This year's keywords for Korea's scientific community are "restoring the research ecosystem" and "artificial intelligence (AI)." After controversy over cuts in 2024, the national research and development (R&D) budget has entered a recovery phase. The National Assembly set the 2026 national R&D budget at 35.5 trillion won. That is up 5.9 trillion won (19.9%) from 2025 (29.6 trillion won).
The question is shifting from "how much did it increase" to "what will count as results, and how will we make discoveries." The government is looking for the answer in AI. Bae Kyung-hoon, Deputy Prime Minister and Minister of the Ministry of Science and ICT, in the 2026 work plan, put forward "AI research colleague," which collaborates across the entire research cycle, and AI foundation models in science and technology, setting a goal of delivering Nobel Prize–level results by 2030.
But the assessment of a Nobel laureate I met on the 11th of last month was subtly different. Tim Hunt (83), the 2001 Nobel Prize in Physiology or Medicine laureate and honorary professor at the Francis Crick Institute, who visited for a lecture at Yonsei University, said, "The achievements of AI 'AlphaFold' in protein structure prediction are astonishing, but science is not a knowledge business—it is a business that deals with ignorance. Nobel Prizes often come from what we don't know, from questions people thought were theoretically impossible," and added, "AI does not explore the unknown on our behalf."
Professor Hunt is a British biochemist who discovered the key protein "cyclin" that regulates the cell cycle, the process by which cells divide and grow, and received the Nobel Prize in Physiology or Medicine in 2001 with Leland Hartwell and Paul Nurse.
He came to Korea to give a talk at the Yonsei University College of Life Science and Biotechnology Global Lecture Concert, established with a donation from Chung Hye-shin, an honorary professor at Hannam University (former chief strategy officer of Alteogen). The lecture is planned to be held annually starting this year to enhance students' exchanges with global scholars. The following is a Q&A with Professor Hunt.
◇ 25 years after the Nobel Prize… "Money" and "80 lectures" were the biggest changes
-It is the 25th anniversary of your Nobel Prize. What was the biggest change in your life?
"To be honest, the biggest change was money. My share of the Nobel Prize money was roughly equivalent to four years of net salary. And in the first year after the award, I gave an extraordinary number of lectures. I counted 80. Excluding vacations, that meant two to three lectures a week. Demand was so high that sometimes I mistakenly double-booked two or three places. Even I was taken aback."
-How was it as a researcher? It seems it would not have been easy to produce the next achievement after the Nobel Prize.
"It wasn't easy. After receiving the Nobel, I was acutely aware that I would 'never make a discovery that important again.' Because of that, I didn't know what to do and felt lost."
-How did you find your research flow again?
"I was busy after the award, and it took time to return to research. Then a truly excellent Japanese postdoc joined me late in my career, and we set sail on a new voyage of discovery. A good colleague and new questions were key."
-What was the new topic you dug into?
"When a cell enters mitosis, many proteins are massively phosphorylated, but for division to end, those phosphates must be removed again for the next cycle to begin. At the time, it wasn't well known how this 'removal process (dephosphorylation)' was controlled. We happened to hit on an important control mechanism at that point."
Professor Hunt explained that ending mitosis requires "large-scale dephosphorylation," in which phosphates fall off many proteins at once, and that this process is triggered by a brief signal of calcium ions. While visiting San Francisco, he heard from another researcher about a "reagent that keeps the mitotic state prolonged," which prompted him to make the reagent himself and experiment; he then captured the phenomenon that when calcium was added, proteins briefly dephosphorylated and then reverted.
Based on this, he suspected the existence of a dephosphorylating enzyme switched on by calcium, and said that through inhibition experiments and more, he came to see more clearly that exiting mitosis is controlled not by a single switch but by different enzymes acting sequentially.
–You stress "serendipity," but in the end isn't it an ability to interpret?
"Correct. Serendipity isn't everything. I don't think I'm an overwhelmingly brilliant scientist, but I do think I'm pretty good at noticing the meaning of small clues. When a clue appears, you must grab it. You never know when it will come."
◇ "Nobel-level questions? Tackle what once looked 'theoretically impossible'"
–What is a "good question" from your long research career?
"It should be interesting, important, and allow me to make an important contribution to the world. For example, 'How does the brain work?' looks like a very good question on the surface, but in practice it may not be. It's ambiguous when you can say it's been 'solved.' I like questions where what the answer is and when you can finish are relatively clear."
–What kind of questions do you think produce Nobel-winning research?
"A good path is to do what people think is theoretically impossible. That's what happened with my important discovery. I discovered cyclin, known as the 'disappearing protein,' but in fact the technology to reveal that protein had already existed for 10 years. But unless you look over time, you can't know that it disappears. I did a simple experiment, and after confirming it disappeared, I knew in a day that 'this is important.' At the time, such a phenomenon was considered theoretically impossible."
-In Korea, there is talk of investing in AI to produce Nobel-level results. Do you think AI helps identify questions that look theoretically impossible?
"That's an interesting question. One of AI's big successes in biology is protein structure prediction programs like AlphaFold. I've heard they are extremely useful. But AlphaFold predicts protein structures based on a vast trove of accumulated experimental structural data. It is very clever, but in essence there is an element of relying on analogy with accumulated past data rather than understanding entirely new principles, so you have to verify with experiments.
Discovering anew questions that look theoretically impossible is another matter. There is too much we do not know about the behaviors of cells, proteins, and organelles. So at this stage, AI seems limited in exploring the unknown in our stead or in designing experiments to find breakthroughs. Moreover, current AI still makes mistakes and can be convincingly misleading."
-Is it risky for the government to focus on AI?
"Doing AI research itself is of course worthwhile. It's the technology of the era. But we should be careful about expecting AI to solve everything. For example, when you see AIs like DeepSeek, ChatGPT, and Gemini give different answers, that in itself may be a signal that rather than a single fixed 'right answer,' verification is needed and trust issues remain.
나는 AI를 도구로는 많이 쓴다. '인천공항 가는 다음 기차 시간' 같은 걸 물으면 꽤 잘한다. 하지만 미지의 세계를 탐사하는 연구에서는, 이미 잘 확립된 지식 위에서 훈련된 특정한 경우를 제외하면, AI가 해줄 수 있는 일이 어디까지인지 아직은 신중하게 봐야 한다."
◇ "한국 과학계, 기초·응용 균형, 독립성, 실패를 견디는 문화 필요"
-한국의 과학 정책 결정자에게 조언한다면.
"과학은 인류에 엄청난 혜택을 줬지만, 비싸고 낭비가 많다. 대부분의 탐색은 큰 돌파구로 이어지지 않는다. 그런데 정부가 '실용 문제만 풀어라'고 하면, 그건 과학을 죽이는 길이다."
-대안은 무엇인가.
"기초 연구와 응용 연구의 균형이 필요하다. 발견이 나오면 그걸 개발·상용화할 사람도 필요하다. 응용을 누가 맡아야 하느냐는 문제인데, 기업이 더 잘할 수도 있다."
-'기초의 시간'이 길다는 걸 보여주는 사례가 있다면.
"메신저 리보핵산(mRNA) 백신이 좋은 예다. 팬데믹에서 생명을 구한 대단한 성과지만, 그게 가능해지기까지는 50년 가까운 고통스러운 축적이 있었다. 수많은 사람이 작은 문제를 하나씩 해결하며 쌓아 올린 결과다. 심지어 어떤 핵심 연구자(2023년 노벨상을 수상한 카탈린 카리코)는 한때 제대로 평가받지 못하기도 했다."
-하지만 정부의 과학 정책은 선거 주기처럼 짧은 주기로 움직인다. 과학 정책은 어떻게 다뤄야 하나.
"과학 정책의 성패는 1~2년 안에 판정되지 않는다. 몇 년, 때로는 수십 년 뒤에야 '무엇을 살렸고 무엇을 망쳤는지'가 드러난다. 그래서 정부 안에는 과학을 '잘 아는 사람'만이 아니라, 과학의 언어를 정치의 의사 결정 언어로 번역해 줄 사람이 필요하다. 정치인들은 대개 과학자가 아니다. 단기 성과만 좇는 선택이 연구 생태계를 망치지 않게 하려면, 그 간극을 메우는 과학계 조언자가 필수다."
-한국이 과학 성과를 키우려면 문화적으로 바뀌어야 할 것이 있을까.
"사실 한국의 모든 사정을 알지 못해 단정하기 어렵다. 다만 내가 경험한 과학의 조건은 분명하다. 첫째는 독립성이다. 조금은 장난기 있고 조금은 말을 안 듣는 사람, 시켜서 하는 게 아니라 스스로 호기심으로 움직이는 사람이 필요하다.
둘째, 좋은 질문을 고르는 문화와, 실패해도 계속 시도하게 만드는 분위기다. 그리고 아주 현실적으로 말하면, 뛰어난 사람들을 알아보고 재능 있는 연구자들이 서로 만나게 하는 환경이 도움이 된다. 나는 과학이 완전히 평등하다고 생각하지 않는다. 어떤 과학자는 다른 과학자보다 더 뛰어날 수 있다. 재능 있는 사람을 모아야 진짜 진전이 생긴다."
헌트 교수는 과거 기초과학연구원(IBS) 자문위원으로 활동한 경험도 언급하며, 재능 있는 연구자들이 모여 일할 수 있는 구조의 중요성을 강조했다.
–결국 'AI냐 아니냐'보다, 과학이 무엇을 하는 일인가로 돌아오는 것 같다. '과학은 지식 산업이 아니라 무지를 다루는 산업'이라는 말을 다시 설명한다면.
"사람들은 과학이 우리에게 확실성과 안정감을 준다고 생각하곤 한다. 나는 반대로 본다. 과학은 모르는 것을 찾는 일이다. '모르는 게 남아 있는 세계'가 좋다. 문제가 있어야 할 일이 생기기 때문이다."
–만약 AI가 언제나 정답을 뱉어내는 세상이 온다면 과학자는 무엇을 하게 되나.
"끔찍하다. 우리는 일을 멈춰야 할지도 모른다. 다들 과학자가 아니라 팝 가수가 되어야 할 수도 있다(웃음). '모른다'가 남아 있는 게 오히려 건강하다고 생각한다."