Division of Ideas: Ants and Crowds

An interesting paper from Arizona State University compared the reliability of choices made by an individual and by the colony. The paper found the surprising result that when a binary choice was easier (one choice much better than another), a single ant outperformed the colony. They found this was due to the fact that ants are likely to trust the word of an ant than check the choice out themselves (after all it is more efficient). It should be noted that the effect was not huge and naturally an easier choice was still more likely to be correct. It was simply that a single ant, checking out both choices for himself was slightly more likely to make the correct choice. I would recommend checking out a summary of it, for example on national geographic (or if you have a subscription check out the paper below).

I don't wish to make a direct comparison with human sociology, naturally our social structure and psychological autonomy is vastly different. However, in a more theoretical and basic sense, I think there are some things we can draw from this. Consider the following thought experiment:

There's a tough choice, lets say: should we give the green light for a certain GM crop (say called moonlight corn)? I'm pretty confident to say that only around 1% of the general public are even vaguely qualified to begin answering the question. Far less than that for people who are directly qualified to answer the specific question of this moonlight corn. So lets say there's 1000 people who know enough about this specific crop; not too unrealistic. Next, say 5 experts think the crop should not be grown. Now consider the dissemination of information about this moonlight corn.

The first amplification of this information would be the wider scientific community, however, say that the percentages are the same for the wider group of relevantly qualified scientists. The next stage would be the media and subject specific NGOs, and herein lies the problem. Again, I stress this is an example, but the amplification process is so efficient (few people informing many) and closed (subject to only a few cultures) that it is open to bias. Here we see real potential for misamplification of information where the people deciding on important decisions aren't the ones who are best informed. This isn't even seeking to demean the general public, we simply don't have the time to properly appraise evidence and decide for ourselves and the crowd decision making process is flawed.


Reference: Ant colonies outperform individuals when a sensory discrimination task is difficult but not when it is easy www.pnas.org/content/early/2013/07/24/1304917110.abstract. Unfortunately it seems Cambridge university doesn't have a subscription, so I couldn't check out the actual paper.

Further Reading: This isn't even considering that people use their world view as a prism to view evidence (confirmation bias), demonstrates that current social systems are not the best way to make decisions about empirical evidence.