Essays

Life in the context of entropy

“The whole is greater than the sum of its parts,” said Aristotle. The phenomenon of being a whole allows living organisms to momentarily resist the trajectory of nature guided by entropy.

Entropy is expressed as the degree of possible rearrangements for components within a system. Being integrated as a whole reduces the number of possible arrangements.

Humans, as living organisms, momentarily resist and defy the course of nature. We consume and digest highly complex and chemically ordered energy sources such as proteins and carbohydrates and attempt to maintain functional unity. The idea of consuming low-entropy food, or “negentropy,” was discussed by Schrödinger in his 1944 book What is Life?

A living organism is characterized by the ability to maintain homeostasis—the fight against disintegration. Our immune system, blood flow, and digestive system circulate and prevent the accumulation of the bad. Yet, we will eventually decompose, following the call of entropy, resulting in a greater number of arrangements of scattered atoms and molecules.

While alive, we develop tightly coupled and sophisticated genetic information in the form of DNA, providing the code for our offspring to continue the fight for survival against nature over billions of years. Our fight continues.

So, how do we live a full life? There are many answers based on one’s perspective. Some are religious points of view. But there is one commonality for all of us in the context of entropy.

We are more defined as human, and we live by doing activities that reduce entropy: the ability to focus, use consciousness, form civilization through laws and physical structures, and remain in control rather than succumb to randomness. However, above all, nurturing and giving birth to a child can be considered the most human and living act possible from the context of entropy.

Another attempt is the use of “mimesis,” the replication of ideas and thoughts that form culture. We also attempt to decrease entropy by providing a blueprint of ideas for the next generation. This was discussed in the book The Selfish Gene by Richard Dawkins.

There is no right way to live as a human being. But we can attempt to live fully by guiding our activities toward a low-entropy state. We must appreciate it and use it while we can. We only have about 1,000 months or so. Then, we are gone.

July 11, 2025, Incheon, South Korea

Procrastination is natural

The subconscious brain can detect whether an upcoming task is associated with pain. Our brains are capable of recording and processing information without our conscious awareness—dreaming is a prime example. Over the past five years, I have had to write and prepare reports, exams, manuscripts, and tests. I have experienced procrastination. In this document, I discuss why procrastination is a natural behavior and how I manage it.

I am not a historian. We are biologically inclined to avoid pain and seek immediate pleasure. I imagine our ancestors only foraged for food when they were hungry, endured great suffering from disease and animal attacks, lived short lives, and faced extremely high child mortality rates. A long-term perspective likely gained importance after the agricultural revolution, which started about 12,000 years ago. Written language and arithmetic must have developed around this time to support planning, maintenance, and recordkeeping for agriculture. Twelve thousand years is not a long time from an evolutionary perspective. That is why I like to think that our primal instincts, shaped before the agricultural revolution, have the dominant influence on our behaviors today. In this context, procrastination is a natural and protective response to avoid immediate cognitive suffering.

Since the agricultural revolution, long-term thinking has become increasingly important and rewarding. The classic Stanford marshmallow experiment demonstrates that children who delay gratification tend to succeed more in life. Still, facing painful tasks like writing introductions or correcting mistakes on graded papers remains difficult. I often remind myself: “Do not overestimate willpower; do not underestimate biology.” It is more effective to engineer our reward systems than to rely on willpower alone. We are wired to avoid discomfort and seek comfort. The key is to use this tendency to our advantage.

The Pomodoro Technique is one of my most effective strategies. It offers a clear endpoint, a “win.” Knowing there is an end to the task matters. I dedicate specific Pomodoro sessions to each task (50 minutes of focused work followed by a 10-minute break). The idea of a short, defined session helps reduce resistance. After four Pomodoros, I reward myself with either a nap or team sports. Over time, this has trained my brain to associate focused work with eventual pleasure.

But there is still activation energy required to begin a Pomodoro session. Having a goal helps, but it’s often not big enough to get me started. For example, writing this blog post is easy due to the joy of writing alone, so often I don’t need the Pomodoro session. But for tasks considered less stimulating, such as writing graded lab reports in my experience, applying controlled pressure is more effective in generating the minimum activation energy to get started.

I have a study partner with whom I agree to compare academic performance, and the lower performer buys a meal at the end of the semester. I also share my goals with family members. This creates external accountability and the risk of disappointing people I care about. The fear of losing trust becomes a motivational factor that leads me to sit down and begin the Pomodoro sessions.

Rest is essential. Without mental and physical rest, a greater activation energy is required to get started. Yes, a few days before an exam, as a student, we can probably run on 5 to 6 hours of sleep per night because the fear of failure alone serves as the driving force to overcome the activation energy to get started. However, this approach carries the risk of not having enough time. Most importantly, it takes away the “fun” of studying, associating learning with a painful and mental process. This then creates a further cycle in the following exam that is associated with pain, so we further procrastinate.

We must recognize a delicate balance between procrastination and genuine fatigue. On regular days, I sleep without an alarm to allow full recovery. If I wake up feeling unexcited to attack the day, I continue sleeping. That is a signal from my body and brain. Listening to these cues helps me avoid burnout while staying productive. Particularly for students and knowledge workers, it’s not about the number of hours awake. It’s the product of Method, Intensity, and Time (MIT). I will write more about this topic later.

Procrastination is not simply a flaw. It is a natural response rooted to protect us. By understanding this and working with—not against—our instincts, we can create systems that reward progress and reduce resistance. Engineer the body through reward and pain, rather than relying solely on willpower.

August 2024, 160 Claremont Ave, New York

The value of poster presentations in scientific conferences

We share scientific value with the broader scientific community in three main ways: on peer-reviewed platforms, non-peer-reviewed platforms, and through research presentations. Here, I focus on one form of the third method: poster presentations. In July 2024, I will deliver a two-hour poster presentation daily for a week. This blog post is written to understand the nature of a poster presentation, my goals through the poster presentation, and my strategy during the poster presentation.

First, I discuss the nature of a poster presentation. A journal paper is passive and objective, while a poster presentation is dynamic and personal. A journal is a final product, pre-compiled. In contrast, a poster presentation is a one-on-one or one-to-few session that we can tailor on the fly. With each sentence and paragraph, we introduce new words or concepts and receive feedback through facial expressions, voice, words, and body language. This gives us a unique chance to customize our value to the engaged person based on immediate feedback. A poster session requires soft skills, such as self-awareness. Each person has a different level of familiarity with the topic, making each poster session a new talk with a different audience.

Unlike an oral presentation, the audience has the freedom to decide to stop by the poster. This freedom allows participants to be more engaged and feel more personal during the session than during an oral presentation, where asking questions may be difficult due to fear of revealing a lack of understanding to the wider audience. Topics during an oral presentation may not interest the audience.

Next, I discuss the goal of a poster presentation. A successful journal paper impacts the field and is cited by other journals. A successful poster transforms the interaction into a professional relationship that serves the interests of both parties during and after the conference. We attend conferences to build new relationships and reinforce existing ones. A poster session is an integral part of achieving this goal. In June 2024, I attended a four-day workshop. Training and tutorial videos were available online from previous years, but they lacked the experience of forming interpersonal relationships. During meals, I attempt to initiate shallow interpersonal relationships by discussing non-science topics and remain as a pleasant person. This is important for my strategy next.

I, first, identify two types of audiences: those to whom I can provide value for their studies or careers, and potential colleagues working on similar problems in my field to discuss ideas and seek feedback. These audiences may turn into trusted professional relationships, which we aim to cultivate.

During my poster session, I create an environment where both parties can feel vulnerable. I start by sharing my own vulnerabilities and weaknesses, making the other person more comfortable sharing theirs. For example, I mention the work I have done and the work of others, highlighting why they were helpful due to my lack of skill. Vulnerability helps us understand each other’s needs. Complementary skill sets form a professional relationship. Additionally, discussing non-science topics during breakfast, lunch, or between sessions is important because we are more likely to be vulnerable with people we have talked and shared jokes with.

I observe non-verbal cues. Poster sessions have no set timeline. Engagement time depends on the other person. A simple way to gauge interest is by the direction of their feet and side-eye movement. If they are pointed away from the poster, the person is not engaged, and it’s time to either ask about their interests or end the presentation.

It is difficult for the audience to read the sentences on the poster and listen to the content simultaneously. The goal is to be interactive and identify how both parties can benefit each other, not one-directional like an oral presentation. We pause, ask questions, seek their feedback, and inquire about their interests, creating a back-and-forth dialogue.

A QR code or URL is not enough. If possible, a poster should have a demo. For open-source tools, I bring my own laptop, place it on a round cocktail table, and demonstrate. The demo must be intuitive, useful, and flawless. For non-harmful materials, I believe it’s a good idea to bring the materials and share them with the audience on the spot. I recall a materials science course where the instructor brought different types of materials, and we could sense and feel them. That is engaging and intriguing for most.

By the end, the poster must have a clear call-to-action. If I have identified the two types of audiences mentioned, I would invite them to stay connected and exchange contact information, have lunch or dinner afterward to build trust.

To reiterate, we attend in-person events to cultivate professional relationships that benefit each other’s careers. A poster is a session that allows one to identify a group of audiences that may turn into a professional network. During a poster, we remain open, interactive, and vulnerable to identify each other’s needs and find complementary skills. If both agree, we collaborate to advance each other’s careers.

Practice is costly

The term “practice” in school is associated with “exams” and “problems”. They serves as a checkpoint to gauge a student’s understanding of learning material and their ability to apply core concepts.

Regrettably, practice demands time and stamina. In Fall 2023, I enrolled in a course on Phonons. One practice problem asked to analytically express the expected position of an anharmonic oscillator from the hamiltonian. My initial attempt had 3-4 pages of unorganized derivations. A mistake early on could necessitate redoing the entire set of derivations. Once I understood the overall scheme, I refused to engage in another practice run. Each practice run was associated with physical pain.

I paused. I had to minimize the physical labor, time, and potential areas for mistakes. I stared at my written work. I used my index fingers to locate sections where I could re-express them in symbolic forms while ensuring clarity for the grader. To reduce strain and increase efficiency during practice, I switched to using a whiteboard. Before each run, I spent more time on the evaluation process. At the end, I streamlined the derivations to 1-2 pages. The day before the exam, I practiced again, to check my preparedness. Yet, I minimized the time required to practice by solving problems with my eyes closed. On exam day, I left the classroom early. I had attained the desired academic performance.

My experience underscores that practice itself is not the goal. The objective of practice is to identify the flaws and inefficiencies in our current techniques. Between practice runs, we consciously refine, reinforce, and eliminate the current technique through imagination, discussions with colleagues, and literature surveys.

The ideal goal is to allocate the least amount of resources to practice runs yet perform at the expected level. Abraham Lincoln famously said, “If I had six hours to chop down a tree, I’d spend the first four hours sharpening the axe.” If Abraham Lincoln were to participate in a chopping competition, he would have experimented with different materials for the axe, modified the saw’s shape, and adjusted the weight balance during those four hours daily. Refinement through the evaluation process between practice runs is what allows us to protect our finite resources and perform at the highest level.

July 8, 2024, 160 Claremont Ave, New York

Update: I have written a guide on My optimization principles that reflects this short essay.

My approach to daily work

“Slow but steady wins the race” is the moral of The Tale of the Turtle and the Rabbit. Unfortunately, the phrase is contradictory in practice. If we are slow and steady with no acceleration, we do not win the race. This was a random thought.

We run a race called a day. A day has no finish line but a timeline. To me, winning the day means improving my skills, feeling fulfilled, and staying motivated for the next day. After years of trial and error since I returned as a full-time student four years ago, I am proud to say I have developed techniques and mindsets that allow me to win daily. Here is my approach.

My race begins the night before. I sleep 8 hours and usually wake up without an alarm. I have examined the number of hours of sleep my body needs to wake up naturally and stay productive throughout the day. It is 8 hours. Sleeping fewer hours than needed indicates I am not productive during the day or I lack time management skills.

During the race, I eliminate visual and sensory cues that might derail me. The book The Power of Habit states that most of our actions are cued subconsciously from the environment. I remove these root cues. I have no email or group messaging apps. The apps are only installed on my mobile phone, which remains muted and out of sight. I willfully check emails and messaging platforms only during breaks. This way, my actions remain controlled, rewarding myself with checking information at the allowed time by following the focus technique next.

I focus for 4-hour blocks using the 50-minute focus and 10-minute rest technique. I view the brain like any other muscle in the body. Fortunately, the brain can be used in full mode for an extended time. However, if I were to maximize the number of pull-ups in 24 hours, I would space out the sessions. During the 4 hours, I turn on a video available on my desktop of a person conducting a 4-hour session on my side monitor with time provided. The person on the monitor provides great accountability for staying in the race and serves as an internal clock. For each session, I record in a plain text editor the time and tasks I have accomplished. I do not use Notion or full-featured note-taking apps to avoid cues. The plain text file filled with completed tasks provides a sense of achievement and momentum.

Every 50 minutes, I take a 10-minute rest to allow my eyes and brain to both relax and consolidate information. During the 10-minute rest, I listen to music or play mini basketball at home. On a normal day, I conduct 2 sets of 4-hour blocks. Beyond the timed sessions, I explore ideas, gain new knowledge, write blog posts, and learn a foreign language without tracking the time. I exercise, either between the two blocks or after the two blocks, by playing basketball or doing a compound body workout of pull-ups and dips.

During weekends and holidays, I do not force myself; a regular race does not apply. I work at less than half the intensity without tracking time. I relax, read books, and enjoy time with my family. I am happy to rest because I have had a fulfilling week. We need to rest to remain happy, appreciative, and fulfilled. Relaxation is often accompanied by daydreaming and the exploration of ideas. When I have good ideas, I record them briefly on my phone. I give myself permission to rest and strategize for my goals. Then, I begin my race again.

July 5, 2024, 160 Claremont Ave, New York

Failure framework: experimental, expensive, pivotal, and avoidable

Failure is the inability to meet an expectation. The expectation is the key component. The position of the expectation dictates the outcome as a success or a failure. The position is associated with quality, standards, regulations, and laws. Individuals, organizations, and nations set different positions.

The expectation sets the state of the outcome. The analogy of a glass being half-empty or half-full is an expectation-based result. If I expected the glass to be full, I view it as half-full. If I expected it to be empty, then it’s half-empty. The state is based on the expectation before observing the water level.

Not all failures are the same due to differences in resource allocation, the magnitude, and the reversibility of the consequences. For example, the failure to maintain a server for financial applications is incomparable to other server failures. While the functional expectation is the same, the reversibility and magnitude of the consequence differ. Here, I present four types:

The first type is experimental failure, characterized by a high level of reversibility and repeatability with minimal resources. It is commonly observed in the research and development stage. Examples include receiving bug reports from users and collaborators. As a student, I strategically use experimental failure for exam preparations by writing exam-like questions on flashcards. I am expected to know the answers a day before the exam. Discovering problems I am unable to solve, I repeat the problems until I meet the expectation with confidence. The phrase, “fail fast, fail often,” is appropriate here. It is a great way to test one’s product and software integrity, provided the consequences are minimal and the process is repeatable.

The second type is expensive failure. “Expensive” often relates to value. I prefer it over “costly,” which solely connotes negativity. Not all failures are expensive, as they require substantial resource allocations. In machine learning, this could be attaining sub-optimal performance in trained models. In simulations, it is failing to reach convergence after a weeks-long effort. In experimental work, it is the failure to validate a hypothesis after 3 to 6 months of dedicated work. For researchers, it includes manuscript and grant rejections. For students, it could be poor midterm and final exam grades. As a junior in college, I enrolled in a graduate-level electrical engineering course called Deep Learning. Despite failing to meet my grade expectation, the failure was accompanied by tremendous knowledge gained.

Professionals encounter expensive failures. While the consequences are significant and could potentially cost one’s job, it is important to recognize that such failures require a substantial mental commitment to achieve high expectations. Those who achieve these expectations often reap benefits not available to those who do not attempt to do so. Therefore, although deciding to take on such risks involves potential downsides, I believe that (1) the willingness to allocate substantial resources to achieve high expectations, (2) the ability to take ownership of failures, and (3) the capacity to make improvements are prerequisites for success.

The third type is pivotal failure. This failure significantly affects one’s life trajectory and is often associated with a great magnitude of consequence and irreversibility. Examples might include failing to find a job in a specific industry, being rejected from programs, failing licensing exams, or losing an election as a politician. Those with high expectations may encounter this pivotal failure more frequently due to the scarcity of available resources.

The fourth type is avoidable failure. These failures are best avoided as they are not only irreversible but also costly. They involve failing to achieve expectations set by regulations, laws, and practices. Examples include failing to meet safety checks required to operate a lab, committing academic plagiarism, or failing to comply with regulations and laws. These failures result in wasted resources and are best learned from the mistakes of predecessors.

This framework offers a way to categorize and understand failure. However, the four types of failure can coexist in varying proportions. For instance, some experimental failures may also be avoidable or even pivotal.

June 16, 2024, 160 Claremont Ave, New York

Goal

I have goals. I think about my most cherished goal every few hours. I sleep and wake up with it. This goal serves as a guiding star, providing a sense of direction regardless of the circumstances. Thinking about the goal itself is magical. It generates a sense of purpose. All my actions and time are directed towards it. Any work that may seem trivial on its own is no longer trivial. It is a step required to achieve the goal.

The goal itself does not provide detailed action plans. Instead, my brain subconsciously explores options and proposes action steps required to achieve the goal. New action steps materialize when I am resting. I record them on a device. I use my conscious brain to filter and prioritize them.

I record what I need to accomplish today, this week, this month, this quarter, this semester, and this year. I have daily to-dos. I do not always check off all the list items. I focus on what I have accomplished. As my brain is explorative with ideas and action steps, there are always more than I can complete in a day. If there aren’t enough, I ask for more.

For my daily hours, I utilize a 50-minute focus, 10-minute rest technique, averaging 10 to 12 hours a day at home with no distractions. I play basketball or listen to music between sessions. I prioritize my physical health above all else. I do not need to force myself. I just consistently work towards the goal.

June 9, 2024, 160 Claremont Ave, New York

Embrace duality for excellence

An electron can be modeled with states such as “spin up” and “spin down,” among others described by quantum numbers. These varying states coexist in superposition until one of the states is observed with a certain probability.

Similarly, multiple states of emotions and thoughts may coexist. Our mental state is not binary. We may express a specific mental state—either happy or sad—only when we state it, similar to how an electron manifests a single energetic state when measured. The written or verbal mental statement may not depict the superposed states. The expressed state merely has a higher probability than other states, similar to the probable observation of the lowest energy state at room temperature in electrons.

Elite athletes, such as Michael Jordan (MJ), exemplify both confidence and humility. MJ scored game-winning points in NBA and NCAA championship games. Yet, he also displayed humility by working to improve his three-point shooting percentage and transform into a mid-range shooter with his trainer, Tim Grover, for nearly two decades. MJ showed a willingness to listen and adhere to practice and diet routines. In practice, he was ruthless, yet he could not hold back his tears after winning his sixth championship, cradling the trophy in his arms. Duality and plurality of emotions may coexist. The probability of each emotional state is merely altered by circumstances, similar to how temperature influences the distribution of electronic states.

I build open-source programs that help experimentalists analyze synthesized crystal data. I design data structures for crystal geometries, develop command-line user interfaces, and generate publication-quality figures. I am proud and confident in my ability to deliver results. Nonetheless, I recognize that my craftsmanship can still be elevated compared to other open-source projects. Kobe Bryant said, “Once you know what it is in life that you want to do, then the world basically becomes your library. Everything you view, you can view from that perspective, which makes everything a learning asset for you.”

I could enhance my code by using matrices to compute atomic distances instead of relying on for-loops. I could improve the flow of the command-line interface by seeking feedback from users without programming expertise. My goal is not just to create good programs that merely work. I aim to craft phenomenal inventions that are loved by my users. I invest my time—a part of my life—in learning and applying unit testing, static type checking, continuous integration, and any other practices that elevate my craft. No audience watches elite athletes’ individual practices in the morning, but that is where their legacy begins.

June 2, 2024, 160 Claremont Ave, New York

Thoughts on rejection

Rejection is a form of failure. Failure is the inability to meet expectations. Here, I present two types of rejections and how I navigate my life.

As of this writing, I encounter rejections every one or two months. My recent paper was rejected by a journal, and my request for collaboration was declined. As I further advance my career in academia, I expect these intervals to decrease.

I do not use the phrase “don’t take it personally.” There is nothing more personal than spending one’s invaluable and finite resource called time.

Nonetheless, rejection is inevitable. We compete for finite resources provided at each level of our career.

We are on a ship called a career. I view rejection as a reef in the ocean. When the ship hits the reef, there are two outcomes: it either sinks the ship or alters its velocity with broken parts.

The first type of rejection may destroy the ship and provide no further opportunity to advance one’s career. In most cases, however, we encounter the second type of rejection, which alters the velocity of one’s career and requires repairs. Not to mention, some ships are equipped with special radars called mentors and knowledge that prevent the ship from encountering the reef. However, once the ship enters uncharted territory, it will inevitably encounter a reef at some point.

When I hit the reef, I retreat and reflect. During the repairs, I locate where the reef was on a map. I ask myself why I navigated toward the reef and whether I can avoid it next time. I seek advice from mentors and books on how they have navigated the path. Then, I embark again.

July 4, 2024, 160 Claremont Ave, New York

The Structure of Scientific Revolutions - Thomas S. Kuhn

It is widely accepted that science evolves linearly based on the body of knowledge curated by predecessors. The phrase “…standing on the shoulders of giants” by Issac Newton embodies the notion that scientific advancements are built atop existing theories and concepts. However, ‘The Structure of Scientific Revolutions’ by Thomas S. Kuhn claims scientific revolutions are neither linear nor cumulative.

According to Kuhn, not all theories and concepts are defined as “revolutions.” Kuhn indicates that scientific revolutions only occur when there is a shift in the existing paradigm to a new paradigm within a scientific community. A paradigm is described as a standard of equations, techniques, apparatuses, and educational systems that a scientific community has embraced and practiced. A paradigm provides the common technical vocabulary that allows scientists in the community to articulate concepts and collectively conduct experiments. In the early seventeenth century, Newton’s Principia Mathematica provided a set of equations and techniques that gave rise to the doctrine of classical physics, also referred to as Newtonian physics. Newtonian physics has become the standard paradigm that models the motion of particles and gravity. However, Kuhn argues that long-standing paradigms often fail to explain the anomalies observed in nature, which causes community members to lose trust in the existing paradigm. Scientists, often young and new to the field, seek a new set of equations radically distinct from the existing paradigm. Eventually, one dominant theory emerges and displaces the old paradigm, as in Einstein’s general theory of relativity, which accounted for the anomaly found in the orbit of Mercury that Newtonian physics failed to explain. Kuhn outlines the process of scientific revolutions with a framework consisting of 4 phases, pre-science, normal science, crisis, and revolution, whose last phase is resolved by a paradigm shift.

The first phase within Kuhn’s framework of scientific revolution is defined as pre-science. While individual scientists attempt to discover new theories during pre-science, there is no dominant set of equations, techniques, and concepts referred to as a paradigm. During the pre-paradigmatic period, scientists observe and collect facts. Due to the lack of a common paradigm, scientists within each pre-paradigmatic school confront one another and interpret these facts in different ways. Pre-science is further characterized by a lack of common scientific vocabulary. The lack of common language hinders collaboration amongst scientists and schools. Thus, Kuhn describes pre-science as the least productive phase in the framework.

The transition from pre-science to normal science occurs as one set of theories and concepts becomes dominant within the scientific community. The distinction between pre-science and normal science is the existence of a paradigm. Kuhn explains that normal science “is predicated on the assumption that the scientific community knows what the world is like,” comparing a paradigm to a “map” that guides scientists towards modeling nature. Kuhn illustrates that research within the paradigm of normal science is also analogous to “puzzle-solving,” where the problems and questions within the paradigm are scattered pieces of solvable puzzles. The puzzle pieces are fit together in a complete shape through refinement and precision. The comparison of a paradigm to a map and puzzle-solving assumes that the scientific community is capable of knowing nature guided by the paradigm. The period of normal science is marked by cumulative and linear developments facilitated by advancements in measuring devices and techniques. Newton’s universal law of gravitation in Principia Mathematica published in 1687 approximated the Moon’s orbital period based upon the principle that attractive gravitational force exists between two objects. Furthermore, using the same principle, Newton predicted the motion of other planets in the Solar system. Within the paradigm of normal science, research questions and facts collected serve to support the existing paradigm. Normal science is not focused on novelty but rather precision and confirmation.

The transition from normal science to crisis takes place when new inexplicable findings referred to as anomalies threaten the foundation of the existing paradigm and cast widespread doubt within the scientific community. As measuring techniques and devices improve, anomalies become easier to detect and harder to be avoided within the scientific community. The anomaly in Newtonian physics was first observed by Le Verrier, a French astronomer, in 1859. Through Le Verrier’s improved mathematical technique of predicting the motion of Mercury, he discovered that there was a 43 arcsecond per century discrepancy between the theoretical value of Newtonian physics and the observed precession of the perihelion of Mercury. Perihelion is the point in the orbit of a planet nearest to the Sun. One of the ways scientists respond to an anomaly is by devising ad hoc modifications of their theory in order to eliminate any apparent conflict within the paradigm. In response to the discrepancy in Mercury’s precession, some scientists that defended Newton’s paradigm assumed that there was an invisible dusk between the Sun and Mercury that affected the precession. Others proposed a new planet, Vulcan, orbited close to the Sun and was responsible for the discrepancy. As the anomaly remains inexplicable within the existing paradigm, scientists in the community become more critical of the paradigm and begin to question its underlying foundations. The widespread acknowledgment of these inconsistencies within the existing paradigm and the introduction of new theories illustrate the defining characteristics of crisis. During a crisis, scientists, often young and less invested in the existing paradigm, seek theories outside the boundary of the paradigm in order to explain the anomaly.

An alternative paradigm is established when a new set of theories and concepts that explains the anomaly becomes widely accepted by the scientific community. In the case of the 43 arcseconds per century anomaly found in the precession of Mercury, it was Albert Einstein’s general theory of relativity published in 1915 that precisely modeled Mercury’s orbit without discrepancy. Einstein’s new theory superseded Newton’s universal law of gravitation and became the standard for predicting a planet’s orbit. The displacement of the old paradigm by a new paradigm marks the defining characteristic of Kuhn’s fourth phase of revolution, in which the newly constituted dominant paradigm entirely reconstructs the fundamental methods, generalizations, and rules of the old paradigm. The shift to Einstein’s theory of relativity in which time and space are not fixed demonstrates that the foundations behind a new paradigm are not cumulative but rather radical. However, Kuhn notes the cyclical and periodic nature of these paradigm shifts or transformations, in which scientific revolution circles back to the period of normal science. After a new paradigm is introduced, the community enters the phase of normal science with scientists of the new order aiming to improve the precision of the paradigm. In the case of Einsteinian physics, the theory of general relativity was further used to predict the movement of the precession of perihelion in other planets. Just as in the shift from Newtonian to Einsteinian physics, according to Kuhn’s framework of scientific revolution, scientists of Einsteinian physics will discover anomalies that lead to new crises, and the established paradigm will again be transformed.

While scientific revolutions accomplished by paradigm shifts within a scientific community seem to advance the knowledge of science towards truth, Kuhn maintains that the role of scientific revolution lies in providing a new “map” that serves to temporarily guide scientists until anomalies are observed. Furthermore, there is no linear progression towards truth but only periodic rise and fall of paradigms. As Kuhn describes, “Einstein’s general theory of relativity is closer to Aristotle’s than… to Newton’s.” Based on Kuhn’s analysis, Newton’s phrase “…standing on the shoulders of giants” is partially accurate during the period of normal science. However, the structure of scientific revolutions as a whole is neither cumulative nor linear but rather a cycle of paradigmatic transformation.

Fall 2021, EID 367, The Cooper Union

The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail - Clayton M. Christensen

Despite seemingly sound managerial practices, such as listening to existing customers and continued investment in technology, great companies are often displaced as market leaders. According to Clayton M. Christensen, in The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail, it is precisely due to these widely accepted practices of good management that leading firms are confronted with great profit loss and eventual bankruptcy. The fall of these incumbents begs the question, “Why do great companies fail?” In his book, Christensen highlights the failure of incumbent firms to recognize and respond to new technological developments as the source of their decline.

Christensen coined the phrase the failure framework which illustrates the process of how certain types of new technologies cause established companies, also referred to as incumbent companies, to fail. Christensen defines technology as either sustaining or disruptive. Sustaining technology iteratively and incrementally improves upon an existing performance measure, a requirement or a specification demanded by customers. In the hard disk drive industry in the 1980s, mainframe manufacturers such as IBM demanded large storage capacity requirements of 300 to 400 MB from the 14-inch hard drive. Disruptive technology, often developed by entrant companies, in contrast, initially underperforms in comparison to the performance measure of sustaining technology but introduces a new performance measure. In the same hard disk drive case study, Christensen cites the architectural innovations of the 8-inch hard disk drive as disruptive technology, which introduced a new performance measure of small size demanded by desktop and mini-computer manufacturers such as Hewlett-Packard. Disruptive technology initially fails to meet the performance requirement of the incumbent companies’ customers and must find a new or niche market that values the newly introduced performance measure for survival. The 8-inch drive with the storage capacity of 40 MB initially failed to meet the storage capacity demand of the mainframe manufacturers. Christensen’s failure framework describes how, despite this initial underperformance, disruptive technology eventually causes the failure of established companies. The failure framework consists of three principle components, technology maturation, performance oversupply, and resource dependence.

Technology maturation, the first component in Christensen’s failure framework, introduces a challenge for sustaining technology to maintain the rate of improvement in performance. Technology maturation occurs at the end of the Technology S-Curve. The Technology S-Curve is a graphical representation which demonstrates the rate of performance improvement with performance on the vertical axis and engineering effort or time on the horizontal axis. As the curve progresses towards the mid-section in the horizontal axis, the slope of the curve increases as technology becomes more understood and more resources are allocated. Technology maturation occurs when performance on the vertical axis asymptotically approaches a natural or physical limit as engineering effort or time further progresses. Returning to the disk drive case study, while the 14-inch drives approached the asymptote of technology maturation with annual performance improvement limited to 22 percent, the 8-inch drives, designed to optimize storage, benefited from the advances in storage capacity with annual storage capacity improvements of 40 percent. The reduced rate of performance improvement due to technology maturation of the 14-inch drives led existing customers of the sustaining technology to pay a premium for the same incremental performance improvement. Customers of the 14-inch drives paid 1.65 USD per megabyte improvement, 13 percent higher than 1.42 USD of the 8-inch hard drive.

The second component of the failure framework, performance oversupply, occurs as the rate of performance improvement exceeds the performance requirement. In continuing the disk drive case study, Christensen presents the storage capacity of 5.25-inch drives, which exceeded 300 percent of the desktop manufacturers’ performance demand. Meanwhile, the 3.5-inch drives, which initially underperformed, ultimately satisfied the storage demand of the desktop manufacturers by maintaining the rate of storage improvement. Consequently, by 1988, the 5.25- and 3.5-inch drives both met the performance demand of the desktop manufacturers. At this point, as desktop manufacturers no longer required a drive with higher storage capacity, customers began seeking other features such as functionality, reliability, convenience, and price. As demonstrated in the desktop computer market, during this period of performance oversupply, existing customers of sustaining technology migrate to disruptive technology. In 1985, only 1 percent of the desktop manufacturers migrated from 5- to 3.5-inch drives. Within 4 years, however, the 3.5-inch drives accounted for 60 percent of total drive sales. The period of performance oversupply and the continued rate of performance improvement of disruptive technology shifted the basis of competition from storage capacity to other features such as portability and price.

The third component of the failure framework, resource dependence, describes the tendency of a company to allocate resources towards serving the existing customers. As the company depends on satisfying existing customers and in return generating profit to maintain the operational expenses of the company, incumbent companies seek more definitive ways to maintain or increase profitability. As a result, incumbent companies aggressively invest in sustaining technology and attempt to lead existing customers to higher-end products with higher gross margins. Incumbents choose not to allocate resources in disruptive technology where gross margins are lower and the market is unpredictable and smaller. In the case of the hard disk drive industry, Seagate, the incumbents of the 5-inch hard drive market for desktop manufacturers, initially canceled the 3.5-inch drive program and continued innovating the 5-inch model where customers paid higher prices for incremental megabytes of capacity. In 1987, despite the emergence of customer migration from 5-inch to 3.5-inch hard drives, Seagate executives initially disregarded the 3.5-inch market due to the smaller market size of 50 million USD and lower gross margins of 22 percent compared to the current 5-inch market with 300 million USD and 25 percent. By 1991, the 3.5-inch market grew to 700 million USD as new customers such as portable laptops manufacturers emerged, and simultaneously desktop manufacturers further migrated to the 3.5-inch disk drive during performance oversupply of the 5-inch drive. While Seagate eventually attempted to allocate resources for the 3.5-inch drive in 1988, Christensen cites that by 1991 none of Seagate’s 3.5-inch products had been sold to portable/laptop/notebook computers. In 1997, Seagate reported a 550 million net loss in sales.

Why do great companies fail? Christensen’s failure framework illustrates the process of how disruptive technology drives sustaining technology developed by incumbent firms in the mainstream market to fail. Sustaining technology incrementally improves upon the performance measure demanded by the existing customers. In contrast, disruptive technology, while initially underperforming in the performance measure of sustaining technology, introduces a new performance measure. As the performance demand of existing customers is met by both sustaining and disruptive technology, customers seek other features such as portability, functionality, and price which are offered by disruptive technology. At this point, customers of incumbent companies migrate to disruptive technology. As the process of migration continues, incumbent companies are displaced by these entrant firms and disruptive technology prevails.

Fall 2021, EID 367, The Cooper Union

The Two Cultures and the Scientific Revolution - C. P. Snow

Steve Jobs, co-founder of Apple Inc., stated, “It’s in Apple’s DNA that technology alone is not enough — it’s technology married with liberal arts, married with the humanities, that yields us the results that make our heart sing.” By recognizing the power of this union between the humanities and the sciences, Steve Jobs created numerous tools that improved the accessibility and user experience of personal computing devices, catapulting Apple Inc. into one of the most innovative companies in the world.

In his lecture at Cambridge University in 1957, later titled The Two Cultures and the Scientific Revolution, C. P. Snow, a British physicist and novelist, warned the British parliament of the widening separation between the two disciplinaries, the humanities and the sciences. Snow asserted that “the intellectual life of the whole of Western society is increasingly being split into two polar groups.” Political leaders who predominantly studied the humanities within the traditional British educational system were ill-equipped to lead the nation in the age of the scientific revolution. Snow argued that the elites often rejected the innovation of scientists, stating that while scientists had the “future in their bones,” the “traditional culture” or the elites responded “by wishing the future did not exist.” Having recognized the threat of this divide to national competition a decade after World War II, Snow demanded for a unification of the two disciplinaries with the final remark, “closing the gap between our cultures is a necessity in the most abstract intellectual sense, as well as in the most practical.”

Steve Jobs’ success in applying his knowledge of calligraphy from the humanities to the development of personal computing devices illustrates the intrinsic role of interdisciplinary diversity in innovation. When the first Macintosh computer was released in 1984, for the first time in the history of machinery, Jobs provided users with a wide assortment of digital fonts and typeface designs such as Helvetica and Times New Roman. The ability to customize the font along with a human-centric user-interface improved the accessibility of personal computing devices. During Job’s commencement speech at Stanford University in 2015, he recalled his calligraphy experience at Reed College in the 1970’s as “…beautiful, historical, artistically subtle in a way that science cannot capture.” He further stated that, had he not studied calligraphy in his 20’s, “personal computers might not have the wonderful typography they do today.” His collective insight into the humanities and software technology changed how humans interacted with machines and demonstrated the significance of interdisciplinary diversity in technological innovation.

Steve Jobs, one of the most innovative entrepreneurs of the 21st century, by combining the knowledge of the humanities and the sciences, improved the accessibility of personal computing devices for users. Conversely, this innovation by Steve Jobs also precipitated the bankruptcies of numerous companies that failed to recognize the threat of interdisciplinary divide as previously warned by C.P. Snow. Therefore, technological leaders and entrepreneurs of today must recognize and embrace interdisciplinarity as an indispensable element of innovation.

Fall 2021, EID 367, The Cooper Union

Two types of innovation and evaluation

In Christensen’s disruptive innovation theory, innovation is categorized into two types. The first type improves on earlier metrics established by the community. In quantum physics and chemistry, scientists develop approximation techniques that solve the Schrödinger equation and match experimental results. The performance of neural networks for image classification was measured based on metrics from the CIFAR-10 and CIFAR-100 image datasets.

The next type of innovation, while underperforms in the primary performance, introduces a secondary performance. This secondary performance appeals to a niche group. Density Functional Theory (DFT) has introduced a new performance measure of computational efficiency by using the three coordinates as electron density to solve the Schrödinger equation. DFT has enabled materials scientists to employ the tool for phase transitions and kinetics.

With the two types of innovation discussed, we strive to measure performance with both objective and subjective measures. For the second type, this is difficult. Jensen Huang from Nvidia said, “I find KPIs hard to understand.” “What’s a good KPI?” “Gross margins, that’s not a KPI. That’s a result.” Nvidia is known for investing in new fields such as computational drug discovery and materials science, beyond its origins as a computer-graphics chip design firm.

There are established measurables in academia and finance for evaluation. These are results. Results are goals. We do not constantly measure goals; they are our guiding stars. We observe and decide which star to follow. The destination is not the star itself; we use it as a tool to navigate our lives. Instead, we measure our velocity and operation aligned with these stars.

Hence, I must develop my own subjective criteria to evaluate my progress for the second type of research. There is no checklist. I ask open-ended questions and determine my progress based on my confidence in answering these questions.

First, I ask whether my research output provides immediate value to a niche group of scientists. Second, I consider whether it has the potential to attract users beyond the existing community. The commonality between DFT and personal computing is their ability to attract a new cohort of users with secondary performance measures—efficiency for DFT and ease of use for personal computing. Lastly, I assess whether my research outcomes have the potential to be adopted by the existing scientific community.

June 12, 2024, 160 Claremont Ave, New York

Leadership

A leader is not defined by a particular style. The role of a leader is to project a vision that aligns with individual interests and unifies members as a cohesive whole. For this reason, a leader does not necessarily need to have technical expertise. A leader is a visionary. A leader is capable of outlining a blueprint that helps individuals see how they might advance their own careers in ways they had not previously imagined.

Writing principles

Words in a sentence are ordered based on purpose. A sentence requires precise words to avoid miscommunication. In fields such as the military and medicine, professionals avoid colloquial words. Surgeons use the term “correct” instead of “right” during procedures. I recently rewrote “Find the number of atoms in a formula” to “Count the unique elements in a chemical formula.” In this context, “formula” could refer to a mathematical formula, and “Find the number” is replaced with “count.” “Elements” refers to unique atom types.

A sentence is abridged with the correct noun-verb pair. Adjectives describe nouns. Adverbs describe verbs. Adjectives and adverbs may not be necessary when the noun-verb pair adequately conveys the meaning.

A sentence should not start with symbols. If an acronym is unfamiliar to the audience, the full name should be used multiple times.

A sentence must be clear, simple, and concise in that order. Refer to Guide for written communication.

A sequence of single sentences is easier to read. A conjunction should be used sparingly.

In programming, we refactor comments, names, and structures after the function is implemented. In writing, words, sentences, paragraphs, and outlines are refactored after the meaning is conveyed.

Repetition is not harmful if it enhances clarity.

Be extremely careful using adjectives and adverbs to prevent miscommunication. These parts of speech are often immeasurable, hence subjective.

• A “significant” problem may not be a significant to another. Rather use countable nouns and use analogy for comparision.

• “Failure” is also a subjective term because it is determined by the expected outcome, which differs for each person. It is better to state what was not achieved based on the expected outcome of the individual or the organization’s expectation.

Sentences without axioms or facts have limited persuasive power in scientific writing.

June 1, 2024, 160 Claremont Ave, New York

Purpose of research

In crystallography and solid-state science, in particular, research serves to (1) discover new substances with potential utility, (2) propose methods, and (3) characterize the underlying structure and phenomena with a category such as space group.

Here, the unifying theme is to produce new knowledge. The main difference between humans and others is our ability to store and retrieve generational knowledge across time and space. Hence, the production and propagation of knowledge is a human-like endeavor. This activity equips us with the materials and methods to become the apex predators in the animal kingdom.

In contrast, this very knowledge also equips us with the capacity to destroy ourselves. Laws, regulations, and orders impose boundaries on individuals, organizations, and nations. The boundaries prevent the misuse of power derived from this knowledge. This is evident in recent history, particularly from the 1940s onwards. Hence, research is a sacred activity; it is a human activity, aimed at advancing our civilization by producing new knowledge.

June 1, 2024, 160 Claremont Ave, New York

“Good luck!”

I favor the word “serendipity” over “luck.” Both words represent an unexpected beneficial outcome. They differ in terms of self-direction and initiation. The well-known explorer isn’t merely lucky to discover new lands. I am not merely lucky to have crossed the Pacific Ocean to be in one of the lands. I decided to be here. Yes, I do not neglect what I was provided with. Nonetheless, serendipity implies a degree of will.

As a student, serendipity is the occurrence of exam problems I’ve solved several times before. As a researcher, it’s the moment I figure out how to optimize data structures and create incredible figures with Matplotlib, or when I discover open-source code that helps me learn computing with matrices. As a writer, it’s the moment when just a pair of a noun and a verb forms a better sentence. The term “serendipity” encapsulates this sense of self-initiative and direction towards solving a problem. So, when I say “Good luck,” it has a meaningful context too.

June 1, 2024, 160 Claremont Ave, New York

학업과 연구 성과를 위한 원칙 - 방식 (M)

메시는 축구 경기 중 걷는다는 악명이 있다 [1-3]. 그런데도 메시가 ‘축구의 신’으로 불리는 이유는 두 가지다.

신은 인간에게 축복 또는 고통을 줄지를 고민한다. 메시 또한 인간을 상대로 고뇌를 마다하지 않는다.

“걸으면서 상대의 위치, 공이 없을 때 우리 팀의 포지션, 저를 마킹하는 상대를 따돌리고 역습을 시작할 수 있을지를 분석해요.”

고뇌를 마친 신이 가뭄에 비와 같은 기적을 보여주듯, 메시는 최소한의 드리블로 공간을 만들어 기적 같은 공격 포인트를 만든다. ‘축구의 신’ 메시는 인간의 비판에도 이렇게 말한다.

“저는 GPS나 통계, 데이터에 크게 신경 쓰지 않아요. 경기 중에 제가 얼마나 뛰었는지에도 관심 없습니다.”

운동 분야에서 ‘신’이 되기 위해서는 공통점이 있다. 신체 사용을 최소화하면서 성과를 내는 방식(Method)을 선택해야 한다. ‘농구의 신’ 마이클 조던도 군더더기 없는 몸동작으로 수비수의 무게 중심을 무너뜨려 공간을 만든다. 여기서 ‘신체’란 뇌를 제외한 손과 발을 의미한다.

인간은 본능을 따라 뇌보다는 신체를 사용하는 경우가 많다. 예를 들어, 시험을 준비하며 자신의 필기를 보고 뿌듯함을 느꼈던 경험은 누구에게나 있다. 굳은살을 보며 만족감을 느끼고, 알록달록한 노트를 보며 하루를 잘 보냈다고 생각한다. 이는 시험이라는 미래의 전투를 앞두고 심리적 방패 역할을 한다.

학업을 위해 자신의 몸을 혹사시키는 것은 축구 선수가 많이 뛰고 과도한 드리블을 하는 것과 같다. 운동선수는 더 많이 뛰거나 더 많이 연습할수록 부상 가능성이 높아지고, 퇴행성 관절염으로 발전하기도 한다. 결국 선수 생활은 단축되고, 몸과 마음의 병을 얻는다.

학업 성과를 얻으려면 시험을 잘 봐야 한다. 시험은 이미 학습된 뇌가 제한된 시간 안에 정확한 답을 인출하여 채점자와 소통하는 과정이다. 노트 필기와 교과서 다독은 약간의 도움이 될 수 있지만, 인출이라는 관점에서는 효과적인 접근이 아니다. 즉, 시험의 본질인 인출 과정에 충실해야 한다. “머리가 하얗게 된다”는 표현은 시험과 같은 조건에서 인출 과정을 반복적으로 연습하지 않았기 때문에 생긴다.

많은 사람들은 동물적 본능에 굴복해 노트 필기나 교과서 다독에 시간을 투자하고, 신체적 고통을 마다하지 않는다. 이는 “시험은 인출이다” 라는 본질을 깨닫지 못한 결과다.

시험을 잘 보기 위해서는 학습과 인출 과정을 따로 보지 않고 병행해야 한다. 인공신경망과 같은 수학 모델 학습 과정에서도 이 점을 확인할 수 있다. 초기에 학습되지 않은 모델은 입력값(X)에 대해 터무니없는 결과값(y)을 인출한다. 이후 실제 값과 비교하여 차이를 기반으로 모델의 매개변수를 조정하면서 점차 정확한 결과를 인출하는 모델로 탈바꿈한다. 이 탈바꿈 과정은 시험을 위해 자신의 뇌를 학습시키는 과정과 같다.

본질에 맞는 접근법임에도 ‘인출-학습-병행’이 실전에 어렵게 느껴지는 이유는 두 가지다. 첫째, 정신적 고뇌가 동반된다. 자신의 학업 실패나 틀린 문제를 직면하는 일은 단기적으로 정신적 고통을 유발한다. 그러나 메시는 이 과정을 피하지 않는다. “저는 항상 자기비판적이에요. 제가 무엇을 잘했고 무엇을 잘못했는지를 가장 먼저 알죠.” 둘째, 신체 사용이 없으면 눈에 보이는 노력의 산물이 쌓이지 않는다. 다독이나 필기를 하지 않으면 허전함을 느끼고, 안정감을 얻지 못한다.

나는 인출과 학습을 병행하며 새로운 전공 과목에서도 좋은 성적을 얻을 수 있었다. 2023년 가을 학기, 양자물리학 과목 ‘포논(Phonons)’을 수강하며 고체 원자의 진동을 모델링하는 방법을 배웠다. 비조화 진동자(anharmonic oscillator)의 기대 위치를 예측하는 문제가 시험에 나올 가능성이 높았다. 처음 시도했을 때는 4쪽 분량의 답안을 작성하는 데 30분이 걸렸다. 나는 현재 값과 실제 값의 차이를 인식하고, 모범 답안처럼 명료하게 인출할 수 있도록 과정을 거쳤다. 검지손가락으로 풀이 과정을 훑으며 채점자가 이해하기 쉬운 형태로 명확성을 유지하면서도 간소화할 부분을 찾았다. 이는 단기적으로 4쪽 분량을 반복해 푸는 것보다 더 많은 정신적 고뇌를 요구했다. 결국 4쪽 분량을 2쪽으로 줄였고, 시험 전날 손을 사용하지 않고 눈을 감은 채로 문제를 풀 수 있었다. 손 근육을 사용하지 않아 같은 시간 안에 더 많은 문제를 풀 수 있었고, 시험 당일 가장 먼저 제출해 가장 높은 점수를 받았다.

여기서 중요한 교훈은 학습과 인출을 병행해야만 학업 성과에 필요한 준비가 완성된다는 점이다. 물리적인 연습량이 줄어 단기적으로는 진전이 없는 것처럼 느껴질 수 있다. 그러나 갓난아이가 단어를 학습하고 인출하는 데 시간이 걸리더라도, 한 번 익히면 자연스럽게 발음하듯이, 인출-학습-병행을 반복해야 교과과정의 ‘원어민’이 될 수 있다. 원어민은 모국어 발음을 쉽게 내뱉는다. 장기적으로 인출에 필요한 노력의 총량도 줄어든다.

나는 유기화학 과목에서도 인출-학습-병행을 적용했다. 유기화학은 미국 의과대학원 진학을 원하는 학생들에게 큰 장애물이다. 암기량이 많아 원하는 성적을 받기 어렵다. 이 과목에서는 매주 쪽지 시험이 있으며, 반응물이 주어지면 합성 과정과 결과물을 손으로 그리는 문제가 출제된다. 대부분 학생은 교과서의 수많은 합성 과정을 노트에 필기하며 시작한다. 그러나 나는 필기를 생략했다. 대신 단원별 예상 시험 문제를 손바닥 크기의 카드에 적고, 뒷면에 모범 답안을 썼다. 초반 개념은 수업과 교과서, 검색을 통해 이해했고, 이해가 완료되면 인출 과정을 반복하며 뇌를 학습시켰다.

이 과정을 반복하면서 합성 패턴과 시험 출제 가능성이 높은 문제가 눈에 띄기 시작했다. 학습된 것과 그렇지 않은 것을 구분해 카드를 주머니에 넣고 꺼내며 반복했다. 단기 기억이 장기 기억으로 전환될수록 상상과 속마음으로 답을 하게 되었다. 외울 필요 없는 공식은 A4 용지에 간단히 정리해 교과서 참조 시간을 줄였다. 기말고사에서 많은 학생이 방대한 분량에 압도되었을 때, 나는 자신이 만든 문제로 과거 내용을 인출하며 이해도를 확인했다. 나의 높은 점수는 뛰어난 암기력이나 기존 실력 때문이 아니었다. 과학고 출신도 아니고, 유기화학을 처음 접했지만 성과에 필요한 질문을 선택하고 인출을 반복해 불필요한 “드리블”을 줄일 수 있었다. 그 결과 교수에게 깊은 인상을 주었고, 다음 커리어에 필요한 추천서도 받을 수 있었다.

연구에서도 인출-학습-병행은 유효하다. 연구 초기 문헌 검토는 교과서 읽기와 같은 과정이다. 논문을 정리하라는 조언도 많지만, 이는 불필요한 드리블일 수 있다. 문헌 검토의 목적은 연구 가설을 세우고, 내 연구가 기존 흐름에서 어떤 위치에 있는지를 인출할 수 있도록 하는 것이다. 논문을 많이 요약해 데이터를 쌓기보다는, 인출 가능한 논리 구조를 갖추는 것이 중요하다. 논문 읽기는 엑셀 실무서를 읽는 것과 같다. 머릿속에서 기능을 인출할 수 있다면 검색을 통해 언제든 찾을 수 있다. 마인드맵처럼 간단한 노트를 활용해도 무방하지만, 인출에 도움이 되지 않는 행위는 피해야 한다.

대부분 사람은 성과의 본질을 꿰뚫지 못하고, 생존 기반의 원시적 본능에 굴복한다. 시험의 본질인 인출에 부적절한 노트를 만들고, 불필요한 학습에 시간을 낭비한다. 굳은살과 수면 부족을 자랑스럽게 여기며 몸을 혹사한다. 자신의 연구가 왜 매력적인지 인출하고 방향을 재평가하는 것보다 논문을 읽고 정리하는 데 시간을 쓴다. 메시는 육체적 노동을 줄이고, 철학자처럼 경기 동안 걷고 자기비판과 고뇌를 마다하지 않았다. 뛰어난 신체 조건이 아님에도 20년 가까이 큰 부상 없이 활약하며 기적을 이뤘다. 이는 신체 사용을 최소화했기 때문이다.

우리는 메시처럼 신체의 사용을 최소화하여 성과를 얻을 수 있도록 접근해야 하며 궁극적으로 자신의 업무와 연구 분야의 메시가 되어야 한다

• [1] 산책 메시, 왜 항상 걷고 있나요? (spotvnews.co.kr)

• [2] 경기 중 쉬엄쉬엄 걸어 다니는 메시 (sportalkorea.com)

• [3] 많이 걸어라, 걸어야 보인다 (xportsnews.com)

2025년 4월, 160 Claremont Ave, 뉴욕에서

인간의 본능을 활용한 단기 및 중장기 목표 달성 방법

인간은 목표 달성에 중요한 업무를 종종 미룬다. 업무가 유발하는 고통을 잠시나마 피하고자 하기 때문이다. 하지만 마감일이 다가올수록 집중력이 높아진다. 결과를 상상만 해도 고통이 느껴지기 때문이다. 인간은 더 강한 고통을 피하기 위해 마감을 앞두고 업무에 몰입하게 된다.

단기 목표를 이루기 위해 고통을 활용하는 것은 효과적인 방법이다. 나 역시 대학원 지원을 위해 높은 학점을 받아야 했고, 이 목표를 위해 일종의 경쟁 모임을 만들었다. 친한 친구 셋이 함께 공부하며, 시험 평균이 가장 낮은 사람이 밥을 사는 규칙을 만들었다. 실시간으로 점수를 공유하며 서로 자극을 주고받았다. 수치심이라는 고통을 피하고자 싫어하는 과목도 더 열심히 하게 되었고, 결국 우리 모두 졸업할 때까지 상위권 성적을 유지할 수 있었다.

하지만 고통을 중장기 목표에 활용하는 것은 오히려 해롭다. 예를 들어, 매일 단어 10개를 외우기로 하고 못 외우면 다음 날 20개를 외운다고 정했다고 하자. 20개를 외우며 “고통스러우니 미루지 말아야지”라고 다짐할 수는 있지만, 뇌는 이 과정을 반복하며 ‘단어 공부는 고통’이라는 인식을 학습한다. 결국 뇌는 그 계획 자체를 회피하게 된다. 새해 계획이 작심삼일로 끝나는 이유도 여기에 있다.

다행히 인간에게는 고통 회피 외에 또 하나의 본능이 있다. 바로 쾌락을 추구하는 경향이다. 여기서 쾌락은 즐거움, 행복, 만족감을 포함한다. 중장기 목표를 이루기 위해선 꾸준한 노력이 필요한데, 이 과정에서 쾌감을 느끼는 것이 핵심이다.

복학 전, 약 3개월 동안 중국어를 공부할 기회가 있었다. 언젠가 중국어로 책을 읽고 영화를 자막 없이 보고, 여러 언어 사용자들과 자유롭게 소통하는 나를 상상했다. 그 상상이 즐거웠다. HSK 6급 시험을 목표로 단어를 외우기 시작했고, 조금씩 중국어가 들리기 시작했다. 내가 성장하고 있다는 느낌은 공부에 몰입하도록 만들었다. 새 단어를 외우거나 발음을 정확히 익힐 때마다 작지만 분명한 즐거움과 만족이 있었다. 덕분에 고통 없이도 학습을 이어갈 수 있었고, 결국 시험에 합격할 수 있었다.

물론 항상 그랬던 것은 아니다. 과정 자체가 고통스러운 순간도 많았다. 하지만 그럴 때마다 나는 꿈을 떠올렸다. 내가 지금 꿈에 가까워지고 있다는 사실을 인식하려 했다. 그 인식을 통해 “나는 성장하고 있다”는 믿음을 얻을 수 있었고, 그것이 성취감으로 이어졌다. 고등학생 시절 SAT 시험을 준비할 때는 대학생이 되어 뉴욕 맨해튼을 걷는 나를 상상했다. 대학 농구팀에 들어가고 싶은 마음도 있었다. 대학원 진학을 목표로 했을 때는, 가고 싶은 연구실 단체 사진에 내 얼굴을 합성해 책상과 바로 뒤에 있는 냉장고 문 앞에 붙여두기도 했다. 그렇게 시각화된 꿈은 동기 부여가 되었고, 고통스러운 과정에서도 작은 성과를 마주할 때마다 “나는 앞으로 나아가고 있다” 있음을 느낄 수 있었다. 동반된 감정이 고통을 상쇄시켜줬다. 때로는 고통을 감수하고서라도 그 과정을 자발적으로 찾아 나서게 했다. ‘노력하는 자는 즐기는 자를 이길 수 없다’는 말이 내게는 실제로 와닿는 경험이었다.

물론 전제 조건은 있다. 꿈을 떠올릴 때 가슴이 벅차올라야 한다. 그래야 고통스러운 과정에서도 성과가 쾌감으로 이어진다. 만약 그런 감정이 들지 않는다면, 지금 세운 목표가 내 삶에 정말 가치 있는지 다시 생각해봐야 한다. “이 꿈을 위해 내 시간을 투자할 가치가 있는가?”라는 질문에 솔직하게 답해봐야 한다.

최근 논문 작성이 진전되지 않아 “왜 손이 안 가는 걸까?” 고민하던 중 이 글을 쓰게 되었다. 지금은 파이썬 코드를 쉽게 공유할 수 있도록 돕는 오픈소스 소프트웨어를 개발 중이다. 초기에는 이 프로젝트를 박사 과정 지원을 위한 연구 성과의 일환으로 바라봤고, 그 덕에 재미를 느낄 수 있었다. 하지만 그 목표를 달성하고 난 뒤부터는 논문 작성이 더 이상 ‘전진’처럼 느껴지지 않았다. 오히려 “논문을 쓰지 않으면 지도교수와의 관계가 나빠질지도 모른다”는 식의 부정적인 동기로 접근하게 되었고, 자연스럽게 손이 멀어졌다.

하지만 지금은 새로운 목표가 생겼다. 학계의 오픈소스 개발 분야에서 의미 있는 기여를 하고 싶다. 지금 만드는 이 소프트웨어가 다양한 연구실에서 널리 사용되는 프레임워크가 될 수도 있고, 여러 나라의 연구자들과 협업할 기회로 이어질 수도 있다. 어쩌면 numpy나 pandas처럼 누구나 쓰는 도구로 발전할 수도 있다. 그 가능성을 상상하는 것만으로도 나는 지금 다시 목표를 향해 나아가고 있다는 기분을 느낀다. 창작의 고통조차 감내할 수 있을 것 같다.

인간의 본능을 이해하면, 단순히 행동을 예측하는 것을 넘어, 행동을 설계할 수도 있다. 고통을 회피하도록 유도하면 단기 목표를 달성할 수 있고, 쾌락을 자극하면 중장기 목표를 추구하게 만들 수 있다. 지금 무언가가 잘 풀리지 않는다면, 내 행동이 본능에 어떻게 작용하고 있는지 되짚어보는 것이 필요하다.

2025년 4월, 160 Claremont Ave, 뉴욕에서