Become Undefinable

Dear Rajdeep,

A category is a competition. When someone can put a label on you, they can compete with you, because the label names everyone else they can call instead. You keep asking which thread is the real path: the code, the markets, the autobiographies, the systems thinking, the philosophy, or the mountains. The question assumes one of them is the work and the rest are the delay. That assumption is the expensive part.

The pull to give one word as the answer is strong. One word is easy to say at a table full of strangers, and it ends the question. It also hands over the list of people who can be called instead of you. Whoever hears the word compares you against that list before reading a line of your work.

The other side of this is quiet. When people struggle to explain what you do, they have nothing to line you up against, so the comparing stops. You become the only person they can think of for the problem in front of them. An only option gets priced. A member of a category gets compared.

Say AI engineer out loud and ten thousand people stand in that box with you. A buyer compares rates, credentials, and output, and the whole comparison takes an afternoon. None of that afternoon is about you. The label turned you into a commodity before the first call, and buyers price a commodity at the cheapest version they can accept.

The number of people who can hold all of it at once sets your price. The letter on the second door made this point from your side, where a second answer is what makes the first one negotiable. That same count runs in reverse when someone buys. A person who needs one skill chooses from thousands and pays the floor. A person who needs three of yours at once chooses from a list short enough to read out loud, and that list has no floor.

A good title is expensive because it works. It pays at once, and it tells people which problems to bring you. It also tells them which problems to bring someone else. The narrowing stays invisible, because every piece of work that arrives is work you can do. Give it long enough and everything that reaches you has the shape of that one word.

Scott Adams built Dilbert out of three skills he never mastered. He was mediocre at art, mediocre at prose, and mediocre at business. One of them alone gave him nothing. Better artists existed everywhere and so did better writers, and neither group can produce that strip alone. Together the three had no substitute, because nobody else stood at that crossing holding all of them.

That combination is a talent stack, and multiplication is the right word for what it does. Every skill you add removes a person you have to convince, along with the delay and the loss that come with convincing them. Each one also changes what the others can see. A person who writes code and reads markets sees a product where a specialist sees a ticket. The multiplying happens when the skills aim at one output, because the handoffs between them disappear. Skills that aim at different outputs stay a list.

The stack has a floor. Each skill has to be good enough to carry weight alone, which means good enough that someone pays for it or trusts it. Below that line the skill is a hobby, and hobbies do not multiply. The test is whether someone can hand you a problem inside that skill and walk away. Run it on each thread and be honest about which ones fail.

Name the stack you already have. Code lets you build the thing you imagined without asking permission. Market knowledge shows you where value moves and why it moves there. Trading taught position size, and it taught the discipline to be wrong fast without arguing about it. Both of them show the world below the surface, where incentives explain behavior that looks irrational from above.

Autobiographies show you how people decided under pressure, which is the only part of a life story you can use. Systems thinking gives you feedback loops and second-order effects, so a symptom stops looking like a cause. Philosophy settles what matters before the week settles it for you. The mountains gave you patience and humility, and they charged full price for both.

As a career plan, that list is a mess. As a life, it is coherent. A career plan bets that the shape of the work holds still. A life bets on the person doing the work, and the person is the only asset that moves when the work does.

Name the cost. Every hour on a biography or a ridge is an hour a specialist spends going deeper. They will be deeper, permanently, and no amount of range closes that gap. You buy range with depth, and the exchange rate is bad for a long time. It turns only when a problem arrives that needs more than one of your skills at once.

A specialist beats you inside their lane, and that never stops being true. Accept it early and it costs you nothing. A specialist cannot stand where two lanes cross, because the crossing needs competence in both and comfort in neither. The people who can stand there are the ones who kept both lanes alive while everyone told them to pick one.

People pay a premium for a connection they cannot make themselves. Someone deep inside one field cannot see the field next door, so a problem sitting between two fields stays open. It stays open because it does not look like a problem in either field until one person holds both views at once. When you walk past it, nobody else is in the room.

A stack pays nothing until one person can describe a result you produced. The letter that put your ceiling at how many people will vouch for you applies here without changes. Someone holding a problem with no obvious owner calls the person they can describe in one specific sentence. Admiration without a sentence produces no call. A result nobody can point at pays the same as no result at all. Finish things where other people can see them, and let someone else name what the finished thing did.

Best is the wrong measure at a crossing. Good enough at several skills, put together, answers a problem that has no name yet, and the buyer of that answer has no second candidate. Go deeper in every thread you keep. A thin thread breaks under the first real problem, and it takes the combination down with it.

You did Goechala. You saw Kanchenjunga at sunrise, and you know what scale does to a person standing in front of it. The mountains taught the part of judgment that a book cannot reach, because a book never makes you tired and responsible at the same time.

Long approaches taught patience, because the walking pays nothing for days and the day at the top depends on all of it. The early mornings taught discipline in the form that counts, the kind with nobody watching and the cold arguing against you. Turning back teaches judgment, and judgment is the expensive one. It is the ability to see that the correct decision and the disappointing decision are the same decision.

The descent is the goal. A summit with a bad descent is an accident that has not finished happening. The mountain has no interest in your title. Nothing you can say to it changes the weather or the time it takes to walk down. It asks whether you did the work, and whether you can go down when going down is correct.

Those rules moved into the rest of the work without asking. You size risk the way you size a weather window. You hold still when a plan collapses, because on a ridge panic costs more than the collapsed plan did. Keep climbing. Nothing else on your list trains that.

Life is scored like a decathlon, and the events arrive in an order nobody publishes. The powerlifter outlifts you and the marathoner outruns you, and neither picks the event. Life tests building, communicating, enduring, adapting, deciding under pressure, recovering from failure, and sitting with uncertainty. It never tests one of them at a time. A decathlon is scored across all the events, so a maximum in one and a blank in the rest loses to a decent score everywhere.

A specialist optimizes for one event, wins it, and loses the decathlon. Optimize for the hopper instead, which means being ready for the event nobody announced. You cannot train for an event you cannot name. So you train the parts every event shares, which are judgment under load and the speed of recovery. When one door closes, three others open, because you hold keys to all of them. When the test is unknown, the generalist wins.

The pairs arrive as forced choices: money or meaning, engineer or trader, builder or reader. Each pair assumes one person cannot hold both at full weight. The two halves compete only if they need the same hour, and most of these pairs do not. Hold both anyway, and the pair stops being a choice. That position stays empty, because everyone who accepted the choice walked away from it.

Build money and give a real part of it away. Write code and study markets. Count a win and stay honest about how much of it came from a moment you did not choose. Holding both costs more time than holding either half, and it removes the limit that choosing puts on you.

One skill breaks in a specific order. The domain shrinks first, and the shrinking is slow enough that nobody inside notices. The first visible sign is a drop in the number of calls. Then a machine does the skill, or the market prices it as a commodity. Nothing stands beside it, so the person goes down with the domain.

I watched this happen to people who were better than me at the one thing. Their skill became a feature inside a product owned by somebody else, and the feature was free. They had nothing beside it to move to, because every year of depth went into the same hole. They optimized so hard for the one thing that adapting was never a skill they practiced. Depth is an asset until the ground moves.

Reading across fields is where the second kind of expertise comes from. A person who reads inside one domain becomes an expert in a domain that can shrink. A person who reads across domains becomes an expert in connection, which is the skill of noticing that a problem here already got solved there. That skill has no title, and it survives the shifts that kill the titled ones.

A stack fails differently. One part stops paying, the other parts keep running, and the only question left is which new pair to try. That question is small, because recombining uses parts you already own. Rebuilding a person around a new single skill is not small, and the rebuild always starts on the worst day available.

The pressure to pick a lane never stops, and it gets louder as things go well. People ask what it is you do, and most of them mean it kindly. Some of them will call you unfocused and point at a specialist who dominated one thing. A clean answer is what they need to file you, and that need is their constraint. Do not let their filing system set your size.

They are not wrong about the specialists they point to. Some specialists win, and they win a real game with real rewards. That game is scored on one axis, and the entry fee is every other thing you were curious about. Pay it if the one thing is what you want.

The guilt that arrives when you open an autobiography instead of a technical manual is that same pressure in your own voice. Guilt carries no information about which book was worth the hour. A month later, name the decision the book changed and name what the manual changed. When someone asks what you do, the true answer is that you solve problems that do not have names yet.

Can another person replace you? If a stranger describes what you do in one phrase, that phrase works as a search query and it returns other people. Then watch what happens when someone tries to explain you to a third person. The stumble is the signal, and the confusion is the moat. When nobody says they have never met anyone like you, you are easy to file again.

The other two are harder. Name the last time you did something that did not fit, and put a date on it. If everything you do makes sense to everyone, you built yourself to be understood, and growth is what that cost. Then ask whether you follow your own curiosity or a plan somebody else wrote. The plan is useful. Curiosity decides where you end up, and when the two conflict, curiosity wins.

Confidence here is evidence. You showed up daily and you stacked the skills, and both of those are countable. Check the record before you check the feeling. The feeling reports this week and nothing else, and this week is the least reliable sample you own.

Keep the record, because dots with no unit stay a feeling. Write down what you read and what it changed about a decision you made later. One line each is enough, and the line about the change is the part that matters. The first of these letters handed you a count of consecutive days, and the record is that instrument aimed at a different target. When someone asks how you got here, the record answers and memory does not.

This week, start one thing that does not fit. Choose the one that makes people ask what it is for. Pick it because it pulls at you and belongs on no plan you wrote. It will feel like a distraction while you do it, and that feeling is not evidence. Before you start, write the cost on one page. The letter on taking the risk while the door is open gave you that page, and a written cost stops being a fear you carry.

Give it real time, which means enough hours for it to produce a fact instead of an opinion. See what it connects to. Next month, start another one. Then count the people who can do all of it at once and write the number down. If you can name five, you are not stacked enough yet. Add one more thread and count again.