Ants vs. Humans: Putting Group Smarts to the Test

And here is the actual paper (linked from the article as well): https://www.pnas.org/doi/10.1073/pnas.2414274121

I remember reading somehwere about the optimal size for human groups to be efficient, but I can't recall the source. If anyone has any pointers, I'd appreciate it!

There was some nuance here. They blocked the abilities of the humans to communicate when achieving that result. So no talking or even gestures, only communicate like ants do.

Since we are so used to communicate when collaborating, it's unsurprising that we fail at collaborating when that is taken away.

This is a very cool study. But the conclusion is probably more nuanced.

That seems to be university PR doing its "magic." The actual study is much more interesting: the humans weren't allowed to speak to each other, and pheromones wouldn't help the ants solve the problem, so both groups were communicating through haptic feedback. Ants do this naturally and demonstrated swarm intelligence behavior by "going with the flow", but the humans kept working at cross purposes by trying to implement a complete solution without coordinating the details.

I agree with the overall conclusion, even if it's phrased misleadingly: human collective intelligence is primarily about individual intelligences accessing group knowledge rather than groups working together to tackle complex problems beyond individual comprehension. Ants are not individually capable of understanding the piano-mover problem at a basic level; research administrators are generally capable of understanding the work of individual researchers, they just don't have the time to digest all the details.

> If it were true wouldn't all democratic societies be in danger.

Democracy is more or less in a permanent state of crisis. This has been discussed thoroughly since the time of Athens, and certainly Rome. The late days of the republic were characterized by squabbling over the specifics of who got to lobby voters, how they were allowed to, and where they were allowed to. For instance they have laws on the books dictating the physical structure of the buildings that people voted in to ensure that the rich couldn't basically station people in the halls leading to the ballots to purchase votes or physically intimidate voters. This is also reflected in the sudden populist turns of eg the Gracchi brothers and Caesar himself.

It's also true of the American republic. Self-conception of us as an egalitarian democracy is still around at best a century old, and more accurately around sixty years old. And we remain extremely far from being an obviously healthy democracy. Of course, the state has vacillated between actions you could argue are wise and those that are clearly not, before and after these divides.

I really would be very cautious at viewing democracies as reflecting of "wisdom". We often can (and often do) come to consensus that is extremely ill-advised from the perspective of the needs of the populace. Democracy is more or less permanently perched on the tension between the will and needs of the constituents which are often at blatant odds with each other. There's a reason why the Philosopher King has held such a cultural weight through the millennia. At best democracy is a best-faith effort to approximate wisdom through consensus—sometimes with better faith than other times.

Given the coordination/cooperation aspect of the problem, this isn’t really the “wisdom of the crowd” as I’ve always understood it.

Something like estimating the number of beans in a jar is a good fit, since there is only one layer of perception to agree on and no coordination required.

This experiment as described seems closer to “design by committee” with (predictably) similar results.

> If it were true wouldn't all democratic societies be in danger.

All democratic societies are in danger -- because the wisdom of the crowd does not have the capability to handle scenarios where the crowd has sufficiently great power.