- April 12, 2024
- Research
Recent research from ESMT Berlin shows that artificial intelligence (AI) can effectively supervise humans involved in large-scale research initiatives, taking on roles related to task allocation, organization, and engagement.
Maximilian Köhler is getting the degree of Doctor of Philosophy at ESMT, along with Henry Sauermann, a professor of strategy at the same institution, are studying the capabilities of artificial intelligence not just as an operator who performs certain tasks, such as collecting and analyzing data, but also as a supervisor who oversees the people who perform these tasks. The concept of algorithmic management (AM) offers a transformative approach to conducting research projects, potentially increasing their scale and productivity.
As the complexity and scope of scientific research increase, the study published in Research Policy demonstrates the ability of artificial intelligence to mimic and even surpass the managerial abilities of humans thanks to its instantaneous, comprehensive, and interactive functions. Focusing on algorithmic management in crowdsourcing and public science efforts, Koehler and Sauermann highlight how artificial intelligence effectively fulfills five key managerial roles: allocating and assigning tasks, providing guidance, orchestrating activities, facilitating motivation, and facilitating learning.
The duo’s research methods included reviewing online materials, talking to project managers, AI developers, and participants, and participating in projects firsthand. These methods allowed us to identify projects that use algorithmic control, understand how AI fulfills management roles, and assess where AM can be most useful. The apparent increase in the use of AM underscores its potential to significantly improve research efficiency. “Artificial intelligence has now evolved to the point where it can dramatically expand the scale and productivity of scientific endeavors by controlling complex, large-scale tasks,” notes Koehler.
Comparing projects using AM to a wide range of initiatives, the results show that projects managed by AM tend to be larger and associated with platforms that offer AI tools. This indicates that, while AM can facilitate the scale-up of projects, it also requires robust technical support, which may be difficult for individual initiatives to achieve. These trends indicate changes in the competitive dynamics within research and have implications for funders, digital research platforms, and large research organizations, including academic and corporate R&D departments.
However, the rise of AI in management capabilities does not make principal researchers or management staff redundant. Sauermann notes, “With artificial intelligence taking care of the more routine and calculative aspects of management, human leaders can redirect their attention to strategic and interpersonal efforts, such as identifying valuable research areas, fundraising, or cultivating a dynamic organizational ethos.”
