personal blog

Many of us may remember Voltaire as an Enlightenment era philosopher who famously preached moderation (i.e. “Use, but do not abuse”) while consuming 40-50 cups of coffee a day. Known for his satirical talent, he also argued that we are trapped in a hall of mirrors of anthropomorphism since the first flickering shadows of a cave fire. To him, our insistence on instilling the human heart, temper (and also bias) onto every god, storm (and now algorithm) was perhaps the best of blasphemies but also perhaps the worst of satires. According to his ledger, we may have paid back God in full as “God created man in his own image, and man has been returning favour ever since”. But what gave such credibility of foresights to those who borrowed our appearance, heart and mind? One explanation may be that these “early revered high priests” thrived by manifesting raw and chaotic signs of nature into a divine inevitability.

In ancient Mesopotamia, early priests known as baru, created the anatomical atlas of clay tablets holding the decoding of “divine script” inscribed upon the livers and organs of sacrificed animals. A couple millennia later, the high priests of Ancient Greece tossed pebbles, beans and bones and interpreted the “divine will” in the Oracle of Delphi. When Persians invaded Greece in 480 BCE, the oracle called the Greeks to meet Persians at sea, which defeated the Persians and paved the way for the Golden age of Athens. In this transition from the bloody entrails of a Mesopotamian sheep to the rolling dice of a Greek oracle, the high priest underwent a profound transformation: he ceased to be a mere “predictor” of future but became an “author” of a rising epoch.

Another few millennia on, the heirs of the high priests have found themselves in the temple of technology. We have traded the ‘divine script’ of the sheep’s gut for the prophetic power of data and algorithm to use the deity of mathematics to counter the terrifying silence of an unscripted future. While our tools have become sharper and our languages more technical, we still search for signals within noise, for direction within uncertainty, and opportunity within crises. 

Today, the high temple of the Federal Reserve has countless committees analyzing labour market conditions, inflation, housing market, credit spreads to set rates. Despite the immense rigor, a familiar dynamic persists. The data rarely points in a single, unambiguous direction. Inflation may be easing, but not evenly (e.g., widening gap between services and goods inflation). Employment may be strong, but with signs of fragility beneath the surface (e.g. there’s a degree of stability, there’s zero net job creation in the private sector over the past six months). Forecasts and models often diverge, and judgment has to inevitably enter. 

There is an inherent feedback loop between the data and the policy. The data informs policy, but policy in turn influences the data that will later be observed. An increase in policy rates transmits throughout the financial system and creates ripple effects across borrowing costs, asset prices, consumer demand, and so on. As a result, the oracle shapes the path of the economy. Markets, businesses and consumers adjust their behaviour by delaying and accelerating investment decisions, hiring plans and recalibrating home purchases.

In ancient Greece, Delphi was not the only site of interpretation. In the agoras of the city states, many civic decisions were guided by a more distributed form of divination known as cleromancy. Citizens and magistrates casted beans, pebbles and bones to predict the future and decide among uncertain choices, including appointments to office, jury service, etc. Cleromancy provided a practical tool for navigating ambiguity in the absence of perfect information. If the Federal Reserve replaced the institution of Delphi, the boards of companies and management teams may have replaced the casting of lots in agora with dashboards, scorecards and KPIs. 

Compared to the agora-era pebble throwing, data and technology have enabled us to invent measurable metrics that can be tracked and reviewed with regularity. Dashboards update in real time, forecasts become iterated and action items are tied to measurable outcomes. There is comfort in this structure by knowing what is being optimized. We create alignment in shared targets. Modern management theory takes pride in such an operational cadence and in their ability to translate ambiguity to quantifiable objectives and measurable execution. It’s the triumph of modern management practices. And yet, for all its admirable precision, there lingers a dread that what is most readily measured does not reflect the most accurate reflection of reality.

What follows is not so much a failure of method as it is a consequence of its success. Once a metric is chosen, it begins to exert a certain influence over its conduct. Revenue targets encourage a pulling forward or generation of demand; efficiency measures invite a trimming here and a tightening there; engagement rises, though not always in the manner that might have been intended. The figures, having been devised to observe, acquire a curious habit of self-directing. Over time, this creates a certain narrowing of perspective. Each number is tracked and each trend is explained, but not everything fits neatly into these measures. The durability of customer relationships, the quality of the decision making process, and the gradual process of trust building are inherently harder to quantify but crucial for success. Therefore, despite the distance from the ancient agora, the resemblance remains. The Greeks used randomness of cleromancy because they accepted uncertainty. We use systems and metrics to reduce it. However, the underlying limitation of fully capturing everything that matters still persists. 

Zooming back out, the same dynamic reveals itself at the macro level, though with consequences of a far larger order. The Federal Reserve, in its charge to steady the economy, has proceeded with studied care. Their citation of reading inflation, employment and financial conditions may sound like the barometer before a storm. Following these readings come raising or lowering of interest rates, which results in cooling or heating of demand. In recent decades, such efforts have led to extended periods of low rates and easy money, particularly in the wake of crises. While these policies were grounded in data and enacted to sustain growth, buoy markets amidst crises and preserve employment, there has been an accelerating accumulation of debt across public and private sectors. Despite the Fed’s denial (e.g., Debt is a fiscal issue), lower rates meant a cost easier to bear, and thus easier to extend. One begins to sense a certain irony where the very instruments designed to steady the system, over time, have encouraged a dependence upon the conditions they create (e.g., we need low rates to keep the cost of the high debt balance low, which also increases the appetite for lending and borrowing).

Looking ahead, as the world turns increasingly towards artificial intelligence, trained upon vast records of past data, we may inevitably deepen our reliance on what can be observed, recorded and quantified. These systems are, by design, brilliant at detecting patterns within what is already known, but nevertheless, they are bounded by it. What lies outside the record, such as subtler forces and emerging shifts, will remain less accessible to their logic and those who rely on them. The risk is not of certain miscalculation of what’s measurable, but of a gradual divergence between a world in which we grow ever more precise in our reasoning, and yet, perhaps, ever so slightly removed from the fuller reality we seek to understand.

Greek tragedies are seldom tales of ignorance, but rather of misplaced certainty. Kings and heroes sought clarity where none could be fully given. It was their eagerness to act that hastened the very outcomes they hoped to avoid. Oedipus, in feeling his fate, walks directly into it; Croesus, assured of victory, brings ruin upon his own dynasty. The oracle spoke, but the weight placed upon its meaning carried consequences of its own. The uncomfortable resemblance exists in the modern oracle of metrics. Metrics and indicators exist as aids to understanding. But over time, they acquire a certain authority. Growth rates, inflation targets, and our scorecards start to guide decisions with increasing conviction. Yet, with enough repetition, they begin to feel less like signals and more like destinations.

After his 61st coffee of the day, Voltaire warned of our tendency to cast ourselves into the unknown. It seems as though we have not abandoned this instinct so much as refined it into our metrics, models and now our machines. We have continued to embed our assumptions of how the world ought to behave. After all, we may still be in the hall of mirrors, asking for guidance from our increasingly polished reflection, rather than discovering the shape of reality.