As artificial intelligence becomes increasingly capable, it is changing how researchers work and prompting us to reconsider what it actually means to do research.
Rethinking research in the age of AI
Image
Speaking about AI’s growing role in science, Dr Xingjian Zhang, whose research focuses on the theory of quantum computation and information, argued that AI should be taken seriously as something that is reshaping the research process itself.
Drawing on a research cycle proposed by mathematician Terence Tao, Xingjian described research as fundamentally a social activity. It begins with learning and understanding ideas developed by others and ultimately relies on those ideas being discussed, communicated and accepted by a wider community.
“To me, I think that doing research is a social activity. So, it starts with us learning ideas from others and ends with your words being accepted by others. What AI does is just a tiny part of this enhanced cycle of developing solutions and verification,” Xingjian said.
From this perspective, developing solutions and verifying them, areas where AI is becoming increasingly capable, represent only part of the research process. Asking important questions, understanding new ideas, communicating discoveries and engaging with other researchers remain central.
“This also raises questions about academic equity. Access to the most advanced AI models is not universal.”
Remote video URL
However, Xingjian argued that this does not exclude researchers from making meaningful contributions. Researchers can continue to play crucial roles as explorers who identify important questions and teachers who help others understand new discoveries. He explained that these roles are central to research as a social activity. As AI takes on more of the work of solving problems, these may also become the most valuable contributions researchers make.
AI may also raise expectations around what constitutes a significant research contribution. Tasks that were once substantial areas of research can eventually become routine as technology advances.
As AI takes on increasingly complex problems, researchers may need to focus more on establishing new frameworks, identifying unexplored questions and extending areas of knowledge.
To me, I think that doing research is a social activity. So, it starts with us learning ideas from others and ends with your words being accepted by others. What AI does is just a tiny part of this enhanced cycle of developing solutions and verification.
Above all, Xingjian emphasised the importance of transparency.
“As researchers increasingly incorporate AI into their work, being open and honest about how it has been used will be essential to maintaining research integrity.”
Rather than replacing the researcher, AI may ultimately encourage us to think more carefully about the uniquely human parts of research: asking meaningful questions, sharing ideas, interpreting discoveries and building knowledge together.