Where there is no gain, loss is not far away

Applying concepts from the world of biology to understanding the development of the Internet has become increasingly more frequent over the last three years. The Web is without doubt a rich ecosystem in constant motion and evolution. Some of its characteristics are particularly thought-provoking. Firstly, when looking at its dynamics from the surface, the first image that springs to mind is that of a colony of fervid internauts who, although they seem to be carrying out their activity in ever growing chaos, never cease to produce organised reference points. Secondly, for some reason, the Web allows a co-operative impulse that is not found elsewhere. This would seem to be characteristic of the activity that goes on in cyberspace. The economy we know is, as yet, incapable of throwing any light on these two questions. On the one hand, how individuals that are so different and belong to such different cultures can find common ground to organise the flow of information they generate. On the other, what are the minimum conditions needed so that this organisation can be supported by co-operative activity? While we wait for human sciences to say something significant on the subject, biology has started to offer some answers, even though they come from the life of ants. However, no-one should think beyond the limits of simulated models and think we are calling them insects.

Brain Goodwin, author of the book How the leopard changed its spots has just published an interesting article in the New Scientist on the “collective intelligence” of ants. Goodwin, like other biologists who work in the same field, wanted to throw some light on the following question: is the behaviour of an ant colony only a sum of the activities of its individual members, or is there a behaviour that just belongs to the colony as a whole which can neither be predicted nor deduced from its parts? Until recently, social biology has dominated this area, clearly endorsing the first part of the question. However, the concept of the colony being a super-organism with an internal life controlled by parameters other than its individuals constituents, is now beginning to emerge. In other words, if the behaviour of these individuals is studied, their behaviour as part of a collective whole can never be predicted. So, the question emerges as to under what circumstances chaotic individual behaviour begins to generate organised behaviour patterns within a colony. The idea that there exists an unpredictable emerging order in natural processes is not new and does not come from biology. We know that water is made up of hydrogen and oxygen, but the atoms of these elements does not explain why water forms a whirlpool when it goes down a drain. For this we need to understand fluid dynamics, and this knowledge is quite beyond individual atoms. In other words, certain phenomena cannot be predicted from their constituent parts. In the same way it is very difficult to predict the organised and co-operative behaviour in Internet from one of its parts, the internauts. To do so we would have to understand the “colony” as a whole. This is so complicated that simulated models would have to be constructed and these too are becoming increasingly complex.

This is what Goodwin along with Ricard Solé (Polytechnical University of Catalonia/UPC) and Octavio Miramontes did: they constructed a virtual ant-hill. Their aim was to answer the questions we ask ourselves when we stop to observe the comings and goings of these insects: how do they manage to look so organised in the midst of such apparently chaotic behaviour? Where does their organising impulse come from? Is it a question of genes alone? The answer to the latter question seems to be negative. As has been recently discovered, over-stimulated, hyperactive ants live together with hedonistic type ants with a clear inclination for “la dolce far niente”, but who, under certain circumstances, throw off their laziness and join their colleagues. What’s going on? What makes them abandon their easy lives as ants and throw themselves into the typical frenzy of the ant-hill? The group of researchers soon found the answer, which will sound familiar to any Internet user. The secret of the colony was in its interaction. When a busy ant passes by a lazy ant, in some way the first interacts with the second and “convinces it” to pull its finger out and get moving.

“That’s all very well but how do ants interact if they don’t have computers, or modems, or even a simple telephone to put to their ears?”, a confused internaut might ask. “Ah, but they have antennae!’ Antennae very sensitive to touch and smell. And, it seems, the necessary interchange of information to begin to weave the colony web is passed through this “interface”. In Goodwin’s virtual ant-hill the ants were simple software agents which moved on a grid that resembled a chessboard. When an active ant passed by a lazy ant the latter was activated and shaken out of its lethargy. The activity of the colony grew in parallel with the activity of the population of active ants. This was due to the greater frequency of stimulation (interactivity) among individuals. When a certain density of active population was reached a rhythmic pattern emerged, and when transferred to a graph, continual, organised waves were observed. The colony started to behave as a super-organism, something which no-one would have guessed from the erratic, chaotic curves generated when the individuals started to move, or when interaction was still very low.

With the simulated model it was also shown that the most important type of interaction for reaching a collective rhythm was that of the active ants with the inactive ones (not active with active). It has just been discovered that this also happens with real ants (Blaine Cole, an investigator at the University of Houston). Another very important factor, which has also been confirmed in real life, is that there is a threshold in the colony which defines how sensitive the ants are to the stimulation of other ants. If this sensitivity is very low, not enough activity waves are propagated throughout the colony to enable a collective rhythm to emerge, regardless of the density of its population. On the other hand, if sensitivity is very high, the group becomes hyperactive, cohesion breaks down and chaos reigns. However, the range of the ‘right’ sensitivity appears to be wide enough to suggest that the rhythmic activity patterns are a powerful consequence of colony life, perhaps its determining factor from the point of view of evolution: in this collective intelligent pattern lies the staying power and competitive advantage of the colony.

What lessons can be learned from “emerging order” and the leap from the chaos of individuals to collective intelligence, in the context of the Internet? Well, I don’t know, each one of us will have delve into the ant-hill of our personality and not forget that unless you interact with others you will not be able to deduce from this personal examination your true collective capabilities. And that’s just the beginning.

Translator: Ruth Belcher