What makes a group smarter?
More connections? Not necessarily.
In a study involving 621 participants, researchers explored how people use social information to find and track a moving resource.
Groups could outperform individuals — but only when participants had access to information about how well others were performing. Without those signals, following others could even make things worse.
The key was not simply connectivity. It was the ability to identify useful information, adjust attention and adapt to changing conditions.
This suggests a different way to think about collective intelligence.
Perhaps what matters is not just how capable each member is, or how many connections exist, but how information moves through the network — and how those relationships change over time.
Now consider a larger question.
What happens when AI becomes part of that network, helping humans identify signals while also changing how they collaborate?
The study did not investigate AI. But it opens a question worth exploring: could the intelligence of a human–AI system depend as much on the organization of its interactions as on the capabilities of its individual members?
ØRIN — Start with what is real. Then change the question.
Source: Chirkov et al., Nature Communications (2026).
https://doi.org/10.1038/s41467-026-77762-z
Visuals are conceptual illustrations, not images of the experiment.
