More Like this

1. Towards Scientific Discovery with Generative AI: Progress, Opportunities, and Challenges

   https://doi.org/10.1609/aaai.v39i27.35084  

   Reddy, Chandan K; Shojaee, Parshin (April 2025, Proceedings of the AAAI Conference on Artificial Intelligence)

   Scientific discovery is a complex cognitive process that has driven human knowledge and technological progress for centuries. While artificial intelligence (AI) has made significant advances in automating aspects of scientific reasoning, simulation, and experimentation, we still lack integrated AI systems capable of performing autonomous long-term scientific research and discovery. This paper examines the current state of AI for scientific discovery, highlighting recent progress in large language models and other AI techniques applied to scientific tasks. We then outline key challenges and promising research directions toward developing more comprehensive AI systems for scientific discovery, including the need for science-focused AI agents, improved benchmarks and evaluation metrics, multimodal scientific representations, and unified frameworks combining reasoning, theorem proving, and data-driven modeling. Addressing these challenges could lead to transformative AI tools to accelerate progress across disciplines towards scientific discovery. 

   more » « less
2. Applications of Statistical Methods and Machine Learning in the Space Sciences

   Poduval, B.; Pitman, K. M.; and the SOC Team (January 2022, Journal of geophysical research)

   Applications and methods of artificial intelligence (AI) are becoming powerful tools for scientific investigations in the space sciences, particularly in the analysis and interpretation of the plethora of spacecraft and ground-based observations of the near-Earth and faraway space. Applications of Statistical Methods and Machine Learning in the Space Sciences was a virtual conference held during 17-21 May 2021 to bring together experts in AI and in the various subfields of space sciences to further explore the utility of AI, machine learning, and statistical analysis techniques while sharing the current status of applications in these fields. The conference concluded by emphasizing the scope of AI techniques available to the space science community for addressing outstanding problems with great success as revealed in the number of research works presented. 

   more » « less
3. Synthesizing theories of human language with Bayesian program induction

   https://doi.org/10.1038/s41467-022-32012-w  

   Ellis, Kevin; Albright, Adam; Solar-Lezama, Armando; Tenenbaum, Joshua B.; O’Donnell, Timothy J. (December 2022, Nature Communications)

   Abstract Automated, data-driven construction and evaluation of scientific models and theories is a long-standing challenge in artificial intelligence. We present a framework for algorithmically synthesizing models of a basic part of human language: morpho-phonology, the system that builds word forms from sounds. We integrate Bayesian inference with program synthesis and representations inspired by linguistic theory and cognitive models of learning and discovery. Across 70 datasets from 58 diverse languages, our system synthesizes human-interpretable models for core aspects of each language’s morpho-phonology, sometimes approaching models posited by human linguists. Joint inference across all 70 data sets automatically synthesizes a meta-model encoding interpretable cross-language typological tendencies. Finally, the same algorithm captures few-shot learning dynamics, acquiring new morphophonological rules from just one or a few examples. These results suggest routes to more powerful machine-enabled discovery of interpretable models in linguistics and other scientific domains. 

   more » « less
4. ML-SD Modeling: How Machine Learning Can Support Scientific Discovery Learning for K-12 STEM Education

   Zhou, X.; Lyu, H.; Luo, J.; Bai, Z. (February 2023, The 37th AAAI Conference on Artificial Intelligence Workshop - AI4EDU: AI for Education)

   The importance of machine learning (ML) in scientific discovery is growing. In order to prepare the next generation for a future dominated by data and artificial intelligence, we need to study how ML can improve K-12 students’ scientific discovery in STEM learning and how to assist K-12 teachers in designing ML-based scientific discovery (SD) learning activities. This study proposes research ideas and provides initial findings on the relationship between different ML components and young learners’ scientific investigation behaviors. Results show that cluster analysis is promising for supporting pattern interpretation and scientific communication behaviors. The levels of cognitive complexity are associated with different ML-powered SD and corresponding learning support is needed. The next steps include a further co-design study between K-12 STEM teachers and ML experts and a plan for collecting and analyzing data to further understand the connection between ML and SD. 

   more » « less
5. Building Consensus for Integrated STEM and Social-Emotional Development: From Convening to Implementation

   Allen, Patricia J.; Noam, Gil G. (May 2023, Connected Science Learning)

   Roberts, K. (Ed.)

   Uniting social, emotional, and academic development is necessary to ensure all young people develop the thinking and feeling skills needed to succeed in a STEM-driven future. Scientific discoveries and technological innovations are transforming society, and while they may improve our quality of life, they also introduce social and ethical quandaries that young people must be equipped to navigate. For example, there are both opportunities and risks to using artificial intelligence, genetic engineering, and renewable/alternative energy sources. Although the public discourse supports bringing STEM and social-emotional development (SED) together, and demands evaluation and measurement of outcomes, integrated STEM+SED in educational research, practice, and policy is largely abstract and aspirational. Given that jobs of the future will be STEM-focused and will require SED/21st-century skills—such as working in diverse teams, solving complex problems, and persevering through failures—it will be important to implement and measure STEM+SED together at the teaching and learning levels. To move the field toward meaningful integration of STEM+SED practices and skills, we convened a National Science Foundation (NSF)–funded virtual conference: Mapping Connections Between STEM and Social-Emotional Development (SED) in Out-of-School Time (OST) Programs. This conference—attended by 49 stakeholders from STEM and SED research, policy, and practice—focused on identifying the measurable STEM+SED qualities and skills important for youth success and prioritized by both fields. From this conference emerged consensus for a common frame to explore STEM+SED integration—focusing on Active Engagement, Agency, Belonging, and Reflection—which we and our partners are using to generate knowledge, resources, and tools to advance the integration of STEM+SED in formal and informal learning environments. The preliminary findings and recommendations from this conference provide a starting point for areas to prioritize, explore, and set the stage for more rigorous, relevant, andhigh-quality research on integrated STEM+SED. We begin by telling the story of our conference, including our initial focus on OST, our choice of the term “SED,” and our approach. We then show how discoveries during and after the conference push this essential STEM+SED agenda forward in research and practice. We conclude with recommendations by and for researchers, practitioners, and policymakers to promote synergy between the fields of STEM and SED across all learning environments. 

   more » « less