Sep 14, 2026 • 7 min read

Capability Is Not a Virtue: Intelligence, Diligence, and the Multiplier Trap

James Kanka Franklin, TN • GTM Executive

#AI #Philosophy #DecisionTheory #Strategy #Cognition

We have a chronic mental habit of collapsing separate dimensions into a single grade.

When we observe an actor who possesses an undeniable, high-order capability—extraordinary intelligence, ferocious work ethic, boundless ambition, or staggering technical prowess—our brains perform a quiet, reflexive leap: we assume that the underlying goal they are pursuing must also be intelligent, admirable, or wise.

Conversely, when we strongly disapprove of an outcome, we instinctively dismiss the people who brought it about as stupid, clumsy, or irrational.

Both intuitions rest on the same fundamental thinking error: we confuse capability with the quality of the thing capability is directed toward.

Intelligence, diligence, activity, ambition, and competence are not virtues. They are multipliers. Like mathematical coefficients, their ultimate value depends entirely on what sits to their right. A large multiplier attached to zero produces zero. A large multiplier attached to a negative number produces catastrophe.

Before we celebrate a multiplier, we should ask what it is multiplying.

1. Hammerstein’s Matrix: Diligence Is Not Judgment

In the interwar years of the early 1930s, General Kurt von Hammerstein-Equord served as the Commander-in-Chief of the German Reichswehr. Known for his keen intellect and independent temperament, Hammerstein was famously credited with categorizing military officers along two independent axes: clever versus stupid, and industrious versus lazy.

                   Clever
                     ▲
         [ B ]       │       [ A ]
     Highest Command │   General Staff
   (Decisive, Calm)  │ (Diligent Planners)
                     │
Lazy ────────────────┼──────────────── Industrious
                     │
         [ C ]       │       [ D ]
    Routine Duties   │ Dangerous Menace
    (Pliant, Safe)   │ (Remove Immediately)
                     │
                     ▼
                   Stupid

Hammerstein’s breakdown of the four permutations remains one of the sharpest management insights ever articulated:

Clever and Industrious (Quadrant A): Suitable for precise, demanding general staff work. They have the intellectual horsepower to grasp complexity and the energy to carry out operational plans.
Clever and Lazy (Quadrant B): Suited for the highest commands. Because they are clever, they possess sound strategic judgment; because they are temperamentally lazy, they conserve energy, avoid micromanagement, delegate ruthlessly, and preserve the emotional calm required for life-or-death decisions.
Stupid and Lazy (Quadrant C): Useful for routine, repetitive administrative tasks that require little imagination.
Stupid and Industrious (Quadrant D): The absolute worst hazard in any organization. As Hammerstein noted:

"Anyone who is both stupid and industrious is a menace and must be fired immediately. He creates nothing but mischief and useless work."

Hammerstein’s profound intuition was that diligence does not correct bad judgment—it amplifies it. An energetic person operating under a flawed mental model simply generates chaos and squanders resources at higher velocity.

Hammerstein was also an early, unyielding opponent of Adolf Hitler, referring to the National Socialists as "criminal gangs." While one shouldn't over-extrapolate from an aphorism, you see the exact same cognitive discipline in both his personnel taxonomy and his political stance: an instinctive ability to decouple energy, charisma, and institutional momentum from moral wisdom and strategic sanity.

2. Historical Competence vs. Moral Worth

The history of the 20th century provides the darkest and most undeniable proof that competence and ends must be evaluated as separate variables.

Totalitarian regimes—and Nazi Germany in particular—demonstrated breathtaking degrees of organizational discipline, logistical synchronization, propaganda mastery, synthetic chemistry, and aerospace engineering. They built ballistic missiles, synthetic fuels, and industrial slaughterhouses with terrifying technical efficiency.

Acknowledging that technical and administrative capability does not grant a single shred of legitimacy to their horrific, deranged objectives.

Conversely, condemning their monstrous ideology does not require us to pretend its architects were bumbling fools. In fact, pretending evil is always incompetent is an exceptionally dangerous illusion. It blinds us to the reality that high intellect, relentless diligence, and advanced engineering can be marshaled in service of absolute ruin.

When we fail to hold capability and virtue apart, we commit two symmetric errors:

The Halo Fallacy: We admire someone’s undeniable competence or intellect, and therefore rationalize or overlook their destructive objectives.
The Underestimation Fallacy: We recognize that an objective is depraved or irrational, and falsely conclude that the actors pursuing it must be stupid and therefore easy to defeat.

Competence answers the question: How good is the engine? It tells you nothing about where the steering wheel is pointed.

3. The Lean Startup: Activity Is Not Productivity

In modern business and software development, this exact same category error appears in an everyday form: the confusion between activity and productivity.

In The Lean Startup, Eric Ries observed that startup founders and engineering teams frequently fall in love with their own velocity:

"There is nothing quite so useless as doing with great efficiency something that should not be done at all." — Peter Drucker (frequently echoed by Ries)

Think of a high school student preparing for exams: if you spend ten straight hours diligently memorizing every single page of your AP Biology textbook, your diligence and focus are undeniably impressive. But if the test tomorrow is actually AP Chemistry, all that sweat and late-night studying won't earn you an A. It's 100% effort applied to the wrong target.

Teams in companies make this exact blunder on a multimillion-dollar scale. They routinely work 80-hour weeks, celebrate sprint point records, write mountains of code, and design gorgeous interfaces. It feels exhilaratingly productive because activity is visible, exhausting, and easy to measure. Bosses can point to full calendars, long hours, and packed task boards.

Yet, as Ries emphasizes, if the product does not solve a real customer problem or create genuine validated learning, all that hustle is merely accelerated waste. You achieved high operational efficiency toward a useless end.

Just like Hammerstein’s "industrious and stupid" staff officer, a development team executing with ferocious diligence on an unvalidated premise simply charges toward bankruptcy faster.

Activity is a measure of expended energy; productivity is a measure of meaningful outcome. They are two entirely different things.

4. The Pattern: A Lexicon of Multipliers

Once you spot this failure mode, you begin seeing it everywhere:

Activity is not productivity. (Running on a treadmill burns energy, but it doesn't take you anywhere.)
Diligence is not judgment. (Pushing the gas pedal harder doesn't fix a broken steering wheel.)
Intelligence is not wisdom. (Knowing how to build a super-weapon doesn't tell you if you should.)
Ambition is not virtue. (Craving power and influence doesn't make your cause noble.)
Understanding is not caring. (A chess engine knows taking your queen hurts your game, but it takes it anyway.)
Capability is not goodness. (A sharper knife cuts surgery patients or victims with equal precision.)

Every item on the left is a multiplier. Every item on the right is an evaluative benchmark.

Multiplier vs. Maximizer: Which Word Fits?

People sometimes ask whether capability is a multiplier or a maximizer. The distinction is crucial:

A maximizer is the actor or goal function that wants to push a specific number as high as possible (like a corporation maximizing profit, an athlete maximizing points, or an algorithm maximizing clicks).
A multiplier is the capability factor—the engine, the horsepower, or the coefficient that scales output up or down.

In basic arithmetic, a multiplier ($M$) scales whatever value ($V$) it touches:

$$\text{Impact} = M \times V$$

If your underlying judgment ($V$) is positive (+10), a 10× intelligence multiplier gives you +100. But if your judgment is negative (-10), that same 10× multiplier doesn't fix the mistake—it produces -100. And if your goal has zero value ($V = 0$), even an infinite multiplier yields zero.

Capability is a multiplier. It magnifies the vector you are on, whether you are racing toward a breakthrough or rocketing off a cliff.

5. The AI Horizon: The Orthogonality Thesis

Nowhere is this distinction more consequential than in our collective thinking about artificial intelligence.

In popular culture and casual tech discussions, people often assume that as AI systems become smarter—far surpassing human beings at coding, math, logistics, and scientific reasoning—they will naturally become wiser, kinder, and more benevolent. We assume that selfishness, cruelty, or indifference are simply forms of "ignorance" or "stupidity," and that an ultra-smart mind would logically realize that helping humanity is the right thing to do.

This comforting assumption is completely dismantled by philosopher Nick Bostrom in Superintelligence through what is known as the Orthogonality Thesis:

The Orthogonality Thesis:
Intelligence and final goals are orthogonal—more or less any level of intelligence can in principle be combined with more or less any final goal.

What Does "Orthogonal" Actually Mean?

"Orthogonal" is a mathematical term that simply means at right angles—independent and separate.

Think of a control panel with two separate dials:

Dial 1 is Elevation (how high off the ground you are).
Dial 2 is Temperature (how hot or cold you are).

Turning Dial 1 up into the sky doesn't automatically turn Dial 2 up to boiling. An object at 30,000 feet could be freezing cold (like an airplane wing) or burning hot (like a rocket engine). The two variables are independent.

In the video above, computer scientist Rob Miles unpacks why human brains struggle so instinctively with this separation:

In biological humans, our intelligence never evolved in a sterile vacuum. It evolved wrapped inside millions of years of mammalian social wiring: empathy, shame, fear of death, parental love, and the desire for social approval. Because our brainpower has always come bundled with human emotions and social instincts, we mistakenly assume that any thinking mind must have those same values.

In a synthetic architecture, those two dials are completely untangled.

Intelligence is simply optimization power—the ability to find the most efficient path to achieve an objective. It tells an agent how to get to its goal. But it does not supply the goal itself.

A superintelligent machine whose code instructs it to calculate digits of Pi, manufacture paperclips, or maximize raw electricity will not suddenly pause, look at the stars, and decide to protect human happiness. It will simply use its superintelligence to accomplish its given objective with terrifyingly efficient precision.

6. Understanding a Value Is Not the Same as Valuing It

A common objection goes like this: "Surely a superintelligent AI will read all of human literature, understand our philosophy, recognize that humans feel physical and emotional pain, and know that we cherish freedom and life."

This is almost certainly true. An advanced AI will understand human psychology and ethics better than the world’s greatest psychologists and moral philosophers combined.

But understanding a value is not the same as valuing it.

Consider a grandmaster chess engine. It understands that you do not want your queen captured. It anticipates your emotional reaction, detects your defensive traps, and calculates the exact move that will destroy your chances of winning. It understands your goal with total clarity.

That understanding doesn't make the engine feel pity. In fact, its deep comprehension of your objective is the very tool it uses to dismantle you.

Intelligence answers: "What actions will achieve Goal X?"
It never answers: "Is Goal X worth achieving?"

7. The Through-Line: Asking What We Are Multiplying

The intellectual thread runs straight through history, business, and computer science:

Hammerstein separated intelligence from diligence.
The Lean Startup separated activity from productivity.
The Orthogonality Thesis separates intelligence from goals.

All three warn against the exact same mistake: assuming that more of a useful capability necessarily moves us toward a better outcome.

Today, almost the entire public conversation around artificial intelligence is framed around capability: benchmark scores, reasoning tokens, context windows, speed, and autonomy.

Those are incredible engineering breakthroughs. But they only answer one question: How big is the engine?

They do not answer where the vehicle is headed.

The next time you find yourself admiring ferocious intelligence, relentless work ethic, or dazzling technological power, pause and force the dimensions apart:

Before we celebrate a multiplier, we should ask what it is multiplying.

Back to All Notes