Two roads diverged in a wood, and I—
I took the one less traveled by,
And that has made all the difference.
— Robert Frost, “The Road Not Taken”
Though popular in graduation speeches, these lines from Robert Frost’s famous poem likely would lack appeal among ants, even the literate ones. Rugged individualism just isn’t a virtue in the ant world. Ants rely on their colony to find the quickest routes between food sources and the nest; this is a phenomenon known as “swarm intelligence.”
Swarm intelligence represents decentralized computing by the ant colony. It relies on individual ants to find the most efficient paths to and from a food source. Scout ants lay a trail of pheromones to the food source, which routes foragers to these paths. As the foragers travel these paths, they lay more pheromones, amplifying the signal. This is an example of adaptive network design where, over time, the fastest routes tend to become more popular, while longer routes die away. For ants, adaptive network design works because the pheromone they deposit is volatile (meaning it evaporates). Thus, less efficient paths sit idle longer and eventually lose all trace of pheromones, while shorter paths are traversed more frequently and, thus, continuously reinforced with scent.
Ants’ foraging strategy for finding the shortest path to and from a food source inspired scientists to create the ant colony optimization (ACO) algorithm. An optimization algorithm is a set of instructions for finding the best solution to a problem. For example, an algorithm for finding the cleanest spoon might be as follows: (1) Compare spoon 1 and spoon 2. Keep the cleanest. (2) Compare cleanest spoon from previous round to spoon 3. (3) Repeat previous steps until you have compared nth spoon. Finding the cleanest spoon is pretty easy with this algorithm until you have more spoons than particles in the known universe, but even then it isn’t impossible.
ACO falls into a class of algorithms used to solve some of the most difficult computational problems, NP-hard problems. NP (nondeterministic polynomial time) problems yield numerical solutions (meaning you cannot find an equation to describe the solution); NP-hard problems yield numerical solutions that can take more than the lifetime of the universe to compute. (And you thought dial-up was slow!) How a protein folds into a three-dimensional shape and the famous traveling salesman problem (TSP) are both examples of NP-hard problems.
ACO has been used to solve various NP-hard problems. Studying actual ant colonies may lead to new principles for solving optimization problems and inspire enhanced ACO algorithms. In recent years, a group of researchers set out to observe how ants respond to dynamic mazes in order to find the shortest path between the nest and the food source.1 By mapping the graphical solutions to an NP-hard problem known as the Tower of Hanoi (a mathematical game or puzzle), the researchers constructed a maze that could test the ants’ ability to find the shortest route. Even though the maze represented a relatively “simple” problem, there were over 30, 000 different possible paths from end to end of the maze.
Amazingly, the ant colonies were indeed able to find the shortest path through the maze from nest to food source. Additionally, the ants responded to blocks in the minimal path by finding a new alternative minimal path. How quickly the maze was solved could be mapped to two key parameters: colony size and exploration pheromones. Larger colonies found solutions more quickly than smaller colonies. Exploration pheromones mark the paths traveled by ants as they forage the maze. Not surprisingly, when pre-exploration was allowed, a new colony could reconstruct a minimal path more rapidly.
Observing how ants respond to changing mazes may inspire programmers to develop algorithms that offer efficient solutions to dynamic problems, like the closure of an airport, road, or manufacturing plant. Ants, as well as bees and slime molds, offer researchers in the fields of biomimetics and computer sciences the opportunity to reverse-engineer (or crack the secrets of) highly efficient optimization algorithms utilized in nature to benefit an advanced society. Yet, lurking in this quest are some important assumptions that bear on our philosophical understanding of nature.
In looking to biological systems to inspire better designs, scientists are signaling that they expect to find that biological systems are regulated by rational rules and structures. Scientists are, in fact, trying to uncover the rules and structures. They do not expect to discover a hackneyed system with limited applicability; what they hope to find are universal guiding principles encoded into natural biological systems that can be applied to human problems.
While these assumptions and examples do not preclude an evolutionary explanation, they do not fit neatly into the popularized models of evolutionary theory. Instead, biomimetics compels us to ask the question, why would we expect to find intricate, sophisticated designs in nature unless they emanate from a rational source? How can it be that the solutions in nature are not just better adapted, but inherently structured as (or even more) elegantly than the best algorithms yet invented by humans?
Christians should not be surprised to find the natural world is rationally ordered, knowable, and useful because, as Scripture affirms, the Creator of the natural world is a rational being. For example, John 1 describes Christ as “the Word” (logos, which can also be translated as “reason”). Neither should the Christian be surprised to find that God has embedded special knowledge within creation. In Job 12, for example, the Bible implores, “Ask the animals and they will teach you.” The concordance we see between natural biological systems and the biblical description of the Creator add weight to the body of evidence for the Bible’s validity. The study of nature is just one of many disciplines that continue to point to Christ as logos, Savior, and Lord.