How a Few Hundred Spaniards Leveraged Two Empires: Si Sheppard on the Aztec and Inca Conquests

1. Guest Background

This episode of the Dwarkesh Podcast is an interview: Dwarkesh Patel speaks with Si Sheppard in an episode titled “How did a few hundred Spanish soldiers topple two empires? – Si Sheppard.” The indexed evidence identifies Sheppard as a professor and military historian. His relevant role in this conversation is not generic commentary on the past, but military-historical analysis of Cortes, Pizarro, Spanish technology, diplomacy, and native alliances in the conquests of the Aztec and Inca empires.

The episode is built around an apparently impossible problem. Dwarkesh opens by emphasizing that a few hundred conquistadors arrived on a continent they did not understand and, within years, subdued empires with millions of inhabitants. He frames the military imbalance in extreme terms: Spanish forces often fought while outnumbered by ratios such as 100 to 1 or 1000 to 1, and much of their fighting strength eventually came from the very native peoples who would later be incorporated into the colonial order. That framing could invite a shallow answer, such as “Europe had guns.” Sheppard’s contribution is to slow the question down and reconstruct the mechanism.

The conversation therefore works as an episode analysis of conquest as a system. The speakers discuss who acted, what each side could perceive, where each side misjudged the other, and why advantages that were not individually overwhelming became decisive in combination. Dwarkesh also gives the episode a modern edge by noting that an outside force with unfamiliar motives, diplomatic cunning, and technological advantages can reorder a larger established system. At the end, he says the analogy to AI is obvious but does not spell it out. This article keeps that boundary: it does not invent an AI theory for the episode, but it does preserve the larger analytical lesson about scale, dependence, and strategic misreadings.

2. What the Episode Covers

Sheppard begins the story in fifteenth-century Europe. In his telling, the Spanish conquests are part of the birth of the modern globalized world. Europe was pressed against the Atlantic and cut off from the eastern centers of spice and silk wealth, which created pressure to innovate at sea. Portugal accumulated knowledge of winds, tides, and coasts while moving around Africa toward India; Spain went west. That background matters because Cortes and Pizarro were not merely isolated daredevils. They were attached to a widening Atlantic system of ships, coastal bases, finance, manpower, and cumulative expeditionary knowledge.

The Cortes story begins with mission drift. In 1518, the governor of Cuba assigned Cortes to explore the Mesoamerican coast. Cortes recruited men for a larger project and turned exploration into an inland campaign to find and seize power. According to Sheppard, Cortes’s political antennae were crucial. He quickly saw that the Aztec Empire was not identical with all of Mesoamerica. The Totonacs resented Aztec authority, and the Tlaxcalans, after initially fighting Cortes, came to view him as a possible way out of strategic encirclement by the Aztecs. They became guides, reinforcements, and political partners.

The episode then complicates the familiar story of Moctezuma II. Sheppard does not present him as simply naive. Moctezuma knew Cortes was coming; he had envoys, scouts, and spies, and he initially tried to make Cortes leave by sending precious gifts. But the gifts confirmed the presence of wealth and intensified Cortes’s desire to reach the center. Sheppard argues that allowing Cortes into Tenochtitlan may have been Moctezuma’s best available option, because an uncontrolled Cortes roaming the empire might have inflamed subject peoples against central authority. The catastrophic error was that Moctezuma himself remained physically available. Cortes used the killing of Spaniards on the coast as a pretext to place him under “protective” custody and then governed through him.

The Aztec campaign was not a smooth Spanish march to victory. When Cortes left the capital to deal with Panfilo de Narvaez, Alvarado misread an Aztec aristocratic and warrior festival as mobilization and massacred the participants. The survivors rallied the city, and Cortes returned to a siege. During La Noche Triste, many Spaniards were killed on the causeways or captured and sacrificed. At Otumba, the Aztecs had numbers and momentum, but Cortes identified the enemy commander, gathered enough cavalry, broke through, and destroyed command and control. The escape depended on Tlaxcalan choices: Sheppard stresses that if the Tlaxcalans had turned against Cortes then, the Spaniards would likely not have reached the coast alive.

Pizarro’s campaign in Peru is presented as an adaptation of the Cortes playbook. Pizarro met Cortes and learned the basic pattern: go to the top of the hierarchy, capture the sovereign, retain as much of the existing structure as possible, and keep a coastal base as a fallback. When he arrived in the Andes, the Inca Empire had just passed through civil war between Atahualpa and Huascar. Atahualpa had won and was moving toward Cuzco to be enthroned. At Cajamarca, Pizarro used a confined courtyard, the shock of artillery and arquebuses, horses, and surprise to capture Atahualpa. Dwarkesh gives the order of magnitude: roughly 160 to 190 conquistadors faced an Inca force at least in the high tens of thousands. The episode treats this not as magic, but as a decapitation operation against a centralized imperial system at a moment of political transition.

3. Core Views: Reasoning, Examples, and Limits

The episode’s central view is that “a few hundred men conquered millions” is the wrong unit of analysis. Sheppard does not deny Spanish battlefield advantages, but he resists the idea that guns explain the outcome. Early firearms had shock value, smoke, noise, and unfamiliarity, yet their physical killing power and reloading speed were limited. The more durable package was steel helmets, steel armor, long Toledo steel swords, inherited Old World military training, horses, and psychological disruption. Horses were not just transportation. They were a shock system. Europeans had institutional knowledge for resisting cavalry with dense formations, long pikes, discipline, and training. Mesoamerican and Andean warfare lacked the same anti-cavalry doctrine, technology, and formation density.

The episode’s examples make that reasoning vivid. When Cortes fought the Tlaxcalans, he had only a few hundred men against multiple thousands. Steel, horses, and training helped him survive, but he also hid casualties to preserve the aura of Spanish power. At Cajamarca, the Inca were veteran warriors who had just won a civil war, but they had never heard gunpowder weapons or seen cavalry used in this manner. Sheppard says the artillery and arquebuses caused relatively limited physical damage; the decisive effect was the explosion, the apparent command of thunder, and the release of horses in a confined space. The casualty figures mentioned by Dwarkesh are episode framing and speaker-cited quantities, not independently established universal measurements. The stronger claim is that unfamiliar technology, tactical surprise, crowd compression, and the ruler’s presence created political consequences far beyond the raw lethality of any single weapon.

A second core view is that Spanish conquest was not a story of Europeans acting alone. It was an insertion into native political fractures. The Aztec and Inca empires looked formidable on maps, but both were multiethnic imperial systems ruled by relatively narrow elites. Tlaxcalans did not simply misunderstand Cortes; they saw him as a tool against Aztec pressure and later demanded territory and a fortress in Tenochtitlan. Totonacs, Chalco communities, and others also saw the Spaniards as an opportunity to escape or avenge Aztec authority. By the final campaign against Tenochtitlan, Spaniards may have been only about one half of one percent of the total force, while the overwhelming majority were other Mesoamerican peoples. The Inca case has a parallel: at Cuzco, subordinated groups such as mitmies, yanaconas, Canarai, Chachapoya, and Ayamaca joined the Spaniards because they saw a chance to break Inca dominance.

A third view is the importance of decapitation vulnerability. Centralized hierarchy can magnify command, but it can also magnify capture. Once Cortes held Moctezuma, he could govern through an increasingly puppet-like emperor. Once Pizarro held Atahualpa, he had leverage over the top of the Inca social order. Sheppard’s reading of Moctezuma is especially useful because it avoids simple mockery. Inviting Cortes into the capital was not necessarily the decisive error; remaining personally available for seizure was. The lesson is not “never interact with outsiders.” It is that central nodes must not be left in positions where a hostile actor can directly control them and then use the system’s own hierarchy as an amplifier.

A fourth view is that the Aztecs and Incas were not passive, innocent, or incapable of learning. The episode lists serious adaptations: confined spaces, rooftops, holes through buildings, flooding cities or rivers, Spanish swords mounted on long poles, bolas used against horses, captured crossbows and firearms, and Manco Inca learning to ride a horse as a demonstration. Inca forces could also win terrain-specific battles. Quizzo Yupanqui used narrow mountain passes, blocked both ends, and dropped boulders from above to annihilate Spanish relief expeditions. These examples matter because they shift the explanation away from native stupidity and toward time, institutional learning, and strategic compression.

The limitation is that short-term contingency and long-term structure point in different directions. Sheppard explicitly says the conquest could have failed in the near term. If Cortes had been killed early, the expedition might have disintegrated. If Atahualpa had kept negotiations at a distance instead of entering Pizarro’s trap, he might have frustrated Pizarro. If the Inca had trapped the entire conquistador force in the right mountain pass, horses, infantry, and metal armor could all have become irrelevant. Yet Sheppard also argues that Eurasian diseases, sea power, coastal strongholds, fortress technology, continued European arrivals, subject peoples’ resentment, and centralized imperial vulnerability would have made long-term resistance extremely difficult.

The conversation then expands from conquest to institutions and globalization. Sheppard treats Potosi silver as the final piece in true globalization: silver flowed into the war finances of Charles V and Philip II, then through Manila into trade with China, Japan, Siam, and other Asian suppliers. But he does not present silver as an uncomplicated Spanish success. He argues that the silver influx imported inflation, raised prices, lowered real wages, and weakened the pressure to develop capital markets and independent banking systems like those emerging in the Netherlands and England. His “quick fix” analogy links silver, gold, and monocrop production: easy extraction can trap a society in an extractive environment and reduce the incentive to build more productive capabilities.

The episode also preserves a moral boundary. Sheppard says Spanish observers may genuinely have been horrified by Aztec human sacrifice and could use Christian mission language to soothe their consciences. But he describes the underlying conquest as a low, extractive undertaking. Dwarkesh emphasizes massacres, rape, assault, and intentional genocide by Spaniards in the New World, and Sheppard does not deny the brutality of the accounts. The important analytical limitation is that religious horror can explain rationalization; it does not convert colonial violence into liberation.

4. Learning and Application

The most useful lesson is not that advanced technology automatically wins. It is that technology becomes decisive when it is combined with training, organization, logistics, morale, political alliances, and speed. In the episode, guns are not sufficient. Horses, steel swords, helmets, cavalry shock, formation knowledge, and psychological effects matter together. Applied to modern organizations, the right question is not “does the outsider have a better tool?” but “can the outsider combine that tool with capital, distribution, partners, expertise, and a tempo we cannot match?” A narrow feature comparison can miss the system-level advantage.

The second application is to distrust the appearance of internal unity. The Aztec and Inca empires looked vast, but many subject peoples did not experience imperial survival as their own survival. External actors usually search for these fractures first: neglected partners, resentful subsidiaries, angry customers, succession losers, regional factions, or communities that believe they can get a better deal elsewhere. A practical diagnostic is to ask who inside the system would welcome an outside entrant as leverage. If key groups see the outsider as a bargaining tool rather than a threat, the system is already permeable.

The third application is to reduce decapitation risk. Centralization improves coordination, but it can make the whole system vulnerable to the capture of a ruler, administrator, credential, platform, dataset, payment channel, or cloud account. Moctezuma and Atahualpa were political apexes; modern equivalents may be root permissions, identity providers, production signing keys, model weights, procurement chokepoints, or a founder whose personal judgment substitutes for process. The response is not to abolish hierarchy. It is to create boundaries around critical nodes: delegated authority, succession procedures, offline recovery paths, dual-control systems, crisis communications, and rehearsed responses to coercion.

The fourth application is to take learning curves seriously. The Aztecs and Incas improvised intelligently, but they did not have enough time to convert improvisation into doctrine. A team facing a new competitor or technology can make the same mistake: a few clever pilots create the feeling of adaptation, while the organization has not changed training, incentives, tools, procurement, incident response, or decision rights. Real adaptation requires institutional memory. It needs drills, feedback loops, operating manuals, and enough time for people to practice under pressure.

The fifth application concerns resource shortcuts. In Sheppard’s account, silver gave Spain purchasing power and war finance, but also encouraged inflation, extraction, and delayed development of banking, capital markets, and industrial capacity. The business version is a company that receives a monopoly rent, subsidy, regulatory loophole, distribution windfall, or speculative boom and mistakes it for durable capability. The boundary question is whether the windfall forces the organization to become more productive, or lets it postpone learning. Easy money can be useful if invested into capability; it is dangerous when it becomes a substitute for capability.

Finally, the episode is a warning about using powerful outsiders. Indigenous elites who allied with the Spaniards were not necessarily irrational. A claimant in Teshkoco could see Spanish support as a path to a throne; other elites could imagine better terms under Spanish rule than under Aztec rule. Their error, in Sheppard’s account, was not that they failed to see the short-term bargain. It was that they failed to imagine the demographic and institutional scale of later Spanish and European arrivals. The modern boundary is similar. Working with a powerful platform, model provider, investor, or state can be a rational lever if exit rights, data control, substitutes, and internal understanding remain intact. It becomes dangerous when the short-term deal creates an irreversible dependency and gives the outsider a path to reorder the system through you.

Source

• Original episode: How did a few hundred Spanish soldiers topple two empires? – Si Sheppard