Insects, birds, fish – they're not the most intelligent of beasts, right?
Creepy crawlies or aquatic creatures are the ones humans feel furthest from; unlike many mammals, they don't engage behaviour that is easy to anthropomorphise. We rarely think that an ant is endearing, see a human likeness in a starling, or signs of intelligence in a trout.
The way they swarm or flock together doesn't usually get good press either. Marching like worker ants might be a common simile for city commuters, but it's a damning, not aspirational, image. Yet a new school of scientific theory suggests that these swarms might have a lot to teach us. The way natural groups, such as ant colonies, behave can offer us problem-solving techniques applicable to business, political or social situations.
American author Peter Miller, a senior editor at National Geographic, investigates what we can learn from animal collectives in his new book, Smart Swarm. "I got hooked on collective behaviour when I wrote an article on 'Swarm Theory', in July 2007," he says. "I knew before I wrote this book that nature was full of intelligence, but I didn't fully understand how interconnected everything was until I talked to biologists about collective behaviour."
He explains, "I used to think that individual ants, for example, knew where they were going, and what they were supposed to do when they got there. But Deborah Gordon, a biologist at Stanford University, showed me that nothing an ant does makes any sense except in terms of the whole colony. Ants serve the same function as components of a colony that our hearts, bones, and brains serve as components of our bodies. Which makes you wonder if, as individuals, we don't serve a similar function for the companies where we work, or the communities where we live. If that's true, then everything we do, whether we know it or not, makes some small difference to everyone else in our group."
Ants aren't intelligent – by themselves. Yet as a colony, they make wise decisions. And as Gordon discovered during her research into red harvester ants in North America's Sonoran Desert, there's no one ant making decisions or giving orders. They use self-organisation – a bottom-up approach where very simple interactions between individuals develop into complex group movements.
Take food collecting. No ant decides, "There's lots of food around today; lots of ants should go out to collect it." Instead, some foragers go out, and as soon as they find food, they pick it up and come back to the nest. At the entrance, they brush past reserve foragers, sending a "go out" signal. The faster the foragers come back, the more food there is and the faster other foragers go out, until gradually the amount of food being brought back diminishes. An organic calculation has been made to answer the question, "How many foragers does the colony need today?" And if something goes wrong – a hungry lizard prowling around for an ant snack, for instance – then a rush of ants returning without food sends waiting reserves a "Don't go out" signal.
As Gordon told Miller, "No ant understands its own decisions. But each ant's decision is linked to another ant's decision, and the whole colony changes." But could such decentralised control work in a human organisation? Absolutely.
Miller visited a Texas gas company that has successfully applied mathematical formulas based on ant colony behaviour to "optimise its factories and route its trucks". He explains, "If ant colonies had worked out a reliable way to identify the best routes between their nest and food sources, the company managers figured, why not take advantage of that knowledge?" Delivering to more than 15,000 customers across America, with a fleet of 700 vehicles – and juggling variables such as the price of energy and fluctuating customer demand – the company struggled to maintain smooth running and a constant supply to customers.
So a team of "complexity scientists" came up with a computer model, based on the self-organising principles of an ant colony. Data is fed into the model about deliveries needing to be made the next day, as well as things like weather conditions, and it produces a simulation determining the best route for the delivery lorries to take. Each driver – the "ant" in this model – would then be responding to its local, immediate environment and needs, rather than to a fixed company schedule. The programme was estimated to bring cost savings of $20m (£13m) a year.
Miller explains that he first really understood the impact that swarm behaviour could have on humans when Tom Seeley, a biologist at Cornell University, described how he used decision- making rules from honeybee colonies to run his faculty meetings. "He convinced me that the "wisdom of the hive" he had observed in bees was based on the same principles James Surowiecki described for people in his bestseller The Wisdom of Crowds. That opened my eyes to all kinds of swarm intelligence, both in nature and in society."
The honeybees that Seeley studied choose as a group which new nest to move to. First, scouts fly off to investigate multiple sites. When they return they do a "waggle dance" for their spot, and other scouts will then fly off and investigate it. Gradually traffic builds up towards one site – which is almost always the best – until the decision is democratically made. It works a bit like a stock market. If scouts watch a dance, investigate a nest site, and are convinced, they can come back and dance for it themselves. The more dancers, the more scouts visit; the more scouts visit, the more likely it is that a site's "stock" will rise, and the more likely it is to be chosen.
The key to a good, quick decision is diversity of knowledge. Many bees go out, but none tries to compare all sites. Each reports back on just one. The more they liked the nest, the more vigorous and lengthy their endorsing waggle dance, and the more bees will choose to visit it. It's a positive feedback system that ensures that support for the best site snowballs and that the best decision is made, collectively. Humans, too, can make clever decisions through diversity of knowledge and a little friendly competition.
"The best example of shared decision-making that I witnessed during my research was a town meeting I attended in Vermont, where citizens met face-to-face to debate their annual budget," explains Miller. "For group decision-making to work well, you need several factors to be in place: you need individuals with a variety of talents and sources of information taking part, who have some incentive to participate; you need a way to sort through the various options they propose; and you need a mechanism to narrow down these options."
Citizens in Vermont control their municipal affairs by putting forward proposals, or supporting others' suggestions, until a consensus is reached through a vote. As with the bees, the broad sampling of options before a decision is made will usually result in a compromise acceptable to all. The "wisdom of the crowd" makes clever decisions for the good of the group – and leaves citizens feeling represented and respected.
The internet is of course an area where we are increasingly exhibiting swarm behaviour, without any physical contact. This too has its precedent in nature: Miller compares a wiki website, for example, to a termite mound. Indirect collaboration is the key principle behind information-sharing websites, just as it underlies the complex constructions that termites build.
Termites don't have an architect's blueprint or a grand construction scheme. They simply sense changes in their environment, as, for example when the mound's wall has been damaged, altering the circulation of air. They go to the site of the change and drop a grain of soil.
When the next termite finds that grain, they drop theirs too. Slowly, without any kind of direct decision-making, a new wall is built. A termite mound, in this way, is rather like a wiki website. Rather than meeting up and talking about what we want to post online, we just add to what someone – maybe on the other side of the world – already wrote. This indirect knowledge and skill-sharing is finding its way into the corridors of power. The CIA has its own collaborative network, Intellipedia, allowing staff to build on each other's intelligence. "Quite a few organisations have been experimenting with ways to tap into the collective intelligence that exists in their groups, in an effort to make better decisions or react more rapidly to changing circumstances," says Miller. "Just think about how many companies you know that are using wikis to promote collaboration, or social media to strengthen informal networks among workers or members."
Swarming behaviour may even be at the base of our prized human intelligence. Miller points out the similarities between swarms and brains: "like ant colonies, [our brains] function through massive parallel-processing. By itself, a neuron in the brain can't accomplish much, but when 100 billion of them are networked together, they can write a book about other swarms."
'Smart Swarm: Using Animal Behaviour to Organise Our World' is published by Collins (£18.99). To order a copy for the special price of £16.99 (free P&P) call Independent Books Direct on 08430 600 030, or visit Independentbooksdirect.co.ukReuse content