Editor's note: The State Council Information Office hosted a group interview on Tuesday themed "Forging Ahead on the New Journey", inviting representatives from the science and technology sector to engage with journalists and answer questions.

In choosing a research path, avoid simply following popular trends or chasing what looks like easy wins. Topics that promise quick publications can be a trap. They drain researchers' time and energy, produce fast results, but can lead to what I call "excellence in mediocrity". To escape this, you may need the courage to take a risk: Can you identify a problem others have overlooked? Are you willing to walk a path no one else has taken? But what if, after years of effort, you still see no clear outcome? This is the time to grind away with patience, stay committed for the long haul, and let time be your ally.

Xu Liangfei, associate professor from the School of Vehicle and Mobility, Tsinghua University

Science takes more than a sense of duty. It takes real passion and interest. For 30 years in grassland research, I've never wanted to do anything else. Yet the deeper I go, the less I seem to know. Here's a small example. I'd worked in Hulunbuir for years and thought I knew every corner of the grasslands. In 2014, two unfamiliar plant species appeared in the sample plots. They bloomed briefly, vanished, and have never returned. Isn't that remarkable? It's the endless mystery of nature that keeps drawing you back. To me, the spirit of a modern scientist rests on two things: responsibility is the root, passion is the leaf. Deep roots let the leaves thrive.

Xin Xiaoping, director of Hulunbuir Grassland Ecosystem National Observation and Research Station in the Inner Mongolia autonomous region

Scientific data acts as a "stabilizer" for artificial intelligence development. By providing large models with principled knowledge grounded in natural laws, it reduces their tendency to generate unreliable or fabricated content, forming the foundation for trustworthy, secure AI that is consistent with natural laws. A key challenge is that high-quality scientific data — essential for AI advancement — is difficult to obtain. It often comes from large-scale advanced scientific facilities or long-term observational studies and requires substantial effort to process. Yet, advances in large models also offer new ways to acquire and refine such data, creating a mutually beneficial cycle between AI and scientific research.

Zhou Yuanchun, deputy director of the Chinese Academy of Sciences' Computer Network Information Center

My understanding of the spirit of scientists in the new era, as embodied by the high-speed rail community, can be summarized in four words: patriotism, innovation, pragmatism, and collaboration. Patriotism is the foundation. It drives us to align our personal aspirations with the national strategies of building a technologically strong and transportation-powered nation. Innovation is the soul. It means daring to venture into uncharted territories and setting the standards that define leadership. Pragmatism is about ensuring the safety of high-speed rail operations through massive data support. Collaboration is the pathway.

Liang Jianying, chief scientist of the China Railway Rolling Stock Corporation

As someone who has worked in aerospace for two decades, what drives me first is a sense of national duty and the aspirations of our people. You've likely seen images of Chang'e lunar probe and Mars rover Zhurong. And you can't miss the bright red flag with its five stars. It's that surge of national pride and historical responsibility when we see the national flag shining in space that keeps us pushing forward. Scientific curiosity is another pull. The idea that every breakthrough in our work could expand the boundaries of human knowledge is deeply compelling.

Wang Qiong, deputy chief designer of the Chang'e 6 mission