AI's Report Card Is In, Opus 4.7 Has Mixed Reviews, and What the First Emperor Can Teach You About Standardization

A Tale of Two Verdicts

In the West, Qin Shi Huang — the First Emperor of Qin, best known to most Western readers as the emperor buried with the 8,000-strong Terracotta Army discovered near Xi'an in 1974 — is often framed as one of history's great unifiers. He ended over 500 years of civil war, conquered the rival states of Han, Zhao, Wei, Chu, Yan, and Qi, and in 221 BCE established the first unified Chinese empire. The very word China derives from Qin.

And while Alexander the Great and Genghis Khan conquered larger territories, their empires fragmented almost the moment they died — Alexander's was divided among his generals within a decade, and the Mongol Empire began splintering into rival khanates after Genghis Khan's death. Qin's framework, by contrast, endured. The administrative structure he built "served as a model for government in China to the present day," and historians credit the Qin "as the originators of the Chinese imperial system that would endure in some form until the Xinhai Revolution in 1911." That's more than 2,000 years of institutional continuity from a dynasty that itself lasted only fifteen.

The other verdict is far less flattering. In this version, Qin Shi Huang is a tyrant who buried scholars alive, burned the books of the Hundred Schools of Thought, and worked hundreds of thousands of peasants to death building the Great Wall and his own tomb.

But here's the interesting twist — and it matters for anyone thinking about leadership. History, as the saying goes, is written by the victors. The tyrant portrait comes almost entirely from one source: Sima Qian's Records of the Grand Historian, written under the Han dynasty — the very dynasty that overthrew Qin and needed to justify its own rule. Modern historians now openly doubt the story. Sima Qian "had every reason to denigrate the earlier emperor to flatter his own." The 2015 discovery of 200,000 Qin bamboo slips in Hubei has further complicated the received picture, revealing a bureaucracy with fines for minor late arrivals (not death penalties), waived punishments for weather delays, and paid wages — even for convict laborers on the Great Wall.

History, in other words, is sometimes a mask decorated by whoever inherits the throne.

The Chaos Before the Standard

To appreciate what Qin Shi Huang actually did, picture what he inherited. Seven warring states, each with its own written script, its own currency, its own weights and measures, even its own cart-axle widths. Roads were rutted tracks where a cart from one kingdom couldn't fit the grooves left by carts from another. A merchant crossing three states needed to relearn the weight system three times. A scholar's book might be unreadable 200 miles from where it was written.

This wasn't a minor inconvenience. It was a structural drag on everything — trade, taxation, troop movement, even basic communication.

Unify Everything. Unify It Now.

Qin Shi Huang and his chancellor Li Si did not wait. Each time a state fell, they moved immediately to dissolve its differences.

They standardized the written script, consolidating regional variants into small seal script so that parts of the empire that didn't speak the same language could still communicate in writing. They standardized currency, replacing knife-money, spade-money, and regional coins with a single round bronze coin. They standardized weights and measures, casting bronze and iron reference models stamped with an imperial edict and distributing them across the provinces. They standardized cart axles, so that a wagon from Xianyang could roll cleanly into a rut left by a wagon from the old territory of Qi. They standardized law, replacing regional customs with a uniform legal code. And they replaced the whole feudal system with 36 centrally administered commanderies run by appointed officials, not hereditary nobles.

Here is the part that deserves real attention: this was against common sense at the time. The default assumption of Chinese statecraft had been the Zhou fengjian system — kings distributed territory to relatives and loyal nobles to maintain harmony across distances that seemed too vast to govern directly. Li Si pushed back. His response cut to the bone: the Zhou kings had distributed land to their sons and brothers in the name of harmony, but within a few generations those very relatives were waging war against one another. Weren't the warring states themselves, he argued, ultimately descended from the same Zhou ancestors? Giving territory away was not the path to peace — it was the seed of the next civil war.

Qin Shi Huang sided with Li Si. He abolished hereditary fiefdoms entirely and imposed the commandery-county system on the whole empire. Against what everyone "knew" about governance, he bet on centralization.

The Cost of Getting It Right

This was brutally hard work. According to traditional accounts, Qin Shi Huang personally reviewed 120 jin — roughly 60 kilograms, or about 130 pounds — of bamboo-strip documents every single day. He worked like a man who knew that the moment he let up, the old fractures would reopen. And in fact, that's exactly what happened. His dynasty collapsed within four years of his death in 210 BCE, partly because the machinery of centralization had been held together by his personal will.

But here is the astonishing thing: the machinery itself didn't fail. The Han dynasty that overthrew the Qin did not return to the old fengjian patchwork. They "largely continued the military and administrative practices instituted by the Qin," kept the script, kept the coinage system, kept the commanderies, kept the standardized weights. Every subsequent dynasty — Tang, Song, Ming, Qing — built on the Qin frame. The political structure survived 2,000 years of dynastic turnover.

Contrast that with Alexander and Genghis Khan. Both conquered more territory in a single lifetime than Qin Shi Huang ever did. Neither left behind an institutional frame that could survive their own funeral. Conquest without standardization is just a large tent in a windstorm.

The Business Lesson (and the Careful Qualifier)

Now transpose this to a company — especially one built through acquisition, or a global firm with a hundred business units that each grew up with their own ERP system, their own procurement contracts, their own HR policies, their own data warehouses, their own way of closing the books. Everyone has a reason their way is better. No one wants to change. Every standardization project hits ferocious resistance.

And the person leading it looks, in the short term, exactly like Qin Shi Huang did to his critics — disruptive, overreaching, insensitive to local culture.

But done well, the payoff compounds the same way. A common back end — shared platforms, shared financial systems, shared data, shared brand standards, shared culture — unlocks economies of scale no collection of independent fiefdoms can match. It makes talent portable across the organization. It makes capital allocation legible. It makes the next acquisition easier to integrate. It outlasts the CEO who built it.

This is not an argument to centralize everything. Qin Shi Huang standardized the script and the cart axle. He did not tell the people of the old state of Chu what to cook for dinner. The art is knowing which layer to unify and which layer to leave alone — and the same discipline applies to a company.

Two dimensions matter: size and function.

Size first. When Henry Singleton ran Teledyne with "extreme decentralization," he was running a 40,000-person company, and most of its business units were still sizable — thousands of employees each, with real scale of their own. That context matters. Decentralization works because each unit is large enough to absorb its own overhead, attract real talent, and operate as a genuine business. A 1,000-person company chopping itself into ten 100-person fiefdoms usually doesn't get nimbleness — it gets ten undersized teams, none with enough leverage to compete. Below a certain size, further decentralization is just fragmentation. The Qin parallel is exact — seven small kingdoms, each with its own script and currency, were not seven nimble competitors. They were seven underpowered duplications of the same overhead.

Function second. Different functions want to live in different places on the spectrum. Back-office functions — IT, finance systems, HR platforms, procurement, legal, data infrastructure — generally belong centralized. They are where economies of scale are most real, where inconsistency creates pure drag, and where the customer never sees the seam. Client-facing functions are the opposite. Sales, customer service, and in some businesses innovation itself (3M is the classic case, with R&D pushed deep into operating divisions close to customer problems) need to be decentralized and empowered, because customer needs genuinely vary — by geography, by age group, by local competitive dynamics — and a head-office pattern imposed from a thousand miles away will miss what a local team sees every day.

This framework is what William Thorndike's The Outsiders — the book Warren Buffett placed at the top of his recommended reading list in his 2012 Berkshire Hathaway shareholder letter — captures in its central lesson. Thorndike profiles eight CEOs whose companies dramatically outperformed the S&P 500 over their tenures: Henry Singleton at Teledyne, Tom Murphy at Capital Cities, John Malone at TCI, Katharine Graham at The Washington Post, Bill Anders at General Dynamics, Bill Stiritz at Ralston Purina, Dick Smith at General Cinema, and Warren Buffett himself. The pattern is remarkably consistent: decentralize operations, centralize capital allocation. Capital Cities put it plainly inside every annual report: "Decentralization is the cornerstone of our philosophy." But Tom Murphy and Dan Burke held the lever that mattered — every major capital decision, every acquisition, every buyback — at the top. Singleton ran Teledyne with fewer than 50 people at headquarters, but he personally directed the capital allocation that let the company repurchase over 90% of its own shares when it was undervalued.

The practical translation is straightforward. Centralize the infrastructure — platforms, capital allocation, brand, culture, core data, back-office systems — where scale, consistency, and common language genuinely compound. Decentralize the work where responsiveness, local knowledge, and ownership genuinely compound. The merit of centralization is that it builds a foundation that outlasts any individual leader; the merit of decentralization is that it keeps the organization close to the ground where real customers and real markets live. Neither is universally right. But most companies err heavily on one side: they over-decentralize the back end (letting every acquired unit keep its own ERP, its own HR system, its own data taxonomy) and over-centralize the front end (dictating customer scripts from headquarters). The Outsiders got the directionality right, and compounded the difference for decades.

The Uncomfortable Truth About Lasting Impact

None of this is easy. Qin Shi Huang pushed his reforms so hard, and worked himself so hard, that the system outran its human sustainability — which is a real warning, not a small one. The modern business equivalent is a transformation leader who burns out the organization and themselves pushing integrations that lose the local knowledge that made the acquired companies valuable in the first place.

But the deeper truth holds. Go against the grain of what everyone around you believes is possible. Absorb the short-term pain of resistance. Build the common foundation that no one else is willing to build. And two decades later, when your competitors are still translating between their own internal dialects, you'll be operating on rails that let you move at a speed they can't match.

Alexander and Genghis Khan were conquerors. Qin Shi Huang was a builder. Only one of those three left something that lasted 2,000 years — and he did it by doing the unglamorous, resisted, back-breaking work of making a hundred different things into one.

Sources: Sima Qian, Records of the Grand Historian (Shiji) | Smithsonian National Museum of Asian Art, "Qin dynasty (221–206 BCE)" | Yunmeng/Hubei Qin bamboo slips (2015) | William Thorndike, The Outsiders (Harvard Business Review Press, 2012) | Warren Buffett, Berkshire Hathaway 2012 Annual Shareholder Letter | McKinsey, "To centralize or not to centralize: generative AI in banking" (June 2024) | BCG AI Radar 2025 & 2026 | AWS Enterprise Strategy, "Centralizing or Decentralizing Generative AI? The Answer: Both" (September 2024)