There are days when I feel our species is a kind of orchestra that forgot the score but still plays with great confidence. Violins screaming in different keys, trumpets improvising in panic, the conductor having a small existential crisis. And yet, we insist with straight faces that this is rational decision-making. Kahneman, Sibony, and Sunstein gave a polite scientific name to this chaos: noise. Personally, I think they were being too gentle. I would call it “the grand human lottery, disguised as expertise.”
I came to this conclusion not during some philosophical retreat, but while watching two doctors give two completely different diagnoses to the same person, based on the same symptoms, on the same day. One said stress. Another said infection. A third, when forced to participate, said maybe both but also maybe neither. At that moment I realised not only medicine, but most human judgment, is like pulling a lever on a slot machine: sometimes you win, sometimes you lose, and sometimes you get something so unexpected you start questioning the meaning of life.
But let me be precise. Noise is not bias. Bias is predictable, like a dog that always barks at postman. Noise is the dog that barks sometimes, sleeps sometimes, bites you sometimes, and occasionally gives philosophical advice. Bias we understand. Noise we underestimate.
And noise is everywhere.
Two judges, same crime → wildly different sentences.
Two HR managers, same CV → different job fate.
Two auditors, same data → different risk rating.
Two people reading this essay → one enjoys it, one suspects I am drunk.
Our judgments drift like weather patterns: shaped by mood, hunger, last phone notification, memory of an email we forgot to answer, even the colour of the room. Kahneman’s work shows this variability is not marginal—it is gigantic, sometimes larger than the effect of genuine skill. We think we are consistent creatures; in reality, our mental machinery rattles like old factory equipment, producing errors not because of ideology, but because of randomness.
And here comes my favourite part: we barely notice it. When human judgment fails, we usually blame bias, stupidity, corruption, lack of training, bad luck, or the alignment of planets. But noise? No, we almost never blame noise, because noise is invisible. Randomness in our head does not come with a warning label.
It would be funny if it was not so tragic. Or maybe tragic if it was not so funny. Hard to tell sometimes.
Let me give an example. Imagine two insurance underwriters evaluating the same case. In a perfectly rational world, their prices should be almost identical, like two clocks ticking in sync. In the real world, the differences can be so large that one might as well be pricing a bicycle and another a spaceship. Not because they disagree on principles. Simply because their minds wander in different directions like two drunk tourists in a foreign city.
This is noise: scattered judgments where the scatter itself is the problem.
The logical question, then, is: why do we tolerate it?
Partly because we enjoy the illusion of autonomy. We like to think our decisions are handcrafted masterpieces, uniquely human. We resist standardisation because it feels mechanical, cold, authoritarian. It feels like losing individuality. But the irony is that our individuality, in decision-making, is often just entropy wearing a costume.
Another reason is that we still live with the romantic myth of “the expert.” The wise doctor. The seasoned manager. The experienced judge. We want to believe these roles come with superhuman precision. But consistent accuracy is not what experience reliably gives us. Experience mostly gives us confidence, which, from scientific view, is not strongly correlated with correctness.
And before anyone thinks I am advocating for robots making all decisions—no, relax. I am not preparing you for some AI uprising. I am simply acknowledging that human thinking includes enormous random variation. Recognising this is not anti-human. It is honest.
There is also a darker angle. Noise is comfortable for institutions because it hides responsibility. If ten judges produce ten wildly different sentences, who do you blame? If three managers give three incompatible performance evaluations, who is wrong? If half of the medical diagnoses contradict each other, who owns the mistake? Noise diffuses accountability like smoke in the air.
But here is the twist. Noise is not an insult to humanity. It is part of our wiring. We are biological machines with chemical moods, imperfect memories, fluctuating attention, and cognitive shortcuts built for survival, not precision. We are noisy by design. Which is exactly why recognising noise matters. Only then we can tame it—not eliminate completely, but reduce the spread, like turning chaotic orchestra into something resembling music.
Structured decision rules, checklists, predictive models, consensus reviews—these tools do not remove humanity. They anchor it. They reduce randomness so that real expertise, when it exists, can actually show through.
Because if randomness dominates, skill becomes irrelevant. And we end up living in a world where important outcomes—sentences, pay, diagnoses, approvals—are determined not by logic but by the mental weather of whoever handles the case.
This is not a dystopian fiction. It is current reality. The dystopian part is that we pretend otherwise.
Still, I remain optimistic. Humanity’s strength is not precision but adaptation. Once we see a flaw clearly, we find ways to improve. And understanding noise is one of those rare insights that can genuinely make society more fair, more predictable, more sane.
But first, we must admit what we truly are: messy creatures whose minds drift like clouds, whose decisions wobble like old furniture, and whose confidence often exceeds our accuracy by impressive margins.
Maybe this admission sounds uncomfortable. For me, it sounds liberating. Finally the orchestra knows it’s playing without a score. And maybe now, with enough humility and science, it can start tuning itself—one instrument at a time.
