The obvious move after a book like Endure is to get busy on a follow-up. But instead I found myself tugged in other directions, none of which sustained my attention for long. Weighing on my mind was my prior history of career swerves. I had started out studying physics. After submitting my PhD thesis, at age twenty-four, I decided to go all-in as a middle-distance runner and train full-time in an attempt to qualify for the Olympics. A little over a year later, I checked my world ranking and my bank balance and decided to be a physicist after all. I took a postdoctoral research position in the National Security Agency's quantum computing group, working out of a lab affiliated with the University of Maryland. It was fun and intellectually rewarding, but two and a half years later, at age twenty-eight, I left my postdoc to start a master's degree in journalism at Columbia University. Was I astutely following my interests? Or, I sometimes wondered, was I a dilettante, chasing whatever shiny new object caught my eye instead of sticking to my original challenges?

What I worried about a decade and a half later, as I wandered in the post-Endure wilderness, was that I might be repeating this pattern: setting an audacious goal, spending years working tirelessly toward it, and then, once success was within reach, walking away to pursue something completely different. A decision like that might seem quixotic when you're twenty-eight, but in your midforties it starts to look pathological. So there was an insistent voice in my head exhorting me to exploit all the effort I'd put in to become "the science of endurance guy." And there was another, quieter voice reminding me that the seemingly irrational decision to explore a journalism career had led to the most rewarding professional years of my life, and that following that urge one more time might pay off again. In other words, I eventually realized, I was on the horns of a ubiquitous and exhaustively studied metachoice that researchers call the explore-exploit dilemma.

In 1991, a professor named James March, at Stanford University's Graduate School of Business, published a paper called "Exploration and Exploitation in Organizational Learning." March was a prolific and influential scholar, as well as a polymath: he published poetry and produced films about the leadership lessons of Don Quixote and War and Peace. Starting in the 1950s, his work with economics Nobel Prize-winner Herbert Simon and others brought nuance and complexity to the study of corporate decision-making. More than a few scholars believe March should have shared Simon's Nobel.

March's 1991 paper highlighted the fundamental tension between what he dubbed exploration, encompassing "search, variation, risk taking, experimentation, play, flexibility, discovery, innovation," and exploitation, encompassing "refinement, choice, production, efficiency, selection, implementation, execution." You can exploit the knowledge and resources you already have, or you can explore in search of an outcome that is uncertain but might turn out to be better. You can devote your corporate resources to churning out widgets as cheaply and efficiently as possible, or you can devote them to inventing a sprocket that will make widgets obsolete. But in a world of finite resources, you can't give your all to both at the same time. You have to choose. March's main argument was that the delayed and uncertain rewards of exploration mean that organizations tend to systematically underinvest in it.

The paper also had an unintended side effect: the "explore-exploit" terminology caught on, crystallizing a concept that researchers in various academic silos had been grappling with in their own specialized languages. In the years that followed, mathematicians who had been toiling for decades on complex optimization algorithms realized that they were addressing the same fundamental questions as economists and business thinkers like March, as well as evolutionary biologists studying human migration paths, ecologists examining animal foraging patterns, neuroscientists decoding the brain's decision circuitry, computer scientists teaching machines to learn, and psychologists and philosophers trying to understand why we want what we want.

As an ex-physicist, I was a sucker for the mathematical approach to explore-exploit decisions. I wasn't naive enough to imagine that I could plug a few details about myself into an equation and get quantifiable advice about my next vacation or my next book. But I began to see explore-exploit dilemmas all around me: in the tug-of-war between my long-standing love of running and my emerging interest in rock climbing; in the music I chose to stream; in the friendships I chose to maintain, neglect, or initiate; in the investment choices I made with my retirement savings; in the search for alternatives in my writing to familiar clichés and overused adjectives. And as I dug into a century's worth of progress on exploring algorithms, I found insights that helped me think through my dilemmas. For instance, the math is pretty clear about the lesser value of pure exploration in your forties compared to your twenties. "Your horizon is getting shorter," a cognitive scientist at Georgia Tech, Robert Wilson, explained to me. There's less time to reap the delayed benefits of a new path, and there's plenty of evidence that humans get worse at exploring and do it less often as they age—though that doesn't mean I should lean into this decline.

The obvious move after a book like Endure is to get busy on a follow-up. But instead I found myself tugged in other directions, none of which sustained my attention for long. Weighing on my mind was my prior history of career swerves. I had started out studying physics. After submitting my PhD thesis, at age twenty-four, I decided to go all-in as a middle-distance runner and train full-time in an attempt to qualify for the Olympics. A little over a year later, I checked my world ranking and my bank balance and decided to be a physicist after all. I took a postdoctoral research position in the National Security Agency's quantum computing group, working out of a lab affiliated with the University of Maryland. It was fun and intellectually rewarding, but two and a half years later, at age twenty-eight, I left my postdoc to start a master's degree in journalism at Columbia University. Was I astutely following my interests? Or, I sometimes wondered, was I a dilettante, chasing whatever shiny new object caught my eye instead of sticking to my original challenges?

What I worried about a decade and a half later, as I wandered in the post-Endure wilderness, was that I might be repeating this pattern: setting an audacious goal, spending years working tirelessly toward it, and then, once success was within reach, walking away to pursue something completely different. A decision like that might seem quixotic when you're twenty-eight, but in your midforties it starts to look pathological. So there was an insistent voice in my head exhorting me to exploit all the effort I'd put in to become "the science of endurance guy." And there was another, quieter voice reminding me that the seemingly irrational decision to explore a journalism career had led to the most rewarding professional years of my life, and that following that urge one more time might pay off again. In other words, I eventually realized, I was on the horns of a ubiquitous and exhaustively studied metachoice that researchers call the explore-exploit dilemma.

In 1991, a professor named James March, at Stanford University's Graduate School of Business, published a paper called "Exploration and Exploitation in Organizational Learning." March was a prolific and influential scholar, as well as a polymath: he published poetry and produced films about the leadership lessons of Don Quixote and War and Peace. Starting in the 1950s, his work with economics Nobel Prize-winner Herbert Simon and others brought nuance and complexity to the study of corporate decision-making. More than a few scholars believe March should have shared Simon's Nobel.

March's 1991 paper highlighted the fundamental tension between what he dubbed exploration, encompassing "search, variation, risk taking, experimentation, play, flexibility, discovery, innovation," and exploitation, encompassing "refinement, choice, production, efficiency, selection, implementation, execution." You can exploit the knowledge and resources you already have, or you can explore in search of an outcome that is uncertain but might turn out to be better. You can devote your corporate resources to churning out widgets as cheaply and efficiently as possible, or you can devote them to inventing a sprocket that will make widgets obsolete. But in a world of finite resources, you can't give your all to both at the same time. You have to choose. March's main argument was that the delayed and uncertain rewards of exploration mean that organizations tend to systematically underinvest in it.

The paper also had an unintended side effect: the "explore-exploit" terminology caught on, crystallizing a concept that researchers in various academic silos had been grappling with in their own specialized languages. In the years that followed, mathematicians who had been toiling for decades on complex optimization algorithms realized that they were addressing the same fundamental questions as economists and business thinkers like March, as well as evolutionary biologists studying human migration paths, ecologists examining animal foraging patterns, neuroscientists decoding the brain's decision circuitry, computer scientists teaching machines to learn, and psychologists and philosophers trying to understand why we want what we want.

As an ex-physicist, I was a sucker for the mathematical approach to explore-exploit decisions. I wasn't naive enough to imagine that I could plug a few details about myself into an equation and get quantifiable advice about my next vacation or my next book. But I began to see explore-exploit dilemmas all around me: in the tug-of-war between my long-standing love of running and my emerging interest in rock climbing; in the music I chose to stream; in the friendships I chose to maintain, neglect, or initiate; in the investment choices I made with my retirement savings; in the search for alternatives in my writing to familiar clichés and overused adjectives. And as I dug into a century's worth of progress on exploring algorithms, I found insights that helped me think through my dilemmas. For instance, the math is pretty clear about the lesser value of pure exploration in your forties compared to your twenties. "Your horizon is getting shorter," a cognitive scientist at Georgia Tech, Robert Wilson, explained to me. There's less time to reap the delayed benefits of a new path, and there's plenty of evidence that humans get worse at exploring and do it less often as they age—though that doesn't mean I should lean into this decline.