We are all using it, we are increasingly surrounded by its effects and consequences and we seem to be asking all the same questions and thinking under the same old frameworks. But at least we are thinking about it and by “it” I mean the impact of AI on science. It’s crucial to distinguish between two intertwined yet distinct themes: AI’s impact on science, and the influence of science on AI development. This distinction lays the foundation for understanding AI’s transformative role, particularly within scientific research and leadership of research performing institutions. 

Understanding AI’s impact on science vs. science in AI 

AI’s integration into scientific processes marks a significant leap towards accelerated discovery and innovation. AI aids scientists in deciphering complex data, enhancing predictions, and automating labor-intensive tasks, thus propelling the frontiers of knowledge across diverse domains, from healthcare to environmental sciences. This form of AI application—where tools and systems are developed to support and extend the capabilities of human researchers—is what we commonly refer to as AI’s impact on science. 

Conversely, the development of these AI systems themselves falls under the banner of science in AI. This involves rigorous scientific methodologies to innovate and refine AI technologies, ensuring they are not only effective but also ethical and reliable. The exploration into neural networks, algorithmic improvements, and machine learning models are intrinsic to enhancing AI’s functionalities and safety, a pursuit rooted deeply in scientific principles. 

The state of the art: AI’s role and challenges in scientific innovation 

Recent policy briefs and reports from the European Commission and few ERA stakeholders elucidate the current landscape and the expansive role AI is destined to play in scientific fields. Documents such as the “Mapping ERC Frontier Research on Artificial Intelligence” and “Trends in the Use of AI in Science” offer insights into how AI technologies are currently employed and the potential paths they might carve out in future research endeavors. 

Still other publications highlight several core areas of focus: the need for robust data infrastructures to support AI applications in science, the importance of developing ethical frameworks to govern AI use, and the urgent requirement for policies that foster an environment conducive to innovation while safeguarding fundamental values. Moreover, these documents raise pivotal questions about AI’s role in accelerating the pace of scientific discovery: How can AI be leveraged to address global challenges effectively? What are the ethical considerations inherent in AI-driven research? How can AI enhance the decision-making processes within scientific leadership? 

INESC Brussels HUB Summer School 2024: Leadership in the Age of AI-Driven Science 

At the INESC Brussels HUB Summer School 2024, these themes will take center stage as we explore the impact of AI on leadership within the scientific community. This includes but goes beyond the common understanding of the impact of AI in science, extending it to a domain that rarely takes the center stage: the leadership of research performing organisations in the age of AI-driven science. This is an event with limited and targeted participation, co-organized with the European Commission, as well as NCBR Brussels (the Polish representation of research and innovation institutions in Brussels) and the active support of several university, RTOs and life-science laboratory EU networks. It is designed to dissect how AI is reshaping the roles and responsibilities of leaders in research and innovation (R&I) institutions. The discussion will be centered on understanding how AI can be a powerful ally in strategic decision-making and foresight planning, and how leaders can cultivate an AI-ready culture that balances technological advancement with ethical considerations and human-centric values. 

The Summer School will address these complex dynamics through a participatory model bridging theoretical understanding with practical insights. By engaging with leading experts and harnessing diverse perspectives, it aims to inform and benchmark, to promote mutual learning and, ultimately, to contribute to start equipping current and future leaders with the tools to navigate and shape the evolving landscape of AI in science. A joint report between all the co-organizers and diverse ERA stakeholders will be produced, in the hope of placing this important perspective at center stage of policy, funding and thought leadership processes.