Reading the World

The earnings call is forty-eight minutes long, and Sonja listens to all of it on a Tuesday evening in April, walking loops around the Stadtgarten because she cannot sit still for this.

Sonja is in her mid-thirties, holds a degree in insurance economics, and has spent ten years doing everything right at a large insurer whose restructuring memos have lately started deleting the career ladder above her. In all those years she has never once listened to one. Earnings calls were for analysts and journalists. Employees got the internal version: the town hall, the CEO’s confident slides, the intranet article with the comment section nobody used. But the notebook she started keeping this winter — part learning log, part weather chart — has changed her habits, and the notebook says: primary sources.

The CFO’s voice is calm, almost bored. Twenty-two minutes in, an analyst from a bank asks about cost development in claims operations. The CFO says the phrase “structural efficiency program” twice in one answer, and then this: “We see a path to an automation-first claims organization over the planning horizon, with corresponding effects on personnel expense ratios.”

Sonja stops walking. Corresponding effects on personnel expense ratios. At the town hall three weeks ago, the same program was called “Colleague Empowerment through AI.” There was a slide with a rocket on it.

At home she opens the careers page and counts, because counting is what she does now. Claims positions posted in North Rhine-Westphalia: two, both fixed-term. A web archive of the same page from eighteen months ago shows fourteen, mostly permanent. Meanwhile, five new postings of a kind she has never seen: AI Operations Specialist. Model Validation — Insurance. Prompt & Workflow Designer, Claims.

Then the small one, the one that actually lands. In the intranet news archive, a two-line notice from February she’d scrolled past when it appeared: the high-potential program — her program, the one that chose her at twenty-nine, the mentor lunches, the executive-floor visibility — is “paused pending review of future skill requirements.” No rocket on that one. No town hall either.

She sits at the kitchen table and writes three lines in the notebook:

What they say: empowerment, no one left behind.
What they fund: automation, validation, exits.
What they quietly stopped funding: me.

It isn’t anger she feels, or even fear, exactly. It’s the specific vertigo of realizing that the official narrative of her career — told to her in performance reviews and program brochures — was never a description of the future. It was a description of what they needed her to believe last year. Nobody lied to her, quite. But nobody was ever going to send her a memo titled The Ladder Has Been Discontinued. If she wanted to know, she was always going to have to read for it herself.

Sensing, not forecasting

The book behind this series argues that agency — the capacity to direct your own course — rests on three dimensions: capability, room to act, and will. Reading the world belongs to the first and is the least taught, most quietly decisive capability there: not “keeping up with the news,” but maintaining an accurate, continuously updated picture of what is actually changing in the part of the world your life depends on.

Start by discarding the wrong model: prediction — the belief that with enough information you could forecast what’s coming (which jobs survive, which skills pay, where the economy lands in 2031) and plan against the forecast. You can’t, and neither can the people paid to. Philip Tetlock’s long study of expert forecasting found the honest record in complex domains to be poor, and the future being built right now, by technologies whose own makers argue about what they’ll do next, is less forecastable than most. Waiting for a reliable forecast is just denial with a research habit.

The Pacific navigators who crossed thousands of kilometers of open ocean without instruments did not do it by predicting the weather weeks out. They did it by sensing continuously — the swell pattern shifting under the hull, the birds, the color of the water, the wind backing a few degrees — and adjusting course early and often. The forecast is unavailable; the signals are everywhere. That is the stance: you don’t forecast the storm; you notice the swell changing. A predictor asks what will happen?, an unanswerable question that produces false confidence or paralysis. A sensor asks what is already happening that changes my position? — an answerable question that produces adjustments.

Where change announces itself

Change in your working world is loud only at the end. Before the layoff round there is a hiring freeze; before the freeze, a budget line quietly not renewed; before that, a shift in what leadership says to the people whose opinion of the stock matters more than yours. The early stages are called weak signals — small, ambiguous, easy to dismiss individually. Reading them is not a mystical skill: it is knowing where to look, and looking on a schedule. For a person with a job and a life — not a strategy department — the highest-yield sources are close to home.

Job postings, yours and adjacent. What your employer and its competitors hire for is the most honest public statement of what they believe about the near future, because postings cost money and commit budgets. Watch the deltas: roles appearing that didn’t exist, roles quietly disappearing, permanent becoming fixed-term, requirements shifting inside a familiar title. Twenty minutes a month.

Money statements over vision statements. Earnings calls, annual reports, investor presentations — not because you own shares, but because companies are legally constrained to be more honest with investors than with employees. When the internal slide says empowerment and the analyst call says personnel expense ratios, the call is the signal and the slide is the sedative.

What quietly stops being funded. Cancellations are rarely announced; they’re just absent. The training program not renewed, the backfill that never happens, the project that stops appearing in updates. Absence is a signal that requires memory to detect — one reason to keep notes: your notebook remembers what the intranet archive buries.

Internal language drift. Memos and town halls are marketing, but marketing changes reveal strategy changes. New euphemisms are load-bearing. When “our people are our greatest asset” becomes “the right skills in the right place,” something has been decided.

Your network’s ground truth. Not the performative feed — the actual conversations. What are people three years ahead of you in your specialty doing? Who is leaving, and toward what? Mark Granovetter’s classic finding on the strength of weak ties applies directly: the most valuable signals come not from your close circle, who know what you know, but from acquaintances embedded in different contexts, who know what you don’t.

Two design rules keep this from consuming you. Bound it: a fixed slot — say, forty-five minutes weekly plus a monthly deep pass. Sensing is a routine, not a mood you live in. And log it: one page a month — signals noticed, what they might mean, what would confirm or kill each reading. The log converts anxiety into observation. Untracked signals fester; tracked ones become hypotheses with due dates.

An evidence hierarchy for one

Signals arrive mixed with noise, spin, and outright manipulation, so the second half of this capability is quality control. Before letting any claim about the future change your picture of the world, place it on a ladder. At the top: things that happened — filings, enacted rules, actual layoffs, posted jobs. One rung down: data about things that happened — statistics, studies, surveys — strong, but ask who collected it, what exactly was measured, and what the denominator is. Below that: informed conjecture — experts reasoning from evidence with stated uncertainty, never a foundation on its own; distinguish those who show their reasoning from those who show their confidence. At the bottom: interested claims — anything from someone selling something: vendors, consultancies whose product is urgency, executives managing a narrative, influencers monetizing your attention. That includes both “AI changes everything by Tuesday” and “AI is all hype, relax” — doom and reassurance are both products with markets. Interested claims are evidence about the claimant, not the world.

The single most protective habit is the boring one: ask what would count as evidence against, then look for it. Kahneman and Tversky’s work on judgment under uncertainty catalogued the machinery working against you here: we overweight vivid stories over flat statistics, see confirmation everywhere once a belief is seated, and anchor on the first number we heard. You cannot switch this machinery off. You can route around it with procedure: the ladder, the log, the deliberate search for the disconfirming case.

Denial, panic, and the nudgeable compass

Between denial and panic lies the narrow channel you’re steering.

Denial is the more comfortable failure, and the more common, because institutions actively supply it: the cocoon — the corporate shell that feels secure while quietly narrowing your options — manufactures reassurance as an operating necessity. Denial’s tell: it evaluates evidence by whether it’s pleasant, not whether it’s solid, and rejects unpleasant signals with meta-arguments (“media hype,” “they said that about the internet”) instead of engagement.

Panic is denial’s mirror: it, too, replaces reading with reflex. The panicking mind treats every signal as the storm itself, forgets that even transformations which reshape whole industries take years — and years are usable — and lurches into large irreversible moves precisely when calm small ones are available. Panic feels like taking things seriously. It is actually a way of ending the discomfort of uncertainty by choosing a certainty, any certainty, including a catastrophic one.

One principle defines the channel between them: signals should trigger moves sized to the signal’s strength — and early moves should be reversible. A weak signal — one odd earnings-call phrase — earns a note and a watch item. A converging pattern — the phrase, plus the postings shift, plus the paused program — earns a small real response: a learning project started, a conversation had, spending trimmed to extend your runway, the time your savings buy you. Only a confirmed structural change earns an irreversible commitment. Sensing without a graduated response is just doomscrolling with a filing system. And this is the honest answer to the fear that paying attention means drowning in dread: dread is perceiving threat without perceiving your capacity to respond, and the response ladder converts the one into the other.

Which raises the modern complication: the instruments most people use to watch the world are engineered to induce exactly these two failure modes. Feeds do not optimize for your accuracy; they optimize for your engagement, and the emotions that maximize engagement are outrage, fear, and tribal vindication. Add cheap synthetic media, and the ambient information supply is adversarial by default. The defenses are unglamorous and effective. Prefer primary documents over coverage of documents. Verify laterally, as Sam Wineburg and Sarah McGrew found professional fact-checkers do: before believing a source, spend a minute checking who is behind it and what others say about it, rather than reading more of the source itself. Treat emotionally perfect content — the story that confirms everything you feared or hoped — as the highest-risk category, precisely because it feels truest. And get your industry signal from the bounded routine, not the feed. You would not navigate by a compass someone else could remotely nudge. Log off the nudgeable compass.

The man who always read the water

Abdellatif’s Thursday call with his brother in Turkey starts, as always, with their mother’s knees and ends, as always, with the real agenda.

Abdellatif came to Germany from Morocco, trained as an electrician, and works in logistics while pursuing official recognition of his credential. “The recognition office answered,” he says. He’s on the balcony of his apartment north of Munich, jacket on against the evening cold, keeping his voice down because Lea, his girlfriend, is on a work call inside. “Partial recognition. They want a Kenntnisprüfung — a skills exam — or eighteen months’ documented practice under supervision.”

“Is that good?” his brother Yassin asks, from his kitchen in Izmir.

“It’s good if I move before the rules move.” And then Abdellatif does the thing he has done since he was nineteen, figuring out ferry prices and visa windows from Tangier: he reads out the water. The skilled-worker rules were loosened again and the ministry is signaling more easing — but the offices processing applications are backed up, so paper rights and real timelines are different things, and he tracks both. Two electricians he knows from the Moroccan community in Munich got full recognition in under a year — through the same office that sat on a friend’s file for two; the difference was a complete document set on day one and an employer letter. So: get the employer letter. His warehouse just lost its second maintenance subcontractor this year, and the shift leader has started asking him, unofficially, to look at things electrical — which is illegal for him to do, and is also information: the company is short exactly what he is. And the wage postings for licensed industrial electricians in Bavaria that he checks — a ritual, first Saturday of the month, phone on the kitchen table, the way another man checks lottery numbers, except this is the opposite of a lottery — have gone up again.

“So the exam,” says Yassin.

“The exam. Six months’ preparation. And I ask the company to sponsor the practice route in parallel — two paths, whichever opens first.” A pause. “The wedding money and the exam money are fighting each other. That’s next month’s conversation.”

Say it plainly, because the world rarely does: what Abdellatif is doing on that balcony is expertise. He carries in his head a live model of two labor markets, an immigration regime in ongoing reform, the processing behavior of specific offices, remittance expectations across three countries, and the difference between what a rule says and what a clerk does — updated weekly through a network spanning Munich, Izmir, and Casablanca, cross-checked against official documents he reads in his third language. No one gave him a course in sensemaking. His life depended on it, so he built it. Sonja, with her degree in risk, is at forty-five minutes a week just beginning to construct what he has run for fifteen years. The professional class is discovering, under pressure, a capability that people who navigate the world without a safety net have always had — and it deserves to be named as what it is. Not street smarts. Not hustle. Judgment, running on better data, gathered by a disciplined sensor, from a position where errors are expensive.

“You always land on your feet,” Yassin says, the old family line.

Abdellatif watches the S-Bahn lights slide past beyond the parking lot. “Nobody lands on their feet,” he says. “I look where the ground is before I jump. It’s not the same thing.”

---

Sonja and Abdellatif are constructed figures — personas built to carry real mechanisms, not case studies of real persons. The book’s method note explains how they’re used.