Perhaps reading also requires introducing a conversion rate. Many books are interesting when you read them, but they are easy to forget afterward or fail to crystallize into deeper mental models.

For example, I recently read Edward Chancellor’s The Price of Time, which recounts the history of interest rates. What stuck with me was that in ancient Greece and Rome, before coinage was invented, trade relied on barter; yet money was not a wholly independent invention. Even before formal currency emerged, concepts like interest rates already existed. Intriguingly, borrowing rates varied drastically across different commodities. In ancient Greece, borrowing wheat carried an astonishing 33% interest rate. The book provides an explanation: grain was borrowed primarily during acute shortages or crop failures—times when people faced starvation. Facing hunger, borrowers were willing to accept 33%, which naturally made it costly. Once the emergency passed and the next harvest arrived, plummeting grain prices allowed borrowers to repay the interest. Such exorbitant rates compensated lenders for steep price volatility. There was another factor: whenever a new Greek ruler took power, much like imperial amnesties in ancient China, aristocrats loved being generous with other people’s wealth, decreeing general debt cancellations. To hedge against the risk of political jubilee, lenders naturally baked a risk premium into baseline interest rates.

What was intended as debt relief for the common people morphed into a default risk that lenders had to price in, driving interest rates higher. A spectacle staged by the aristocracy to win popular favor ended up making survival even harder for ordinary citizens—who likely remained grateful nonetheless. To me, economics is most fascinating when revealing these warped causal loops, much like a thriller film where discovering the underlying cause leaves you reflective.

The book also recounts the story of John Law. As a close confidant of the French Regent, Law leveraged royal power to experiment with his financial theories, establishing the Banque Royale, issuing fiat paper currency, and artificially suppressing interest rates. This triggered an epic speculative bubble. An insightful detail is how low domestic rates triggered capital flight: depressed yields across France prompted capital to flee toward higher-yielding European jurisdictions and across the Atlantic into the Americas. This unfolded in the 18th century, yet it describes 21st-century macro dynamics with uncanny accuracy.

The lesson of the book is that time has a price: interest. Reinvesting accumulated interest yields compounding—a phenomenon humans have practiced since antiquity.

Recently, I began thinking about compounding in the context of reading. Accumulating knowledge can also generate compound returns. Yet when reading, we retain only a fraction of what we consume. From The Price of Time, I primarily recall those two anecdotes. They might prove mildly useful—if someone asks about ancient Greek grain loans, I can readily quote 33%—but the vast majority of the book faded. In my experience, we only retain knowledge that intersects directly with our personal experiences and innate curiosity.

Hence, reading has a conversion funnel:

Books * Curiosity => Retained Knowledge

This filtered knowledge represents potentially useful insight. Yet whether it is ever applied (such as calculating ancient Greek wheat rates) faces another hurdle: can it connect with real-world practice?

Books * Curiosity * Practice => Actionable Knowledge

Thus, the actual conversion rate of reading is extraordinarily low. Most content never captures our attention, and the small portion that does is rarely put into practice.

This realization struck me after reading Colin Huang’s public essays. His writings contain profound reflections that command admiration (though I credit fate for letting me read them late; had I read them earlier, I would have lost far more money on PDD stock). Huang frequently cites Bertrand Russell’s The Conquest of Happiness. It is evident that he reads deeply and integrates insights with lived experience. Once PDD’s share price recovers, I plan to write an essay on my takeaways; writing it now would sound like the rationalizations of a trapped bagholder, so I will wait for a rebound.

Huang’s reflections on Russell’s work—specifically his synthesis of common sense and anxiety—made me realize that although I read the same book years ago, my personal application and synthesis were virtually zero. I had merely turned the pages without retaining the substance. Re-reading Russell felt like a slap in the face. Russell warned against reading for vanity: reading solely to log a review or check off a “read” milestone means reading to show off and soothe personal anxiety, not out of authentic curiosity. No wonder nothing useful stuck.

Huang distilled the essence of The Conquest of Happiness into one maxim: respect common sense. I agree. Most human disappointment and anxiety stem from defying common sense. In a bull market, everyone fancies themselves an investing genius; novices recklessly add leverage—classic symptoms of disregarding common sense.

Common sense must be deliberately cultivated. Absorbing fundamental common sense from reading represents true knowledge accumulation; even if sparse, this baseline common sense is what actually generates compounding returns. Russell’s advice to “respect common sense” is conspicuously absent from many people’s mental models. By reading and continuously sharpening our judgment, this knowledge integrates into every decision we make.

Books * Curiosity * Practice * Common Sense => Compounding Knowledge

“Common sense,” in a literal sense, is knowledge readily accessible for everyday decision-making. Consider a concrete example:

In Philip Fisher’s Common Stocks and Uncommon Profits, there is an explicit chapter arguing that one should boldly buy equities during wartime. Fisher’s thesis is straightforward: war is inherently bearish for fiat currency, making the hoarding of cash during conflict foolish. Investors should steadily accumulate quality equities during wartime, anticipating that markets typically rebound vigorously once hostilities cease.

Regrettably, though I had read the book, I had zero recollection of that chapter, causing me to miss the market opportunities during the US-Iran geopolitical conflict. The insight had failed to spark my curiosity on first reading, and thus never transformed into intuitive common sense.

Books do not yield value as easily as we imagine. Reading does not inherently compound; only knowledge rigorously distilled into “common sense” yields compound returns.

This reminds me of modern Large Language Models. An LLM ingests the entirety of internet knowledge, having scanned nearly every digitized book in existence. In theory, it possesses boundless intelligence—yet it cannot inherently determine how to apply that knowledge.

While an LLM understands computer science fundamentals and programming paradigms, it lacks innate purpose when writing code. It requires human direction to marshal the right knowledge to accomplish a task. Modern coding agents focus heavily on training or prompting models to adopt adaptive problem-solving behaviors. Developers are essentially teaching the LLM how to utilize its knowledge. The prompts, agent harnesses, and evaluation frameworks accumulated by developers function precisely like “common sense,” continuously compounding within the agent’s development loop.

Returning to the formula:

Books * Curiosity * Practice * Common Sense => Compounding Knowledge

Life is finite; we cannot read every book. As Russell noted, we should read what genuinely interests us, not read to impress others. Since curiosity is bounded, maximizing “compounding knowledge” requires focusing our energy on “practice” and “common sense.” Rather than racing through more volumes, we are better served curating the ideas that truly captivate us, consciously practicing them in reality, and extracting durable common sense.

LLMs govern the knowledge base; humans understand how to apply it.