黑料福利网

PhD researcher completed his PhD on March 3, 2026 at the Department of Industrial Engineering and Innovation Sciences. He studied how scientific inquiry functions as a shared effort by simulating how communities reason, exchange evidence, and navigate human limitations.

Why now

Science increasingly develops through collaboration. Michelini鈥檚 work shows that scientific progress does not only depend on individual skills and performance but also on how people question one another, express doubt, and learn together. This matters for organizations making choices under uncertainty, from climate adaptation to responsible AI use, as well as for entrepreneurs balancing data, intuition, and time pressure.

Unexpected differences

Michelini鈥檚 simulations reveal that adding more data does not always lead to agreement. When people rely on different background assumptions or evaluate evidence using distinct standards, new information can intensify disagreement. Yet the models also show that bringing diverse perspectives into conversation can keep groups from locking into a narrow view too early. Michelini demonstrates that diversity in reasoning styles and standards supports more open exploration.

Useful imperfection

Bias and limited attention are often treated as flaws to eliminate. Michelini鈥檚 findings paint a more nuanced picture. In his simulations, communities can use these imperfections to support collective reasoning. Individuals defend their own positions as well as they can, while others challenge them. This interaction helps the group advance even when no one reasons perfectly. Michelini shows that collective structures matter just as much as individual traits.

Real world meaning

For entrepreneurs, this work encourages building teams with intentionally varied approaches instead of aiming for uniform thinking. In product development, for instance, tensions between safety, usability, and cost can strengthen decision making if managed well. In public challenges such as health policy or climate planning, maintaining reasonable disagreement can improve the quality of outcomes rather than delay them. Michelini鈥檚 results highlight how structured dialogue can transform polarization into constructive exchange.

What it shows

Michelini鈥檚 research illustrates that the strength of collective knowledge depends on how communities are organized. Differences in standards, productive disagreement, and even imperfect reasoning can support better outcomes when interaction encourages exchange and revision. These insights are relevant for how research programs, innovation teams, and policy groups are designed.

Looking ahead

Agent based models make it possible to study situations that would be difficult or ethically problematic to test in real scientific communities. This creates a research environment for exploring how polarization arises, how discoveries spread, and how consensus forms. The PE group continues to use this approach to understand how science flourishes when people work together not despite but because of their differences.

Matteo Michelini defended his thesis on March 3, 2026.
Title of the thesis: .
Supervisors: Wybo Houkes, .

CURSOR published a , written by by