Our AI Journey

Unfolding AI's Impact on Science - Powering Breakthroughs and Challenges

Richard Van Noorden is a features editor with Nature's journalism team. The reporters at Nature cover all aspects of science and science-policy - including discoveries, controversies, and politics insofar as it affects science and scientists. They have won many journalism awards for their work and they are editorially independent from the journal staff. The aim is to better inform their readers: primarily scientists and those whose working lives revolve around science, and also the interested public who want to know what's going on in research. The features team, in particular, commissions, edits and writes deeper explorations of research issues by both staff and freelance contributors; they also conduct their own exclusive investigations and surveys. While Richard primarily covers the physical sciences and scientific-community issues such as scientific publishing and misconduct in science, his area also includes data-driven reporting and technology. Over the past year, exploring how AI is changing the practice of science has become a more prominent area.

Learn more about Richard's contributions to the launch of Nature's digital special issues on 'Science and the New Age of AI', directly from him

Author

Richard Van Noorden

Features Editor

Nature Magazine

In 2023 we launched a Nature digital special issue on Science and the new age of AI, in order to dive more deeply into the way AI is changing the practice of science, focusing on generative AI as well as more established machine-learning. This kicked off with a survey of what 1,600 researchers think about AI in science, and has continued with stories on AI in scientific publishing, AI deepfakes, AI in education, and mis-use of AI in research, with further material to come. This goes alongside our reporting on how AI is being used in fields such as materials science, medicine and biology.

I conducted and analysed the survey of AI researchers, with the help of colleagues, and have been commissioning much of our AI-related journalism content. I do this amid a much wider team of Nature reporters and editors who are also reporting, editing and commissioning on AI topics (including a newly-launched Nature Briefing AI & Robotics newsletter). We share expertise and brainstorm ideas together. Nature’s video team, for instance, recorded a video with me on how ChatGPT works.

AI techniques, like any statistical tools, are a double-edged sword in science. They have enabled remarkable discoveries and accelerate scientific work, but are also poorly applied or mis-used. The latest wave of generative AI and foundational models are already helping scientists do their work more efficiently, our survey showed - and are also leading to the pollution of scientific literature with worthless spam. Training AI on scientific data (rather than language) might produce new ways of finding patterns in data faster than a human can, and could lead to a blossoming of discoveries - and also a confusing tangle of unreliable pattern-matching. AI will also change power relations in science: partnerships with big tech firms are likely to become more important to remain at the cutting edge.

Scientists may end up using open-source AI models rather than relying on firms' closed-source products.

The journalism team's work supports Nature's mission statement: "to provide a forum for the reporting and discussion of news and issues concerning science..." and "... to ensure that the results of science are rapidly disseminated to the public throughout the world, in a fashion that conveys their significance for knowledge, culture and daily life." The team’s vision is to explain and communicate AI's role in science and to get past some of the misunderstanding and hype in this area - and ultimately to help researchers apply AI where it can be useful while calling out areas of concern.

Back to overview