Sensemaking: The Power of the Humanities in the Age of the Algorithm

There is a kind of understanding that arrives only after long study, and it does not feel like the everyday act of putting yourself in someone's shoes. It is slower, more disciplined, and it reaches further. Ordinary empathy lets you guess what a person feels in a moment. Analytical empathy lets you reconstruct the world that makes their feeling reasonable—the beliefs, the history, the shared meanings that a person inside a culture never has to explain to themselves because they breathe them.

The distinguishing move is that this empathy is theory-supported. It draws on the frameworks the humanities have built for reading human worlds: how meaning gets made, how context shapes perception, how a practice looks from the inside rather than from the analyst's chair. Without that scaffolding, you are left projecting your own assumptions onto other people and mistaking the echo for insight. With it, you can interpret a worldview on its own terms.

This is why it depends on being there and on going deep. Rigorous engagement with the humanities supplies the interpretive tools; real-world immersion supplies the material those tools work on. Thick data—the textured, situated record of how people actually live—is the raw stock of the whole enterprise. You cannot feel your way analytically into a world you have only skimmed.

Done well, this empathy pays off as cultural insight: an understanding of why people act as they do that holds up because it was built from the inside out. The empathy is not the soft part of the work. It is the rigorous part.

How to

1. Move past the first level (recognizing an emotion) and second level (imagining yourself in their place) to the third: interpreting their world through their own cultural categories.
2. Ground your interpretations in thick data—extended immersion and situated observation—rather than surveys or single interviews.
3. Import a relevant humanities theory (phenomenological, hermeneutic, anthropological) to structure what you notice and to test your inferences against.
4. Continually check whether you are describing their worldview or smuggling in your own.

Watch out for

• Mistaking your emotional reaction to someone's situation for an understanding of their perspective.
• Stopping at 'walking in their shoes'—which still centers you—rather than reconstructing the world as they inhabit it.

Insight tends to arrive when you stop gripping it. The state that lets it come has a specific quality: you hold your preconceptions loosely, you suspend your expectations about what the answer should be, and you let understanding come through you rather than forcing it out of you. The distinction matters. Forcing produces the conclusion you already carried in. Receiving produces something you did not know you were looking for.

This openness is not passivity. It is a demanding form of attention that refuses to close prematurely on an interpretation. Most people meet the world already sure of what they will find, and their certainty filters out everything that does not fit. To remain unattached is to keep the filter off longer than is comfortable, to sit with ambiguity until the material itself begins to speak.

Serious engagement with the humanities cultivates this state, because reading a great work well requires the same surrender—you let the text reorganize you instead of forcing your reading onto it. That habit, practiced, becomes a stance you can bring to people and situations. And the stance is what makes genuine cultural insight possible: the truth about another world can only reach you if you have left the door open for it.

How to

1. Enter fieldwork or interviews with questions rather than conclusions, and physically record moments that contradict your expectations.
2. Notice when you feel the urge to steer a conversation toward a preferred answer, and deliberately let the silence extend instead.
3. Separate your data-gathering from your interpreting sessions so you are not filtering perception through nascent theory.

Watch out for

• Confusing receptivity with credulity—openness does not mean accepting every claim at face value.
• Treating grace as a fixed trait you either have or lack, rather than a state you can enter and lose within a single conversation.

Start with a hypothesis and you will mostly confirm it. That is the trap of the tidy method: it rewards you for finding what you set out to find. Abductive reasoning refuses the fixed starting point. You begin without the answer, cast a wide net for data, gather more than seems efficient, and hold off on the conclusion while the evidence is still shapeless.

The hard part is the middle, where you have to tolerate doubt. Deduction gives certainty and induction gives probability, but both work by narrowing early. Abduction stays open through the uncomfortable stretch when you have a great deal of material and no clean story to hang it on. What eventually happens is a leap—not a random guess, but an educated move toward the most reasonable explanation of everything you have collected. That leap is where genuinely new ideas come from, because it is not constrained to recombine what you already believed.

This is why it depends on thick data. A wide net over rich, textured observation gives the leap something worth landing on; a thin net over numbers gives you back only the categories you brought. Gather deeply, doubt honestly, and the best available explanation will present itself as cultural insight—the reading of a world that feels, once stated, obviously true and was invisible until someone made the jump.

How to

1. Delay committing to any explanation until you have gathered data from unexpected and adjacent sources, not just the obvious ones.
2. When an anomaly won't fit your existing frame, treat it as the seed of a new explanation rather than noise to discard.
3. Ask repeatedly 'what would have to be true for this to make sense?' and follow the most economical answer.

Watch out for

• Collapsing into premature closure because sitting in doubt is uncomfortable.
• Mistaking a wild guess for an abductive leap—the leap must be the most reasonable explanation of the full evidence, not the most exciting.

The expert does not run through a checklist. The wine taster, the seasoned physician, the person who has lived inside a craft long enough—they perceive the right answer before they can tell you the rule that produced it. This is phronesis, practical wisdom: an intuitive, context-dependent knowing that no procedure can capture because it lives in the situation rather than above it.

It is arational, which is not the same as irrational. The judgment is sound; it simply does not arrive through explicit steps. It is earned through stages of development, from the beginner who clings to rules toward the master who has absorbed so many particular cases that the general sense of a domain becomes second nature. At that point the person navigates a whole body of knowledge by feel, exercising aesthetic judgment—recognizing what fits, what is off, what matters—without decomposing the decision into parts.

The cost of this is that it cannot be rushed or downloaded, and the person who has it often cannot fully articulate it. The gift is perspective: the mastered domain gives you a place to stand and a way of seeing that outsiders, however clever, do not possess. Connoisseurship is not decoration on top of competence. It is the form competence takes when it has gone all the way in.

How to

1. Immerse yourself in the primary material of your domain—the objects, texts, cases, or markets themselves—not summaries of them.
2. Seek out and study exemplars of excellence and failure side by side until you can feel the difference before you can articulate it.
3. Apprentice yourself to a recognized connoisseur and pay attention to what they attend to, not only to their conclusions.

Watch out for

• Rushing to the confident-expert stage by memorizing frameworks while skipping the years of situated practice that produce real judgment.
• Dismissing intuitions you cannot yet justify—early mastery often knows more than it can explain.

You know it by the reaction it produces. Someone hears the observation and says, quietly, that is so true. Not that it is clever or surprising, but that it names something they already lived without being able to say. Cultural insight is a truthful, contextual understanding of a shared world—an account of what matters to people and why they act as they do—that carries explanatory power rather than mere description.

It is not a fact you can verify in isolation, because it belongs to a context and dissolves outside it. It reveals the structure of meaning inside a particular human world: the reasons behind the behavior, the values under the choices. And it has to explain, not just record. A tally of what people do is not insight; the why that makes the tally make sense is.

This kind of understanding is built, and the inputs are demanding. It comes from analytical empathy that reconstructs a worldview from within, from the receptivity that lets the truth arrive unforced, and from abductive reasoning that leaps to the best explanation of rich material. Each supplies something the others cannot.

There is a rival way of seeing that erodes it. The reductive habit of treating human worlds as data points to be optimized—counting behavior while ignoring meaning—produces confident conclusions that miss the point entirely. Where that habit dominates, insight thins out. What genuine cultural insight finally offers is perspective: a standpoint from which the world becomes legible in a way no aggregate of numbers can match.

How to

1. State your candidate insight as a claim about what matters to people and why, then test whether it explains behaviors you didn't originally set out to explain.
2. Bring it back to the people it describes and watch for recognition, not politeness.
3. Anchor every insight in the specific context that produced it; strip away the context and you have a platitude.

Watch out for

• Confusing a vivid quote or surprising statistic with an insight—data points describe, insights explain.
• Generalizing an insight beyond the culture it came from until it becomes universally true and therefore useless.

A perspective is what lets you decide what matters. Not the data itself, but the finely honed point of view that selects context, determines relevance, and interprets how the scattered facts fit together into a portrait you can actually act on. Neil deGrasse Tyson put the limit of the alternative plainly: "In science, when human behavior enters the equation, things go nonlinear. That's why physics is easy and sociology is hard." Hard data explains material nature well. It does not explain us. And no quantity of brain scans or market segments will yield an insight if you have no perspective on the human behavior underneath them.

The way you earn a perspective is by imagining other worlds. If you work for a pharmaceutical firm, you need to understand the world of a person with diabetes, or your drug development fails. If you make cars, you need to know what life is like for a driver in western China, or you fill your vehicles with features nobody wants in the world's largest car market. This is what cultural engagement trains—the humanities teach you to inhabit a world not your own. The payoff runs in an unexpected direction: when you can fully imagine someone else's world, you develop a sharper perspective on your own. You start to see when the models and the financial innovations are drifting away from the truth.

That is where the real value sits. A perspective distills patterns from both scientific fact and practical reality, from what exists now and what might come next. It navigates by a North Star rather than by whatever the dashboard happens to show. And a perspective, over any meaningful stretch of time, proves more profitable—to your bank account and your life—than staying confined inside a cage of data. Human intelligence is the only intelligence that can cultivate one. That is not sentiment. It is the precondition for making sense of a nonlinear world.

How to

1. Articulate your North Star explicitly—what you believe matters most—so you can consciously use it to select and weight information.
2. When drowning in data, ask what portrait it composes rather than what each point says in isolation.
3. Pressure-test your perspective against the cultural insight it should be built on, revising it when the two diverge.

Watch out for

• Mistaking a strong opinion held early for a honed perspective earned through insight and mastery.
• Letting perspective calcify into dogma so it filters out the disconfirming data that should update it.

From the penthouse office at Ford headquarters, Mark Fields looks out over a campus so large that visitors need a car to reach their meetings. From that singular vantage point he makes decisions affecting hundreds of thousands of employees and more than 150 billion dollars in annual revenue. His problem is not intelligence or experience—he ran Mazda, he has decades in the industry. His problem is his vantage point itself. Like most CEOs, he is protected from the world by official and unofficial layers of people. Employees spend months rehearsing a one-hour meeting with him. Some gloss over issues to avoid being the bearer of bad news; others simply edit for efficiency. With every edit, Fields loses touch with the human intelligence he needs.

For years his intuition carried him, because Ford's world overlapped with his own. What matters to engineers in Detroit overlaps with what matters to a Texan buying a truck. It does not overlap with a young creative in Shanghai, or an entrepreneur in Chennai seeking space for reflection in a grinding schedule. Ford invested in Lane Assist for cities without demarcated lanes, and explored driverless cars for markets like New Delhi, where employing a driver signals status. The high-quality features became irrelevant, even detriments. Executives in this position describe the experience the same way: "I've lost my intuition." They all want it back quickly.

The reason wise decisions are hard is that business is almost always a bet on human behavior—which product sells, which employee succeeds, what price a customer will pay. Companies that excel at these bets thrive; companies that ignore the thick data of how people actually live are working from a faulty model of humanity. Getting the intuition back is not a matter of more analytics. It comes from rigorous engagement with the worlds those bets depend on.

How to

1. Trace each major decision back to the specific cultural insight and perspective that justify it, and be able to name them.
2. Distinguish decisions that require courage from those that require more analysis, and stop gathering data once further study only postpones commitment.
3. Judge outcomes over the relevant time horizon—an enduring product or resolved negotiation—not by immediate reception.

Watch out for

• Using analysis as an anxiety-management ritual that delays the decision indefinitely.
• Attributing a good outcome to the last-mile decision while ignoring whether the underlying insight was sound or lucky.

Reading a culture's seminal work is not the same as skimming a summary of it, and the difference shows up later, when you are trying to make sense of people whose world is not your own. Deep engagement with texts, art, music, history, and languages builds something that market research cannot manufacture: a felt understanding of how other people hold their reality together. You do not acquire this by sampling. You acquire it by staying with difficult material long enough for it to change what you notice.

The pattern is one of earned familiarity. When you have spent years inside a culture's stories and quarrels and forms of beauty, you begin to recognize its meaningful distinctions from the inside—what counts as sacred, what counts as funny, what counts as an insult. That recognition is the raw material of analytical empathy. A person who has only glanced at a culture can describe its behaviors; a person who has lived in its texts can interpret them.

There is a second effect, quieter and harder to defend on a spreadsheet. Sustained immersion cultivates receptivity—a willingness to be met by something you did not construct in advance. The abstraction of a focus group flattens this out, because it asks people to report on themselves in a setting stripped of their world. The demanding path keeps the world attached.

The cost is real: this kind of engagement is slow, and it resists proof. You cannot show, in a quarterly review, that reading the whole tradition rather than the executive summary made you wiser. What you can say is that the difference is not decorative. It is the difference between knowing about a people and understanding them.

How to

1. Choose primary sources—the actual novels, liturgies, foundational documents, and vernacular practices of a culture—over summaries, dashboards, or persona decks.
2. Read in the culture's own languages and idioms where possible, or work with someone who does, since translation strips the ambiguities that carry meaning.
3. Spend time inside lived practice—markets, meals, rituals, ordinary transactions—not just archives, so you feel how the texts operate in daily life.
4. Return to the same works after your project questions have shifted, treating them as instruments that reveal different things at different depths.

Watch out for

• Do not mistake fluency in a culture's clichés for engagement with it; recognizing surface signifiers is exactly the superficial exposure this construct is defined against.
• Resist the pull to convert everything you observe into abstracted variables too early—market-research categories destroy the particularity that made the immersion worth undertaking.

Thin data records what people do; thick data records what it means to them. A log of clicks tells you that a hand moved. It says nothing about the mood behind the movement, the story the person would tell about it, the gesture that accompanied it, the smell of the room where it happened. Thick data collection is the patient gathering of exactly those things—the contextual, narrative, sensory detail that carries significance rather than mere fact.

The reason to bother is that meaning does not survive abstraction. Strip a behavior of its context and you are left with a trace that can be counted but not interpreted. Keep the context, and you have material that lets you reason toward an explanation. This is why thick data feeds abductive reasoning: you cannot form a plausible account of why people act as they do without the texture that suggests the account in the first place. The story is not decoration on the number; the story is the evidence.

It also feeds analytical empathy, because thick data is empathy in collected form. Recording a person's mood, their offhand story, the way they hold a tool, is already an act of taking their world seriously as a world. The detail you preserve is the detail you will later be able to feel your way into.

The discipline is knowing what to keep. Everything is context; not everything matters. Thick data is not a promise to record more—it is a commitment to record what expresses what matters, and to resist the comfort of the tidy metric that leaves the meaning behind.

How to

1. Observe people in their own worlds—homes, shop floors, waiting rooms—rather than pulling them into your setting where their behavior is already distorted.
2. Record the sensory and emotional texture: what people touch, hesitate over, complain about, and work around, not just what they choose.
3. Capture the stories people tell to explain their own actions, treating their interpretations as data rather than noise to be cleaned.
4. Write field notes that preserve ambiguity and contradiction instead of resolving them prematurely into categories.

Watch out for

• Don't convert thick observations into tidy quantitative codes too early—the moment you tally gestures into frequencies you have thrown away the context that made them thick.
• Beware confirmation staging: arranging observations to illustrate a hypothesis you already hold, which turns fieldwork into set decoration for a predetermined conclusion.

People behave differently when they are watched in the settings they actually inhabit than when they are asked, in a conference room, to account for themselves. The savannah is the phrase for where behavior really lives: the street, the kitchen, the shop floor, observed at eye level, in its own social context, on its own terms. Surveys, focus groups, and the view from the executive suite all share a defect. They remove people from the situation that gives their actions sense and then ask them to reconstruct it from memory or imagination.

Eye level is not a metaphor for humility, though it is that too. It is a methodological commitment. To study behavior where it occurs is to accept that the researcher must go to the phenomenon rather than summon a decontextualized version of it into a controlled room. What you lose in tidiness you gain in truth, because the context you refused to strip away is exactly the context that produces analytical empathy—the ability to understand why an action makes sense to the person performing it.

There is a force working against this, and it is worth naming plainly. A mindset that trusts big data and frictionless optimization treats direct observation as slow, unscalable, and anecdotal. Under its influence, immersion gets cut first, because it cannot be automated and does not fit a dashboard. The savannah loses out to the abstraction that is easier to buy.

The recognition is uncomfortable for anyone managing a budget: the cheapest data to collect is often the data least connected to how people actually live. Immersion costs time and attention that no shortcut replaces.

How to

1. Go to the actual setting — the shop floor, the living room, the checkout line — and observe behavior in its native context before you ask a single question.
2. Position yourself at the participant's level, not the manager's: study the phenomenon as the person living it experiences it, not as the org chart frames it.
3. Record what people do alongside what they say, and treat the gap between the two as your primary finding, not noise to be cleaned.

Watch out for

• Do not mistake site visits staged and narrated by an executive host for immersion — you are then studying the company's self-image, not the world.
• Avoid arriving with a coding scheme that lets you only see what you already expected; premature categories quietly convert immersion back into abstraction.

You can only see the distinctions that matter to you. A sommelier tastes differences in a wine that another person, drinking the same glass, cannot detect—not because the sommelier's tongue is built differently, but because the wine matters, and mattering sharpens perception. This is care in the sense meant here: not kindness, but the disposition in which something is genuinely significant to you. Where nothing matters, everything looks the same.

Care is what makes aesthetic judgment possible. To say that one thing is finer, truer, or more apt than another, you need a stake in the difference. Absent that stake, judgment collapses into indifference or arbitrary preference. The person who cares can hold the distinction and defend it; the person who does not simply shrugs. This is why care functions as the connective tissue of interpretation—it is what binds a set of observations into a reading rather than a list.

Its most important work is on perspective. Perspective is not just where you stand; it is what you are able to make of what you see from there, and care determines how much that vantage yields. Two people can occupy the same position and come away with wildly different understanding, because one is invested in the meaning of what is in front of them and the other is merely present. Care moderates the whole relationship: it decides whether a perspective becomes insight or stays inert.

The practical implication is that you cannot outsource caring, and you cannot fake it. Interpretation without care produces the motions of analysis with none of the seeing.

How to

1. Choose problems and populations you have a genuine stake in understanding, not merely ones assigned or profitable.
2. Before analyzing, articulate what specifically matters to the people you study and why it would matter to you in their position.
3. Cultivate aesthetic judgment by immersing in the objects of the world—texts, spaces, rituals—until you can feel their weight, not just describe them.

Watch out for

• Confusing care with sentimentality or advocacy, which distorts what you see rather than sharpening it.
• Manufacturing false enthusiasm for a subject you are indifferent to—it produces conspicuously hollow interpretation.

A particular state of mind now sets the terms for what counts as knowledge. It trusts the objective and the quantitative, celebrates disruption, and treats friction as a defect to be optimized away. Big data is its native evidence; the smooth, scalable, measurable solution is its idea of progress. Within this mindset, cultural and humanistic knowing register as soft—interesting, perhaps, but not serious, because they cannot be reduced to a number or run at speed.

The consequence for understanding people is direct. This way of thinking devalues exactly the methods that produce cultural insight, because those methods are slow, contextual, and resistant to abstraction. Insight into a shared human world does not scale the way a dataset does, so it gets treated as a luxury rather than a foundation. The mindset does not merely ignore cultural knowing; it actively crowds it out, redirecting attention and money toward what can be counted.

It also works on real-world immersion, and the effect is corrosive. Where the impulse is to optimize and quantify, direct observation in natural settings looks like an inefficiency. Why send someone to watch people on the street when the traces are already in the database? The mindset moderates immersion downward, substituting the abstracted proxy for the thing itself, and then mistaking the proxy for the whole.

What makes the condition hard to resist is that it is not stupid. Quantitative knowledge is genuinely powerful, and optimization genuinely useful. The failure is one of monopoly—when the objective and the frictionless become the only admissible forms of knowing, the world of meaning goes dark, and no one notices what was lost because it was never counted.

How to

1. Audit your questions for a quantification bias—ask what you are ignoring because it resists a metric.
2. Treat 'disruption' and 'frictionless optimization' as contestable values, not neutral goods, and ask whom the friction protected.
3. Pair every dashboard with at least one interpretive question no dataset can answer on its own.

Watch out for

• Letting the availability of data dictate which problems you consider real—the streetlight effect institutionalized.
• Dismissing cultural knowledge as 'soft' because it does not scale, when its non-scalability is precisely its analytic value.