Researcher

1. What does a researcher actually do? 🤔

In one sentence

It feels like a mix of detective + mountaineer + record-keeper. Except here the “culprit” is a truth that no one in all of humanity knows yet, and the mountain you’re climbing is one without even a map. 🔬

A researcher (Researcher / Research Scientist), in one line, is “a person who creates knowledge that didn’t exist in the world before.” More concretely, here’s what they do:

Asking questions: “Why doesn’t this cell die?” “Can we make this material lighter?” It all starts with a question nobody knows the answer to.
Literature review: Reading and organizing hundreds of papers to figure out what others have already discovered (no point re-solving what someone else already solved).
Forming hypotheses: Making a testable guess like “It’s probably because of X.”
Designing & running experiments: Designing an experiment to test the hypothesis and actually running it. In biology or chemistry, with your own hands; in computer fields, with code.
Data analysis: Picking apart the results with statistics and computation. “Is this a real signal, or just chance?”
Writing papers & presenting: Writing up your discovery, submitting it to journals, and presenting at conferences. Writing is half of research. A discovery that hasn’t been verified isn’t a discovery yet.

Let me sketch the mood of “a day in a researcher’s life” (it varies wildly by field):

Morning: You check the results of yesterday’s experiment. Half the time it’s “just as expected!”, the other half it’s “huh? why is it doing that?” That “why?” is actually the truly fun part.
Midday: Lab meeting. You show your results to colleagues, sometimes get torn apart, sometimes get new ideas. Research seems like a solo thing, but it’s really a team sport.
Afternoon: Run more experiments, write code, read papers, or analyze data. And a fair amount of time goes into writing grant proposals (this is a bigger chunk than you’d think).
Evening: If a deadline is near, you polish the paper draft. It’s the work of turning a good discovery into “writing that others can understand and believe.”

The coolest part? Every day you work at the very outer boundary of human knowledge. There comes a moment when you’re the first person ever to see something not yet written in any textbook.

Why this job is awesome ✨

I’ll be honest: research is hard. But the “why” that makes people choose this path is genuinely powerful.

You see something that never existed before, first: There comes a moment when you’re the only person on Earth—in this entire universe—who knows that fact. People endure years for that 5-minute thrill.
The impact is real: Katalin Karikó’s mRNA research became the COVID vaccine and saved roughly 20 million lives in its first year alone. Research that nobody believed in at first. Your curiosity could save millions of people.
You learn for life: Heaven for someone who’s bored by doing the same thing over and over. It’s a field where you have to learn something new every year just to survive.
High autonomy: (Especially in academia, or once you become a PI) you get to decide “which question to solve.” Not many jobs in the world give you that.

There are quieter rewards too:

When a problem that wouldn’t budge for months suddenly clicks with an “aha!” at 2 a.m.
When a researcher on the other side of the planet who read your paper emails you, “Thanks to this, our research moved forward.”
When a junior says, “I decided to go into this field after hearing your talk.”

And right now is a genuinely exciting moment. As AI accelerates literature review, hypothesis generation, and simulation at insane speed, we’re entering an era where a single researcher can do—far faster—what used to take an entire team 10 years.

The cold reality (reality check) ⚠️

If you’re even slightly considering becoming a researcher, you deserve to know the real thing—not the Instagram highlight reel.

The barrier to entry is really long:

4 years of a bachelor’s + (for most real research jobs) a 4–6 year PhD
After that, a 2–5 year postdoc is almost the default
Only then do you finally stand at the starting line of being an “independent researcher”

And the essence of research is failure. Really. 80–90% of experiments don’t work. Confirming that your hypothesis was wrong is just everyday life. You have to be able to endure “didn’t work again today” for months, for years.

What nobody tells you:

Grant proposals often have an acceptance rate of 10–20%. Even with a great idea, if you can’t get funding, you can’t do the research. That’s why a researcher actually spends a lot of time as a “proposal writer.”
Papers get rejected too. Karikó and Weissman submitted their mRNA discovery to Nature and Science and were rejected by both. The world’s top journals rejected Nobel Prize-winning research.
During the PhD and postdoc years, working 60–80 hours a week is common. Yet the pay is often around $50,000–$60,000 a year for a postdoc (in the US).

The economic reality:

During the PhD you’re a student, so income is low (or nearly nonexistent), and the “opportunity cost” is large. While your peers are employed and earning money, you’re still in school.
Permanent academic positions (tenure track) are really scarce. Far more people earn PhDs than there are professorships, to the point where people call it “the worst academic job market in a generation.”

Setting the record straight: most of a researcher’s day is not the movie “Eureka!” moment. It’s reading literature, cleaning data, writing proposals, debugging failed experiments—repeating this doggedly. But at the end of that doggedness, every so often comes a discovery that changes the world.

2. Will this job still be promising in the future? 📈

A reality check on the job market

Here you have to split “researcher” into two branches. The truth is completely opposite for each.

Industry R&D (corporate research labs) is growing. According to the US Bureau of Labor Statistics (BLS), computer and information research scientists are projected to grow about 20% from 2024 to 2034. That’s more than 6 times the average across all jobs (about 3%). Demand for corporate research roles in AI, bio, new materials, and semiconductors is strong. In Korea too, R&D demand is steady at companies like Samsung, SK, LG, and bio firms, as well as government-funded research institutes (KIST, ETRI, etc.).

But academia (university professorships) is the opposite. Tenure-track positions are genuinely scarce, and even postdoc positions are trending downward, creating a severe structural bottleneck of “too many PhDs, no professorships.” That’s why these days, PhDs increasingly head to industry instead of academia.

Bottom line: demand for research itself is strong, but “where you’ll do the research” decides your fate. If you only look at being a university professor, it’s narrow and treacherous; but if you look broadly—companies, national research institutes, startups—there are far more paths.

Will AI replace this job?

This is what Reputo really wants to talk about. Rather than replacing researchers, AI is wholesale changing the way researchers work.

Things already happening:

Accelerated literature review: It used to take weeks to read hundreds of papers. Now AI summarizes the key points, cutting it down to days.
Hypothesis generation: In February 2025, Google’s “AI Co-Scientist” independently derived—in 48 hours—a bacterial gene-transfer mechanism that an Imperial College team had taken 10 years to confirm.
Simulation & structure prediction: AlphaFold solved the 50-year-old protein-folding problem, cutting protein structure prediction from months-to-years down to minutes. It completely changed the pace of drug development.
Experiment design: Self-driving labs are emerging, where AI proposes “running the experiment under these conditions will give you the most information,” in automated experiment loops.

So where does a researcher’s value shift to? It moves to these three places:

The ability to ask good questions. AI finds answers fast, but it still can’t decide what to ask. Questions worth solving, angles nobody has looked at—this becomes the human researcher’s core weapon.
Being a discoverer who uses AI well. Whoever uses AI like an assistant to accelerate literature, hypotheses, and simulations—and spends the freed-up time on the truly creative parts—becomes overwhelmingly more productive.
Reproducibility and verification. AI also tells plausible lies (hallucinations). The work of verifying through experiments whether the hypotheses and results AI spits out are real, and ensuring that others get the same results when they follow along (reproducibility)—this has become more important. The more AI pours out results, the higher the value of the person who can judge “is this actually correct?”

Successful researchers use AI not as a threat but as a superpower. You don’t become “a researcher whose job is taken by AI”—you become “a researcher who takes jobs from researchers who don’t use AI.”

💰 The actual salary

What students always ask: “So… how much do researchers make?”

United States (USD, as of 2026):

PhD student/postdoc: about $50,000 – $65,000 (postdoc basis, roughly 70–90 million KRW) — effectively a “training stage” salary
Early-to-mid-career industry research scientist: average about $130,000 – $138,000 (roughly 180–190 million KRW)
Full range: about $75,000 – $160,000, with a median around $135,000
Big Tech AI research scientist: including stock, even $300,000 and up (a scarce field, so the ceiling is very high)

Korea (KRW, as of 2025–2026):

Starting salary for a PhD at a government-funded institute: about 52 million KRW ~ (+ project incentives and allowances). KIST starting salary is about 46 million KRW, and average annual pay is around 95 million KRW (varies widely by experience and project funding won)
Corporate research labs (master’s/PhD): tend to have higher starting pay than government institutes (there’s even a comparison that government-institute starting pay is only about 60–70% of the corporate level)
A Korean researcher’s salary is made up of “base pay + own-project incentives + other-project incentives + tech transfer/consulting,” so it’s a structure where the better you win projects, the more your pay jumps.

Reality check: research jobs are structured to stabilize late rather than bring “big money right away.” You earn little during the PhD and postdoc years, then it shoots up once you go to industry or become senior. If you want to make money fast, another path might be better; if “I want to solve this question” is your real motivation, it’s a path you won’t regret.

Is it right for me? (self-assessment)

Think of it like a game character build. Research rewards certain stats.

It’s a perfect fit if you’re someone who:

Has insanely strong curiosity (can’t stand not knowing “why?”)
Sees failure as data (when something doesn’t work, you don’t get crushed—you go “oh, so it’s not this” and move on)
Has grit (research is not a sprint, it’s an ultramarathon)
Loves writing and explaining (making others believe your discovery is half the job)
Can do both deep solo digging and team discussion
Tolerates ambiguity (having no answer key is the default)

Honestly, it could be tough if you’re someone who:

Needs fast, certain rewards (research rewards are slow and uncertain)
Only feels secure when there’s a clear answer or manual
Is deeply wounded by rejection and criticism (paper and grant rejections are everyday life)
Absolutely needs a 9-to-6 with a predictable schedule (experiments don’t respect your schedule)

Work-life balance:

Early PhD/postdoc: generally brutal (60–80 hours a week is common, tied to experiment timing)
Senior/stable phase in industry: much better, but it can still get brutal during deadline and project seasons

3. The cold truths you must know: the downsides ⚠️

The reality of work-life balance

I’ll be honest: the training period (PhD and postdoc) is the hardest.

PhDs and postdocs commonly work 60–80 hours a week.
Experiments don’t conform to human schedules. Cells grow even at 3 a.m., measurement instruments have to be booked in line, and once you start an experiment you can’t just cut it off.
When deadlines (papers, conferences, grant submissions) pile up, you get ground down for weeks at a time.

In other words, especially early on, you give up a lot of weekends, plans, and hobbies. You have to go in knowing this isn’t “busy for a moment”—it’s a years-long lifestyle.

Stress and mental health

The pressure of this job is hard to describe until you experience it yourself:

Uncertainty is the default. You live every day carrying “nobody knows whether this will work or not.”
Failure is everyday life. Most experiments fail, and hypotheses are often wrong.
Rejection is everyday life. Grant acceptance rates of 10–20%, papers torn apart by reviewers, then revised again.
The comparison trap. Comparisons like “they got into Nature but I…” eat away at your mind.

In reality, burnout and mental health issues among PhD students and postdocs are treated as a major issue in academia. So “resilience” isn’t just nice to have—it’s survival gear.

Economic reality & opportunity cost

During the 5–6 years of a PhD plus a few years of postdoc, your peers are employed, drawing salaries and building careers. Your “opportunity cost” is as large as that gap.
Postdoc pay is around $50,000–$65,000 a year in the US, and many feel it’s meager relative to a PhD degree.
The reward comes late. It gets better once you go to industry or become senior, but you have to endure the long tunnel until then.

Career risk

The academic bottleneck: Tenure-track positions are absurdly few relative to the number of PhDs produced. If “becoming a professor” is your only goal, it’s a cruel probability game.
Time on market: Academia tends to read it as a weakness if you can’t land a position within a few years of finishing your PhD—“you’ve been on the market too long.” The clock is ticking.
Field volatility: Funding swings with policy and trends. A hot field cools off, and suddenly money pours into a different field.

Stories from people who quit

Things people who left research commonly say:

“I left not because I lacked talent, but because I could no longer endure the uncertainty and instability.”
“After bouncing from postdoc to postdoc, I suddenly found myself in my late 30s with no stable position.”
“I loved the research itself, but I got worn out spending half my time writing proposals just to get funding.”

But there’s a twist: a good number of PhDs who leave academia actually do even better in industry, startups, and data fields. PhD training (the ability to dig deep, define unknown problems, and prove things with data) is a powerful weapon anywhere. Drop the equation “researcher = professor” and the path gets far wider.

Bottom line: If curiosity drives you, you can endure uncertainty and rejection, and you have the desire to “be the first to see the answer to this question”—then research is a path you wouldn’t trade for any other job in the world. If you need fast, certain rewards and a predictable schedule, think it over one more time.

4. The legends of this field 🏆

You probably think the people who changed the world in the history of research were all “elites recognized as geniuses from the start”? Not even close. People who were dismissed, demoted, rejected, and discriminated against—yet held on to their question to the very end—are the ones who ultimately flipped the table. Look at these five stories.

Marie Curie — From a shed-laboratory to two Nobel Prizes

Did you know that Marie Curie did her research without even a proper lab, refining several tons of ore by hand in a leaky, shed-like space?

Born in Poland, she crossed over to Paris in an era when women couldn’t attend university, and studied amid poverty and cold. Together with her husband Pierre, she discovered that uranium ore gave off radioactivity too strong to be explained by uranium alone, and from there she found polonium (named after her homeland, Poland) and radium. And she revealed the fundamental fact that radioactivity comes from the atom itself, not from some molecular arrangement—a discovery that changed the paradigm of physics.

She faced discrimination head-on, too. For the 1903 Nobel Prize in Physics, only her husband Pierre and Becquerel were nominated at first, with Marie left out—even though she was a key figure in the discovery. It was only because Pierre insisted, “Marie’s contribution is equal; if she’s left out, I won’t accept it either,” that she received it alongside him. In 1911 she won the Chemistry Prize as well, for isolating radium, becoming the only person to win Nobel Prizes in two different scientific fields. After Pierre died in an accident in 1906, she became the first female professor at the Sorbonne. She showed us what it looks like to hold on to a good question with grit to the very end.

Katalin Karikó — Demoted four times, then won a Nobel Prize

Did you know that Katalin Karikó was demoted four times at the University of Pennsylvania, and was called “that crazy mRNA lady” by her colleagues?

Hungarian-born, she believed disease could be treated with mRNA. But in the 1990s, nobody believed her. When she couldn’t win funding, the university demoted her in 1995. Her grant proposals were rejected one after another, and the key paper she co-wrote with her colleague Drew Weissman was rejected by both Nature and Science. The world’s top journals rejected Nobel Prize-winning research.

The secret to how she held on? It wasn’t “because she believed her long effort would someday be rewarded with global recognition.” She didn’t care whether she was recognized or not, and just focused on the next experiment. The mRNA technology she developed after moving to BioNTech in 2013 became the foundation of the Pfizer and Moderna COVID vaccines, estimated to have saved roughly 20 million lives in the first year alone. And in 2023, she and Weissman won the Nobel Prize in Physiology or Medicine. Her one line: “Focus on the next thing. That’s all there is.”

Jennifer Doudna — Gene scissors that began with a drawing on a café napkin

Did you know that Jennifer Doudna’s CRISPR research wasn’t actually “research that was useful right away,” but pure curiosity research that started simply because she wondered how bacteria fight viruses?

Born in Washington, D.C. and raised in Hawaii, she fell in love with chemistry even amid a school atmosphere of “what’s a girl doing in science?” After majoring in chemistry at Pomona College and earning her PhD at Harvard, she dug into the three-dimensional structure of RNA. It all began in 2006, when geomicrobiologist Jill Banfield sketched out a CRISPR diagram on a napkin at a café on the Berkeley campus.

She and Emmanuelle Charpentier revealed that the bacterial CRISPR-Cas9 immune system could be turned into “a tool that cuts and edits any DNA you want.” This became a revolution in gene editing—hundreds of labs around the world now use it, transforming genetic diseases, agriculture, and food. The two won the 2020 Nobel Prize in Chemistry, becoming the first case of two women winning a Nobel Prize in science together. Her message is clear: “Basic science that starts from curiosity ultimately changes the world.” Even if you can’t see the immediate use, a good question is worth digging into to the very end.

Demis Hassabis — From chess prodigy to an AI Nobel Prize

Did you know that Demis Hassabis was a chess master (rated 2300) at age 13, worked as a game developer, then earned a PhD in neuroscience before founding an AI company? He didn’t dig just one well—he dug several wells.

Born in London, he relentlessly pursued the question “what is intelligence?” through chess, game design, and cognitive neuroscience. He earned a PhD in cognitive neuroscience at UCL and co-founded DeepMind. And the team he led solved a 50-year-old grand challenge in biology with AlphaFold—the problem of predicting a protein’s three-dimensional structure just from its amino acid sequence.

To grasp how big a deal this is: it used to take months to years to figure out a single protein’s structure. AlphaFold cut that down to minutes, and it predicted and released the structures of nearly all known proteins. The pace of drug development, genetic disease research, and even neglected-disease treatment changed wholesale. For that achievement, he and John Jumper won the 2024 Nobel Prize in Chemistry. An AI researcher, not a chemist, won the Chemistry Prize. What he showed is this: AI doesn’t replace research—it’s a tool that helps researchers solve problems they could never crack in a lifetime. He showed us, ahead of time, what the future researcher looks like.

Donna Strickland — A Nobel Prize for one discovery during her PhD

Did you know that Donna Strickland did the core research that earned her the Nobel Prize while she was a PhD student, and that there was a 55-year gap in between?

From Guelph, Canada, she invented Chirped Pulse Amplification (CPA) in 1985, during her PhD at the University of Rochester, together with her advisor Gérard Mourou. The principle is clever: amplifying a strong laser pulse directly would burn out the equipment, so you stretch the pulse out to lower its instantaneous power, amplify it safely, then re-compress it to crank the intensity back up. Thanks to this technique, laser pulse intensity climbed to the petawatt (10¹⁵ watt) range.

Today this technique is used in LASIK surgery, precision machining, and ultrafast science. If you’ve had LASIK, you’ve benefited from Donna’s discovery. She won the 2018 Nobel Prize in Physics—the third woman to win the Physics Prize, after Marie Curie (1903) and Maria Goeppert Mayer (1963). There was a 55-year gap between the second and third female laureates. Her story tells us two things: first, that even research done as a student can change the world. Second, that diversity in science still has a long way to go—and that’s why there’s a place for you to step in.

5. How do I prepare? 🎯

If you’re still a student (high school / college)

You don’t need to be a “genius.” What you need is curiosity + consistency + tools you’ve made your own.

Foundations to lay down:

Math & statistics: The common language of almost all research. Statistics especially is the key to distinguishing “is this a real signal or chance,” so you absolutely have to get strong at it.
Programming: Python or R is now essential regardless of field. Data analysis, simulation, and using AI tools all come out of here.
The core science of your field: The fundamentals of whatever field draws you—biology, chemistry, physics, CS, and so on.
Reading & writing in English: Papers are almost all in English. Build the muscle to read and write in English early.

Things you can start right now (for real):

Join an undergraduate research lab (UROP): This is the most important. You experience firsthand what real research is like, get recommendation letters, and confirm your aptitude. Don’t be afraid to email professors.
A habit of reading papers: Pick a paper in a field you’re interested in and read one a week. It’s fine if you only understand 10% at first. In six months you’ll see 50%.
Small side projects: Try analysis with public datasets (Kaggle, etc.), or design a simple experiment yourself. The key is internalizing the “question → hypothesis → verification” cycle in your bones.
Get your hands on AI tools early: Use AI for summarizing literature and writing code, but also build the habit of verifying what AI says. This is the core competency of the future researcher.

The goal isn’t résumé-padding. It’s to test-drive the question: “Am I really someone who can grab hold of a problem I truly don’t know for months and enjoy it?”

If you’re switching from another field

Good news: research is a surprisingly transferable path, and other experiences can even become a strength.

Things that transfer well:

Domain knowledge: A nurse with clinical experience moving into medical research, or an engineer moving into materials research—knowing the field is a big weapon.
Data & coding skills: If you come from a developer or analyst background, you connect naturally to computational research. It’s especially powerful in this AI era.
Writing & communication: Half of research is writing. Experience in planning, journalism, or marketing turns out to be surprisingly useful.

Realistic expectations: a true research job (especially in academia) usually requires going through graduate school (master’s/PhD) again. Think of it not as a “quick transition” but as a redesign onto a new track. That said, industry R&D and data science have paths you can enter without a PhD, so before diving into full-time grad school, it’s also a good move to aim for those positions first. And before you decide, definitely get your foot in a lab—even as an intern or on a contract.

Essential skills

Organizing a practical skill stack by priority:

Top priority: Statistics & data analysis. The foundation of all research. (Resource: the statistics courses in Section 6)
Top priority: Programming (Python/R). The common language of analysis, simulation, and using AI.
Top priority: Critical thinking & asking good questions. The least replaceable ability in the AI era. The power to doggedly ask “is this actually correct?”
High: Scientific writing. The ability to make others believe your discovery. (Resource: The Craft of Research)
High: Experiment design & a sense for reproducibility. The habit of checking “will this result come out the same if someone else follows along?”
High: Using AI tools + verification. The ability to accelerate with AI while filtering out hallucinations. This is the differentiator of the next-generation researcher.
Medium: Grit & resilience. The mental toughness to endure rejection and failure. It’s built through training.
Medium: Collaboration & presentation. Research is a team sport. Getting torn apart and learning at conference talks and lab meetings.

6. Learning resources 📚

Online courses (you can start right now)

Coursera — Methods and Statistics in Social Sciences (University of Amsterdam): Build a solid foundation in research methodology and statistics. https://www.coursera.org
Coursera — Johns Hopkins “Data Science” / R Programming: The gold standard intro to data analysis and R. https://www.coursera.org/browse/data-science
edX — HarvardX “Data Analysis: Basic Probability and Statistics”: The best starting point for probability and statistics. https://www.edx.org/learn/statistics
edX — MIT/Harvard statistics & data science courses: Step by step from R to regression to machine learning. https://www.edx.org/learn/data-science
Kaggle Learn (free): Hands-on mini courses for learning Python, data analysis, and ML by doing. https://www.kaggle.com/learn

Books (a researcher’s must-reads)

The Craft of Research (Wayne Booth et al.): How to set up a good research question, gather evidence, and write persuasively. The bible of research regardless of field.
A PhD Is Not Enough! (Peter Feibelman): A survival guide to a scientist’s career—choosing an advisor, choosing between academia vs. industry vs. national institutes, and even research interviews.
Authoring a PhD (Patrick Dunleavy): A practical guide to carrying a PhD thesis from idea all the way to completion and publication.
How to Write a Lot (Paul Silvia): Building the habit of writing consistently without procrastinating. An unexpected must-read, given that half of research is writing.

Sites & tools

Google Scholar (scholar.google.com): The starting point for searching papers. Map out a field by following citation links.
arXiv (arxiv.org): A treasure trove of the latest preprints (papers before formal publication) in physics, CS, AI, and more. The frontline of AI research shows up here first.
PubMed (pubmed.ncbi.nlm.nih.gov): The standard database of life-science and medical research papers.
Connected Papers / Semantic Scholar: Next-generation literature-review tools that visualize the relationships between papers and summarize them with AI. They drastically cut down literature-review time.
NobelPrize.org (nobelprize.org): Free lectures, interviews, and biographies of the legends. The best when you need motivation.

Try it yourself (the most important thing)

Getting your feet wet in a real lab is 100 times faster than courses and books:

Apply to your university’s undergraduate research program (UROP)
Email a professor in a field you’re interested in saying “I’d like to learn in your lab” (you get replies more often than you’d think)
Summer research internships (university, national institute, or corporate R&D programs)
Do your own mini research with small public data and write it up on a blog

Keep this in mind: a researcher isn’t “a person who knows the answers”—it’s “a person who holds on to a good question to the very end.” The more we live in an era where AI finds answers fast, the greater the value of you knowing what to ask. On that journey of turning today’s “I don’t know” into tomorrow’s “I know,” there’s plenty of room for you, too. 🔬✨

TL;DR