Collective problem solving ants vs. humans

We designed a "piano mover puzzle", a geometrical problem where the solver has to maneuver an odd-shaped "piano" through a tight environment. This puzzle can give us some measure of the solver's cognitive skills. Dogs often find these problems difficult.
 
 
Although they are only 2.5 mm big longhorn crazy ants do much better at these sorts of problems. For this work, we designed a much more complex piano movers problem that would be challenging even for these tiny "professionals".
 
 
The nice thing about the piano movers' problem is that it can be prepared at any size. The nice thing about longhorn crazy ants is that they can cooperatively transport large-size objects together. Ant groups did much better than individuals!
 
 
Outside ants, almost no other species can cooperatively transport large loads. A rare exception is humans. We prepared the same puzzle on a larger scale - and added wheels. Individual people used memory and planning to perform quite better than an ant:
 
 
How would people cooperate to solve the same puzzle? We prepared an even larger version of the puzzle so that 26 people could tackle it together. We used two experimental conditions: Restricted communication, similar to what the ants use, and free communication.
 
 
People in restricted communication groups did worse than individual people. This is a stark difference from the ants that, using the same communication scheme, showed huge improvements. In fact, the slowest human groups did worse than the best ant groups.
 
Why do ants improve in groups? We find that, when in a group, the ants acquire a new cognitive skill - emergent short term memory. It is fascinating that they use this emergent memory to solve the maze in a way that begins to resemble the way people tackle the puzzle.
 
Why do people in groups do worse? We find that in our puzzle, people prefer being part of the consensus over their personal well-thought-over opinions. Thus, the group always tries the most evident options first, and this bears a cost.
 
What happened when people were allowed free communication? In this case, group smarts did not exceed the best individual in the group or even matched this person. In fact, the performance of a communicating group of people was similar to that of the average person in the group.
 
To learn more about cooperative transport and problem-solving click here:
https://www.weizmann.ac.il/complex/feinerman/research/cooperative-transport
 
 

And now, especially for kids!

You can do these experiments yourself! Wanna know how?

 
 You can learn all about it in our graphic novel: "The Anteater's Culinary Guide" (in Hebrew, coming out soon in English).
 

Further Reading:

 Dreyer, T., Haluts, A., Korman, A., Gov, N., Fonio, E., & Feinerman, O. (2025). Comparing cooperative geometric puzzle solving in ants versus humans. Proceedings of the National Academy of Sciences, 122(1), e2414274121.