Towards Agentic AI for Science: Hypothesis Generation, Comprehension, Quantification, and Validation
Year: 2025
Type: conference-paper
Abstract: AI models for science [14] have the potential to harness large datasets, accelerate scientific discoveries and transform numerous fields. For example, AlphaFold [13] has accurately predicted protein structures, significantly reducing research time in biology and medicine. Similarly, AI has been used to identify compounds for diseases like COVID-19 [10] and Cancer [21], dramatically shortening the timelines for breakthroughs in healthcare. In mathematics, AlphaTensor [7] has discovered faster algorithms for matrix-matrix multiplication than Strassen's algorithm. AI systems have been applied in fields such as healthcare [10, 21], material science [15], physics [19], mathematics [7], education [16], psychology [18], and software development [20]. However, these models are often developed ad-hoc. In contrast, agentic AI [14] has recently emerged as a Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). WWW Companion '25, April 28-May 2, 2025, Sydney, NSW, Australia © 2025 Copyright held by the owner/author(s). ACM ISBN 979-8-4007-1331-6/2025/04 https://doi.org/10.1145/3701716.3717754 systematic approach to advancing scientific discovery, offering the ability to autonomously perform complex tasks and interact with the researchers.
Authors: Lifu Huang, Danai Koutra, Adithya Kulkarni, Temiloluwa Prioleau, Qingyun Wu, Yujun Yan, Yaoqing Yang, James Zou, Dawei Zhou
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4411549718