.
I love this sentence... "He describes AI as a “funhouse mirror”: it makes good performers dramatically better and mediocre ones worse, because it amplifies whatever is already there."