Ants use algorithms similar to those of the internet

Researchers are beginning to understand how ant colonies can make complex decisions. It’s best understood, they say, as something like an optimization algorithm:

Scientists found that ants and other natural systems use optimization algorithms similar to those used in engineering systems, including the Internet. These algorithms incrementally invest more resources as long as the signals are encouraging, but quickly retreat at the first sign of trouble. The systems are designed to be robust, allowing parts to fail without harming the entire system. Understanding how these algorithms work in the real world can help solve technical problems, while engineering systems can provide clues to understanding the behavior of ants, cells and other natural systems.

Cold Spring Harbor Laboratory“Deciphering algorithms used by ants and the internet” on ScienceDaily The paper is open access.

The researchers explain in more detail:

The same algorithm used by internet engineers is used by ants when they forage for food. At first, the colony may send out a single ant. When the ant returns, it provides information about how much food it got and how long it took to get it. The colony would then send out two ants. If they return with food, the colony can send three, then four, five, and so on. But if ten ants are sent out and most don’t come back, the colony doesn’t reduce the number it sends to nine. Instead, it reduces the number by a large number, a multiple (say half) of what it sent before: just five ants. In other words, the number of ants increases slowly when the signals are positive, but decreases drastically when the information is negative. Navlakha and Suen note that the system works even if individual ants get lost and parallels a certain type of “additive increase/multiplicative decrease algorithm” used on the Internet.

Cold Spring Harbor Laboratory“Deciphering algorithms used by ants and the internet” on ScienceDaily The paper is open access.

Computer programmers learned a better solution to the Traveling Salesman problem, in part from ants:

Navigation expert Eric Cassell, author of Animal Algorithms (2021), has done much work in this area, especially on how life forms not engaged in abstract ‘thinking’, as we know it, navigate with precision. In his opinion, similar to that of the Cold Spring Harbor authors, something like an algorithm kicks in. It may be possible to look for the way the algorithm is generated in the ant’s neurons or genes.

A bigger conundrum is that the ant (along with many other insects) manages these complex patterns with brains of just 100,000 to a million neurons. (Humans have 3 billion neurons.) It doesn’t seem that the original source of the intelligence that makes the algorithm possible is the ant itself. But that keeps nature interesting.


You may also want to read: For ants, building a bridge is not an “easy” task. Richard Stevens: There is nothing “simple” about designing neural systems and the computer systems to receive and interpret neural sensory inputs. The Quanta piece promotes the idea that software algorithms are “simple.” Rather, it would take an army of engineers to do what ants instinctively do.

and

How do insects use their tiny brains to think clearly? How do they deal with complex behavior involving only 100,000 to a million neurons? Researchers are discovering that insects have a number of strategies for making the most of relatively few neurons to enable complex behavior.

Leave a Comment

%d bloggers like this: