The Useful Box and the Bengali Problem of Too Many Choices
Acronyms and terms used:
AI: Artificial Intelligence, software systems that perform tasks usually associated with human intelligence, such as prediction, classification, summarization, or decision support.
EHR: Electronic Health Record, the clinical software system where patient care is documented, ordered, reviewed, billed, reported, and often quietly made more complicated than anyone planned.
FHIR: Fast Healthcare Interoperability Resources, a modern healthcare data exchange standard that tries to make clinical information easier to move between systems.
HL7 v2: Health Level Seven version 2, an older but still heavily used healthcare messaging standard that moves clinical events such as admissions, lab results, orders, and discharges.
SQL: Structured Query Language, the common language used to query and manage data in relational databases.
UX: User Experience, the practical feel of using a product, form, website, device, or system, including whether it behaves like a helpful tool or a sulking government clerk.
Too much freedom is not freedom. It is a room full of open cupboards.
You know the room. One cupboard has old tax papers. One has three phone chargers, none of which fit anything currently alive. One has a plastic box containing another plastic box, because Bengali civilization rests on the sacred principle that no container must ever be thrown away. Now imagine your work life like that. Your phone is buzzing. Your inbox is breeding. Your browser has twenty-three tabs open. Your to-do list has become a municipal ward office. Everybody has a demand. Nobody has a plan.
This is where the romance of freedom begins to look slightly damp.
We are taught to love possibility. More choice. More options. More apps. More features. More money if possible, but let us not become comedic. The modern promise is that if we remove enough boundaries, a better version of ourselves will walk out, bright-eyed, efficient, flexible, and probably drinking green tea from a stone-colored mug.
In practice, many of us remove the boundaries and find only a tired man in a faded T-shirt wondering whether brushing teeth counts as momentum.
This is not a moral failure. It is a design problem.
A human brain is not a vast marble palace of reason. It is more like a crowded suburban train trying to remember where it kept the ticket. It can do astonishing things, yes. It can invent calculus, compose music, build bridges, debug software at 2:17 in the morning, and remember the exact insult from Class VII that should have been forgotten around the time Doordarshan still had dignity. But it cannot handle infinite open loops without becoming noisy.
Every unfinished thing leaves a hook in the mind. Reply to that email. Check that bill. Fix that sentence. Apply for that job. Call the dentist. Update the website. Make tea. No, first wash the cup. No, first buy milk. No, first decide whether life has any remaining administrative meaning.
And then the afternoon is gone.
The strange truth is that a good constraint can feel like rescue. Not because life becomes easy, but because life stops arriving as a swarm.
One task. One hour. One page. One walk to the shop. One problem actually named. There is mercy in smallness when the alternative is being beaten by infinity with a wet towel.
But here is the catch, and it is a large one, wearing shoes.
Not every constraint is good. Poverty is a constraint. Illness is a constraint. Heat in Calcutta in May, when the ceiling fan rotates like a philosophical argument and achieves roughly the same cooling effect, is a constraint. Joblessness is a constraint. Rent is a constraint. A collapsing public system is a constraint. These do not automatically make people creative. They often make people exhausted, frightened, ill, and less able to think.
So let us not sell hardship as wisdom. That is a nasty habit of comfortable people.
A useful constraint focuses capacity. A cruel constraint removes capacity.
A poet choosing fourteen lines has a useful constraint. A hawker being bulldozed without livelihood support has a cruel one. A writer deciding to use plain language has a useful constraint. A patient standing in a hospital queue because three counters do the work of one functioning system has a cruel one. A product team choosing three features and doing them well has a useful constraint. A family cutting protein because prices have gone up has a cruel one.
The difference is not decorative. It is the whole argument.
Good constraints are chosen, visible, humane, and revisable. Bad constraints are imposed, hidden, punitive, and defended by people who do not pay their cost.
This matters in ordinary life. It matters in writing. It matters in software. It matters in healthcare. It matters in government. It matters in the tiny empire of your own desk.
Take the to-do list, that innocent-looking snake.
Most people keep adding to it until it becomes less a list and more a museum of personal disappointment. Buy medicine. Renew domain. Write post. Clean room. Fix resume. Call mother. Read article. Watch lecture. Exercise. Learn AI. Sleep properly. Become financially stable. Repair entire personality before lunch.
Then the list becomes unbearable, so we ignore it. Ignoring it produces guilt. Guilt produces more list-making. More list-making produces a new notebook. The new notebook produces hope for three days. Then it joins the old notebooks, like retired ministers in a dusty drawing room.
The solution is not a better notebook.
The solution is subtraction.
Not glamorous. Not heroic. Not the sort of thing that looks impressive on LinkedIn, where everyone is “excited to announce” something every four minutes, as if professional life were a permanent drumroll. But subtraction works because it asks the adult question: what can be removed so the important thing has room to breathe?
One meaningful task at the top of the day. Not seventeen. One. If that is done, the day has a spine. Everything else is furniture.
This is not laziness. It is respect for the bottleneck.
In most systems, the bottleneck is not where people think it is. A factory may buy more machines when the real bottleneck is design approval. A hospital may buy a new analytics tool when the real bottleneck is bad coding discipline. A writer may buy a new laptop when the real bottleneck is fear. A middle-aged man in the southern fringe of Calcutta may blame his entire life, not without evidence, but the immediate bottleneck at 9:10 in the morning may still be that he has not made tea and opened the document.
Reality is often humiliatingly specific.
I learned this the hard way in healthcare IT. Systems rarely fail because nobody added enough things. They fail because everyone added too many things without deciding what anything meant. One extra field. One exception. One interface. One report. One dashboard. One “temporary” workaround that becomes permanent because the person who knew why it existed left in 2018 and now the organization treats it like an ancestral deity.
After a decade, the system is no longer designed. It has accreted. Like limescale. Like bureaucracy. Like WhatsApp misinformation during election season.
And when the data becomes confusing, someone says, “We have a data quality problem.”
Sometimes yes. Often no.
Often the deeper problem is representation. The system never agreed on what the data meant. It moved the message but lost the meaning. It transmitted the code but not the clinical intent. It stored the date but not the timing logic. It captured the value but not the context. The pipe worked. The sense leaked out.
This is one of the most important distinctions in modern technical life: transport is not meaning.
A parcel can reach the correct address and still contain the wrong medicine. A train can arrive on time and still go to the wrong city. A message can move perfectly between systems and still be useless because nobody agreed what the message meant when it left home.
FHIR cannot magically fix that. HL7 v2 never could. AI certainly will not. AI sitting on top of confused data is not intelligence. It is a parrot in a smoky room, repeating things with confidence because confidence is cheaper than truth.
The same thing happens in daily life. We think we need more information, but often we need a better question.
“What should I do with my life?” is too large. It is a sky-sized question. The brain looks at it and quietly leaves the building.
“What one thing would make tomorrow less rotten?” is smaller. Therefore better.
“Why am I failing?” is a courtroom.
“What is the bottleneck?” is a workshop.
See the difference? One produces shame. The other produces a handle.
This is why problem definition is underrated. People worship ideas. They should spend more time respecting good problems. A good problem is a box with air holes. It contains the chaos without killing the creature inside.
The great mistake is to think constraint means lack of imagination. Often it is the opposite. Constraint is what prevents imagination from becoming a fish market at 8 a.m.—alive, loud, slippery, full of smell, and impossible to invoice.
A children’s book with only fifty words can become unforgettable because the limit forces rhythm. A song with three chords can carry more feeling than a musical palace with marble staircases. A small kitchen can produce better food than a giant buffet where every dish tastes faintly of steam and resignation. Anyone who has eaten a proper egg roll from a modest stall knows this. The magic is not infinite choice. The magic is hot paratha, egg, onion, chili, lime, paper wrap, and no lecture from a consultant.
The boundary makes the thing.
But boundaries must not become cages. This is where nuance is needed, otherwise the argument turns into management wallpaper.
A student from a poor family does not need someone saying, “Constraints will make you creative.” He may need money, sleep, safety, a quiet room, and a functioning internet connection. A depressed person does not need a productivity sermon. He may need treatment, patience, food, sunlight if the weather allows it, and one task small enough not to look like a mountain wearing a police uniform.
A city does not become disciplined by attacking only the powerless. It becomes humane by designing workable rules, enforcing them fairly, and offering transitions that do not turn survival into trespass.
So yes, constraints help.
But only the right kind.
The useful box is not the one somebody locks you in. It is the one you draw around the work so the work does not spill everywhere and drown you.
For organizations, that means fewer vague initiatives and more named problems. Not “digital transformation,” which is usually a fog machine wearing a blazer. Say instead: discharge summaries must reach the next clinician within six hours, medication changes must be understandable, and every data element must have an owner. Now we can work.
For teams, it means stop starting and start finishing. Work in progress is not productivity. It is inventory with anxiety. Every half-started thing steals attention from every other half-started thing until the whole team is running like a ceiling fan with a bent blade.
For personal life, it means designing a day that does not require a mythical version of you to appear. Do not plan for the heroic self. Plan for the actual self: underslept, mildly suspicious, financially worried, easily distracted, still capable of one decent act if the door is not made too heavy.
Make the door lighter.
Put the phone in another room for thirty minutes. Write the one sentence. Pay the one bill. Remove one useless obligation. Close three tabs. Say no to one decorative meeting. Eat something with protein. Walk to the corner if the heat has not turned the road into a frying pan. Return. Continue.
This sounds small because it is small.
Small is where the machinery catches.
The modern world keeps telling us to expand. Expand your network. Expand your skills. Expand your reach. Expand your brand. Expand your breakfast into a lifestyle platform. But sometimes the humane thing is to contract. Narrow the day. Name the bottleneck. Remove one dead process. Choose the next visible step.
A life cannot be optimized like a spreadsheet. A mind cannot be treated like cloud storage. A city cannot be run as if poor people are formatting errors. A system cannot be rescued by adding features to a bad definition.
The useful question is not, “How do I do everything?”
Nobody does everything. The people who appear to do everything are either lying, delegating, inheriting money, or silently collapsing in expensive shoes.
The useful question is: “What must not be allowed to swallow the day?”
Answer that, and you have the beginning of a box.
Not a prison.
A box with a window. A box with a chair. A box where the work can sit down, stop shouting, and finally become visible.
P.S. References: David Epstein, Inside the Box.