Maximilian Koehler, a PhD candidate at ESMT, and Henry Sauermann, a professor of strategy at ESMT, delve into the dynamic role of artificial intelligence (AI) within scientific research. Rather than merely functioning as a “worker” executing specific research tasks like data collection and analysis, AI is explored as a “manager” overseeing human workers engaged in these tasks. This concept, known as algorithmic management (AM), heralds a significant paradigm shift in research project execution, potentially enhancing scalability and operational efficiency.
As the landscape of scientific research becomes increasingly intricate and expansive, Koehler and Sauermann’s study highlights AI’s capacity not only to replicate but also to potentially surpass human managers. This is achieved by harnessing AI’s instantaneous, comprehensive, and interactive capabilities. By examining algorithmic management in crowd and citizen science contexts, the researchers showcase examples where AI effectively fulfills five vital managerial functions: task division and allocation, direction, coordination, motivation, and supporting learning.
Their investigation involved scrutinising projects through online documentation, conducting interviews with organisers, AI developers, and project participants, and actively participating in some projects. This multifaceted approach enabled the researchers to pinpoint projects utilising algorithmic management, comprehend how AI executes management functions, and discern scenarios where AM might offer heightened efficacy.
The proliferation of use cases underscores the potential significance of adopting AM in enhancing research productivity. Koehler asserts, “The capabilities of artificial intelligence have reached a point where AI can now significantly enhance the scope and efficiency of scientific research by managing complex, large-scale projects.”
In a quantitative analysis comparing projects utilising AM with a broader sample, the study uncovers that AM-enabled projects tend to be larger and often associated with platforms providing access to shared AI tools. This trend suggests that while AM facilitates scalability, it also necessitates technical infrastructures that standalone projects might struggle to develop. These findings hint at evolving sources of competitive advantage in research and may hold significant implications for research funders, digital research platforms, and larger research entities such as universities or corporate R&D labs.
However, the integration of AI into management functions does not imply the obsolescence of principal investigators or human managers. Sauermann emphasises, “If AI can assume some of the more algorithmic and routine managerial functions, human leaders could redirect their focus towards strategic and interpersonal tasks such as identifying high-value research objectives, securing funding, or cultivating an effective organisational culture.” This underscores the potential for AI to augment human decision-making rather than replace it entirely, fostering a symbiotic relationship between AI and human expertise in scientific research management.
More information: Maximilian Koehler et al, Algorithmic management in scientific research, Research Policy. DOI: 10.1016/j.respol.2024.104985
Journal information: Research Policy Provided by ESMT Berlin