Scientific discovery in the age of artificial intelligence

• Published in: Nature (London) Vol. 620; no. 7972; pp. 47 - 60
• Main Authors: Wang, Hanchen, Fu, Tianfan, Du, Yuanqi, Gao, Wenhao, Huang, Kexin, Liu, Ziming, Chandak, Payal, Liu, Shengchao, Van Katwyk, Peter, Deac, Andreea, Anandkumar, Anima, Bergen, Karianne, Gomes, Carla P., Ho, Shirley, Kohli, Pushmeet, Lasenby, Joan, Leskovec, Jure, Liu, Tie-Yan, Manrai, Arjun, Marks, Debora, Ramsundar, Bharath, Song, Le, Sun, Jimeng, Tang, Jian, Veličković, Petar, Welling, Max, Zhang, Linfeng, Coley, Connor W., Bengio, Yoshua, Zitnik, Marinka
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 03-08-2023; Nature Publishing Group
• Subjects: 631/114/1305; 639/705/117; 639/705/531; 706/648; Algorithms; Artificial Intelligence; Data collection; Datasets; Deep learning; Experiments; Generative artificial intelligence; Humanities and Social Sciences; Hypotheses; Knowledge; Labeling; Machine learning; Model accuracy; multidisciplinary; Neural networks; Review; review-article; Science; Science (multidisciplinary); Scientists; Self-supervised learning
• ISSN: 0028-0836; 1476-4687
• DOI: 10.1038/s41586-023-06221-2

Artificial intelligence (AI) is being increasingly integrated into scientific discovery to augment and accelerate research, helping scientists to generate hypotheses, design experiments, collect and interpret large datasets, and gain insights that might not have been possible using traditional scientific methods alone. Here we examine breakthroughs over the past decade that include self-supervised learning, which allows models to be trained on vast amounts of unlabelled data, and geometric deep learning, which leverages knowledge about the structure of scientific data to enhance model accuracy and efficiency. Generative AI methods can create designs, such as small-molecule drugs and proteins, by analysing diverse data modalities, including images and sequences. We discuss how these methods can help scientists throughout the scientific process and the central issues that remain despite such advances. Both developers and users of AI tools need a better understanding of when such approaches need improvement, and challenges posed by poor data quality and stewardship remain. These issues cut across scientific disciplines and require developing foundational algorithmic approaches that can contribute to scientific understanding or acquire it autonomously, making them critical areas of focus for AI innovation.
The advances in artificial intelligence over the past decade are examined, with a discussion on how artificial intelligence systems can aid the scientific process and the central issues that remain despite advances.