After my last essay, Dhilbar sent me Richard Hamming's 1986 talk, You and Your Research, and told me to read it. I finished it on a Wednesday night and spent a few hours the next day writing about it. This is that.
The famous lines from this talk are everywhere on the internet. I had seen most of them before, the way you see quotes from a movie you have not watched. So I expected the full thing to feel like a greatest hits album. It did not. The actual talk is weirder, more uncomfortable, and a lot more impatient than the quotes make it look. Hamming is not selling you a system. He is an old engineer who has watched smart people waste their lives for forty years and he is slightly annoyed with you for being about to do the same thing.
He starts with the question that bothered him for forty years. Why do so few brilliant people, in the same building, end up doing anything that lasts? He says he became obsessed with this at Los Alamos, watching Feynman and Fermi and Bethe, and realizing that physically he was the same kind of human as they were and yet they were doing something he was not. He calls it envy. He says it bluntly. He wanted to know why they were different from him.
That alone is a strange way to start a talk. Most adults will not admit to envy on stage. Hamming uses it as the first piece of evidence. The reason he has spent forty years studying great work is that he was once, in a small office at Los Alamos, jealous of people he thought were no smarter than him. The whole talk runs on that engine.
And his answer, when he finally gives it, is not what you would expect from someone who has watched Nobel winners up close. He says it is not about brains. He thinks most people in the room have enough brains. He says it is not really about luck either, although luck is real. The thing he keeps coming back to is courage, and self-management, and a specific kind of taste that lets you tell the difference between a problem worth bleeding for and a problem that just sounds important.
The first line that actually got me was about importance. I have heard work on important problems a thousand times. It sounds like a fortune cookie. But Hamming gets specific in a way the people quoting him never do.
It's not the consequence that makes a problem important, it is that you have a reasonable attack. Richard Hamming
That second clause does all the work. He says the three biggest problems in physics, the ones that would guarantee a Nobel and any sum of money you wanted to name, were never worked on at Bell Labs. Time travel. Teleportation. Antigravity. Not because Bell Labs was unambitious. Because they had no attack. No first move that was not obvious nonsense or a thirty year fishing trip.
When I read that, I had to put the talk down. Because most of how I have been thinking about ambition is wrong in exactly this way. I pick topics by how impressive the noun sounds. AI. Agents. Memory. AGI. The grammar of looking serious, at least online, is to stand next to the biggest word in the room. The whole point is to be seen near it. Whether you have any move on it almost does not enter the conversation.
What Hamming is saying is that importance is not a property of the problem. It is a property of the relationship between you and the problem. A problem is important for you if you have an angle on it that nobody else does, or a tool nobody else has bothered to pick up, or a piece of taste that lets you see a small move that turns into a big one. The corollary is brutal. A lot of the things people sound ambitious about are not, by his definition, important to them at all. They are just impressive. The two are not the same word.
I think this is the single most useful idea in the talk for someone my age. At 16, almost nobody around me has any real attack on anything yet. We are all standing next to large nouns. The work, the actual work, is to find one tiny corner of one of those nouns where you, specifically, can do something that someone older cannot. That corner almost always looks smaller from the outside than the noun does. That is fine. The size of the corner is not the point. The fact that it is yours is the point.
The part of the talk I keep returning to is the bit about the door. Hamming says he watched, over decades, the people who worked with their office doors closed and the people who worked with theirs open. The closed door people got more done that day. They were focused. They protected their time. They did not get interrupted. They were, in modern language, doing deep work.
And ten years later they had nothing to show for it.
He who works with the door open gets all kinds of interruptions, but he also occasionally gets clues as to what the world is and what might be important. Richard Hamming
The closed door person was more productive. The open door person became more important. There is a real gap between those two words and almost every system I have set up in my life is built to optimize the first one.
Notifications off. Notion locked down. Headphones on. Pomodoro timer running. I am, right now, very good at executing a list. The hours where I feel most productive are the hours where I am just clearing the queue. The hours where I do the thinking that actually changes my mind look completely different. Those hours are messy. I am in DMs. I am at lunch with someone whose work I do not fully understand. I come away with three new questions and nothing shipped.
Hamming would say the second kind of hour is what compounds. The first kind of hour just empties a queue that refills tomorrow. The hard part is that the second kind of hour feels, in the moment, like you are wasting time, and the first kind of hour feels good, and almost every adult around me will reward the first.
I do not have a clean answer for this. I know I cannot just flip a switch and become an open door person. The closed door is also doing something real. It is how things ship. What I can do is notice when I am hiding behind it. There is a version of focus that is just being scared of a wider conversation. That version does not compound. It just feels safe.
About halfway through, Hamming tells a story I have read three times now and will probably read again. He had been eating lunch at the chemistry table at Bell Labs. After a few weeks of getting comfortable, he started asking the chemists what they thought the most important problems in their field were. Then which of those problems they were working on. Then, finally:
If what you are doing is not important, and if you don't think it is going to lead to something important, why are you at Bell Labs working on it? Richard Hamming
He says he was not welcomed at that table after that.
One of the chemists came back to him months later and said the question had been crawling under his skin all summer. That person went on to head the department and become a member of the National Academy of Engineering. The other people at the lunch table, the ones who got annoyed and changed the subject, Hamming says he never heard their names again.
Sit with that for a second. This is a top tier research lab in its peak era. Every person at that lunch table had passed every filter on earth. They were brilliant. The difference between the one who became someone and the ones who disappeared, in Hamming's telling, was not IQ. It was whether they could tolerate the question being asked out loud.
I do not know yet which kind of person I am going to be. At 16 I am probably both, depending on the day. But the story made me think about the small social cost of asking the real question, and how casually I have been paying it not to ask. Most of the time I do not ask the question because I do not want to look like the kid who is too earnest. That is, on its own, a small price. But Hamming's whole point is that small prices, paid every day, become very large bills over a career.
The thing that surprised me most, and the part I would have missed if I had only seen the famous quotes, is how much of the talk is about self-management. Not science. Self-management. The word "manage" comes up over and over. He manages himself. He talks about how he managed his boss. He talks about how Bode managed him. He talks, repeatedly, about how Shannon failed to manage himself after information theory and basically stopped doing scientific work.
The story we get told about doing great work is romantic. Follow your passion. Trust the process. The work finds you. You are either built for it or you are not. Hamming's view is the opposite of all of that. He thinks great work is closer to engineering. You build a system around yourself, you tune it, you design around your specific faults.
He is vain, so he brags about unfinished books in public to force himself to finish them. He prefers the door closed but knows that costs him taste, so he forces it open. He notices that famous scientists tend to only want to work on big problems and stop planting new acorns, so he refuses to read papers in the field he is already known for. He is, in cold blood, building a machine that produces important work, and he himself is part of the machine.
Knowing many of my own faults, I manage myself. I have a lot of faults, so I've got a lot of problems, i.e. a lot of possibilities of management. Richard Hamming
That sentence is, in a way, the entire talk. Most people treat their faults as fixed weather. Something to apologize for, something they will fix one day, something that will go away once they grow up. Hamming treats them as inputs to a control system. Vanity, ego, laziness, FOMO, the desire to look smart at dinner. Each one is energy. If you point your vanity at finishing the project instead of looking clever in the group chat, your vanity works for you. If you turn your FOMO into deadlines you have already announced to people, FOMO works for you.
The trick is not to become a different kind of person. The trick is to know exactly what kind of person you are and arrange the room so that person ends up doing good work anyway. That is a very different posture from what I see in most teenage advice. The teenage version of self-help is almost always about becoming someone new. Hamming is saying the opposite. Get extremely specific about who you actually are, faults and all, and use the faults.
There is a section on compound interest that I underlined twice. Hamming says two people of equal ability, one of whom puts in ten percent more thinking per day, will more than double the other's output over a career. He believed this so strongly that he openly admits to neglecting his marriage to study, which is a wild thing to drop on a stage in 1986. He is a little sheepish about it. He still says it.
The compounding argument is correct, and it is also the place where I think Hamming and I are sitting in different worlds. He spent his career inside three or four fields. Numerical methods. Coding theory. Filters. Computing. The fields stayed put long enough for ten years of compounding to actually pay off. You could go deep in one direction and the direction would still be there at the end of it.
That is not the world I am stepping into. The half life of a specific technical skill in AI right now feels like eighteen months. Frameworks shift. Models shift. The right way to do retrieval changes twice a year. A lot of what looks like compounding turns out to be context dependent and quietly resets when the underlying tooling changes shape.
So I have been trying to figure out how to read Hamming in this world. I think the answer is to read him one level up. The thing that compounds is not the technique. It is the taste. The question asking. The ability to walk into a new domain and quickly notice what has an attack and what does not. Hamming himself shifted fields several times on purpose. He just did not have to do it as often as people my age will. What survives across regimes is the meta-skill, not the specific tool. The lunch table question travels. A good prompt template, probably, does not.
The part of the talk that I am still chewing on is the part about ambiguity. Hamming says great scientists tolerate ambiguity well. They believe their theory enough to keep going, and they doubt it enough to notice the flaws. If you believe too much, you never see what is wrong. If you doubt too much, you never start. Most people fail one of those two tests, almost always quietly, and almost always for years.
He says Darwin had to write down every piece of evidence that contradicted his beliefs, because otherwise his mind would lose track of them. Read that line again. The man who came up with one of the deepest theories in human history did not trust himself, by default, to remember the things that disproved his own theory. He had to keep a notebook of his own counterevidence. That is one of the most humbling sentences I have come across in any context. Most people I know, including me, do not have the basic honesty to write down the thing that contradicts what we want to be true.
Hamming's whole framing of greatness, when you stack the pieces, is something like this. You pick a problem you actually have an attack on. You commit hard enough that your subconscious starts working on it in the background. You leave the door open even when it costs you a clean afternoon. You write down the evidence against your own theory. You manage your faults instead of pretending you do not have any. You are willing, when the moment comes, to ask the question at the lunch table even if your friends would prefer you stopped.
Almost none of those steps are technical. None of them require talent that is not already inside the room. Which is, I think, exactly the part of the essay he wants you to notice.
Near the end, Hamming basically admits the advice is easy. He says it more than once, in different words. I've told you how to do it. It is so easy, so why do so many people, with all their talents, fail? Then he answers his own question. Drive. Commitment. Anger pointed at the system instead of the work. Ego in the wrong direction. Alibis. The exits are well marked. He paints arrows on the floor.
I think the gap between his advice and his readers, including me, is not a knowledge gap. It is a courage gap. Courage does not compound the way knowledge does. You do not accumulate it once and bank it. You have to spend it, fresh, every time. Every important problem feels stupid before it works. Every door you leave open costs you an afternoon. Every time you ask the lunch table question, someone is annoyed at you for a week. Hamming's model says you pay these costs anyway, because the alternative is a comfortable life nobody remembers.
He ends the talk with one line, almost as an aside, after fifty minutes of dense, lived through advice: Go forth, then, and do great work.
Reading that at 16, with no track record to defend and no reputation to protect, the line lands harder than I think it would at any other point in my life. I do not yet know if I will live up to it. I do not yet know what my attack is, or what corner I am going to spend the next ten years on, or whether I have the courage Hamming keeps asking me for. I just know that something about the talk has made it impossible to keep doing the comfortable, decorative version of the thing I was already doing.
Which, I think, is what good writing is supposed to do. It should not let you stay where you were. It should make your old plans look slightly embarrassing in the morning. By that test, this is one of the best things I have ever read, and I am very glad it found me before I had any other plan.