The Cycle of Reason

“I’m not scared of a smart man riddled with doubt; I’m scared of idiots figuring things out.” - Madae Uppedo (2026)

We will be talking about social psychology: the development and cycle of mental states that repeat indefinitely across industries.

The Machine

Here’s a claim that should bother you: nobody ever ruins an industry on purpose.

The people who hollow out consulting, or journalism, or venture capital, or your favorite subfield of software, are not villains twirling a mustache over a spreadsheet. They are, almost without exception, competent. They are good at their jobs. The problem is that somewhere along the way, “their job” quietly stopped meaning what it used to mean, and nobody sent a memo.

This is the part that’s hard to accept, so sit with it: decay is not a character flaw. It’s a structural inevitability, and it happens on a schedule.

There’s a four-beat rhythm underneath almost every institution you’ve ever watched go stale, and once you see it you can’t stop seeing it.

Beat one: Reason. A small group of people, usually broke, usually outsiders, solve a real problem using nothing but judgment. There’s no process yet, because process is a luxury good and they can’t afford it. What they have instead is taste — the accumulated, half-conscious sense of what’s actually true versus what merely sounds true. This is the fox phase: cunning, adaptive, uncredentialed.

Beat two: Power. The judgment works. It scales. And the moment it scales, it has to be turned into something teachable, because you cannot clone a founder’s intuition, but you can write it down as a checklist. Power enters here — not as corruption, but as infrastructure. Rules, titles, a hierarchy. This is the lion phase: force replacing cunning, structure replacing feel.

Beat three: Institution. The rules now need people to run them, and running rules is a different skill from making the original judgment call. A new caste rises — not worse people, just people optimized for a different game. They are excellent at navigating the org chart, terrible at the thing the org chart was built to protect. Nobody notices the swap, because the paperwork still says “reason” at the top.

Beat four: Greed. By the third or fourth generation of this — and it is almost always three or four, a suspiciously consistent number across totally unrelated domains — the people running the institution have no living memory of the scarcity that produced the original judgment. They inherited the throne room, not the war. Their incentive is simple: protect the position, not the purpose. This is where an industry starts producing decks instead of decisions, thought leadership instead of thought, and — eventually — grifters, because grifters are just the final, most efficient form of an institution that’s stopped checking its own homework.

And then, somewhere on the outside, broke and unfashionable, beat one starts again.

This isn’t a metaphor. It’s been independently discovered at least three times, by people who never read each other, describing wildly different centuries.

Three People Who Already Wrote Your Article, Sort Of

Vilfredo Pareto, an Italian economist with a taste for blunt aphorisms, spent the early 1900s arguing that history is “a graveyard of aristocracies.” His mechanism: power alternates between foxes (cunning, adaptive, comfortable with ambiguity) and lions (forceful, rigid, comfortable with hierarchy), and no elite — however brilliant at founding — survives contact with its own success. Every governing class eventually gets replaced, either by slow absorption or by outright revolt, because competence at getting power and competence at keeping power are almost never the same skill.

Six hundred years earlier, a North African scholar named Ibn Khaldun watched dynasty after dynasty rise out of the desert, hold power for three or four generations, and collapse — and noticed the pattern was so regular he could practically set a clock by it. His term for the glue that held the founding generation together was asabiyyah, usually translated as “group solidarity,” though “the thing that makes hardship worth it” gets closer to what he meant. His genuinely uncomfortable insight — the one that should keep you up at night if you’re building anything — is that decline isn’t moral. Nobody has to get lazy or corrupt for a dynasty to fall. The conditions that produced the discipline in the first place simply stop existing once the discipline succeeds. You cannot manufacture scarcity to keep your people sharp. Success is the disease that kills the thing that caused it.

And then, in 1911, a German-Italian sociologist named Robert Michels looked at socialist parties — organizations founded explicitly to prevent elite capture, run by people who had read Marx and meant it — and found them oligarchic anyway. His conclusion has since been quoted into a bumper sticker: who says organization, says oligarchy. Not because socialists are hypocrites, but because the moment any group gets big enough to need a hierarchy, the hierarchy needs people who are good at running hierarchies, and those people will — reliably, structurally, without any of them individually intending it — end up serving the hierarchy’s survival over the hierarchy’s original purpose. This is your epigraph, formalized. The idiots aren’t idiots. They’re excellent administrators who happen to have optimized for the wrong target, and nobody told them, because the org chart doesn’t have a column for “still cares about the original problem.”

Exhibit A: A Company That Used to Own Your Pocket

In 2007, Nokia controlled roughly half the smartphone market on Earth. By 2013, its phone business was gone, sold to Microsoft for parts. Six years, ninety percent of market value, erased. The postmortems love to blame Symbian, the aging operating system nobody could fix in time — but the operating system wasn’t the disease, it was a symptom.

Researchers at Aalto and INSEAD spent years afterward interviewing seventy-six Nokia executives, managers, and engineers to find out what actually happened inside the building. What they found wasn’t stupidity. It was fear, running in both directions at once. Top management was terrified of missing quarterly numbers and losing investor confidence, so they leaned on middle management to hit impossible targets. Middle management was terrified of being blamed for the shortfall, so they quietly polished the truth on the way up — moved the data points on the chart, in one manager’s own words, “to give a better impression.” Engineers who knew Symbian couldn’t compete watched their honest assessments get softened at every layer between them and the boardroom, until what reached the top was optimism with the inconvenient parts filed off.

Nobody lied because they were bad people. Everybody in that chain was protecting something real — a job, a target, a team, a reputation for delivering good news. That’s beat three and beat four collapsing into each other: an institution so committed to protecting itself that it becomes structurally incapable of hearing the thing that would save it. The reason that built Nokia’s early dominance — engineers with actual taste for what a phone should do — was still in the building in 2007. It just couldn’t get a message past its own management layers anymore.

Exhibit B: Why Good Management Is the Trap, Not the Fix

Clayton Christensen spent a career studying exactly this pattern and arrived at a conclusion that should unsettle anyone who works in tech: the very practices that make a company excellent are the ones guaranteed to blind it. Listen closely to your best customers. Chase the highest margins. Invest where the return is provable. Every one of those is textbook good management, and every one of those systematically starves the unproven, low-margin, weird-looking thing that will eventually eat you — because your best customers don’t want it yet, and your margin targets can’t justify it yet, and by the time either of those things changes, someone smaller and hungrier has already built it. Christensen watched titans like DEC and Sears do everything right on their way to irrelevance. This is Khaldun’s asabiyyah in a business-school suit: the very discipline that wins the market is the discipline that can’t survive winning it.

The Loop Is Running Again, Right Now, in Your Field

Here’s the part that should feel less like history and more like a weather report. The current AI landscape is a live beat-one-versus-beat-three fight in real time: heavily funded labs with enormous compute and enormous process, versus smaller open-weight teams with a fraction of the resources and none of the institutional caution, moving faster on narrower problems precisely because nobody’s asked them to protect a quarterly number yet. Nobody knows yet who’s the fox and who’s the lion in this round. That’s usually only obvious in hindsight — which is exactly why it’s worth paying attention to now, while it’s still ambiguous, instead of five years from now when someone writes the Nokia-style postmortem.

The tags on this article aren’t decoration. This is, not incidentally, the whole argument for why unglamorous infrastructure work matters: an attempt to build the next fox phase somewhere durable enough to survive its own success — which, historically, is the one trick almost nobody manages to pull off.

The Uncomfortable Question

Here’s the part that should actually bother you, more than the historical pattern-matching: if decline is structural rather than moral, virtue doesn’t save you. You cannot hire your way out of this with better people. Michels tested that hypothesis on the most idealistic organizations of his era and watched them fail anyway. Good intentions are not a control mechanism. They’re barely a speed bump.

What seems to matter — in the rare cases where an institution slows its own decay — isn’t the moral quality of its people but the mechanical quality of its feedback loops: whether the people at the top can still lose, whether the founding scarcity gets deliberately re-imported instead of engineered away, whether “administrator” and “practitioner” stay the same job for as long as possible before they’re allowed to split.

That’s a smaller, less romantic thesis than “good people beat bad systems.” It’s also, probably, the only thesis with a track record.

So: not scared of a smart man riddled with doubt. Scared of the idiots who figured it out — where “it” is not the problem, but the org chart. They’re not stupid. That’s the whole problem. They’re the most competent people in the building, at exactly the wrong thing.

The cycle doesn’t break because someone was virtuous. It breaks — if it breaks at all — because someone, somewhere, still broke, still outside, still hungry, is doing beat one again, right now, in a field nobody respectable is watching yet.

Go find them. Or be them.