Posts Tagged ‘slime mold’

The Smarts of Slime Mold

October 8, 2008

When looking up the Ig Nobel prizes for Monday’s post, I was intrigued by the mention that slime molds can solve puzzles. As near as I can tell, here’s what was going on:

The Cast of Characters: Your everyday Physarum polycephalum, otherwise known as slime mold. Normally this mold likes to hand out in dark places, like under logs, but according to sites like this it can be useful to scientists because the cells are huge and easy to see. (“Cell” might be the wrong word; apparently lots of cells fuse together into one giant creeping amoeba-like thing.)

The Temptation: Slime mold gets hungry and wants to eat. Normally it eats things like rotten plants, but for illustrative purposes we’ll use a carrot and a cupcake.

The Problem: What if the slime mold is in a maze, and there are two pieces of food? If it were just one, then the mold would check out the entire maze and head over to the nutrients. With two different pieces of food, however, it runs into a problem: it can’t be in two places at once, and it’s not patient enough to eat one item and then move to the other.

The solution: It stretches between the two nutrients using the shortest path. It was for this recognition − that the slime mold was finding the shortest path − that Toshiyuki Nakagaki, Hiroyasu Yamadam, and Ágota Tóth won the Ig Nobel prize in Cognitive Science (for their September 2000 article “Intelligence: Maze-Solving by an Amoeboid Organism,” in Nature magazine). The article requires a subscription, but even without one you can look at a series of photos here. [From the photos, it looks to me like the slime is actually checking out a few different paths, but among them is the shortest.]

The follow-up: Almost three years later, the first two authors (Toshiyuki Nakagaki and Hiroyasu Yamadac) and a third (Masahiko Hara) pointed out that when the mold was trying to get to food in several different locations, it did the same trick of trying to be at each food source at once. Furthermore, “These findings indicate that the plasmodium can achieve a better solution to the problem of network configuration than is provided by the shortest connection of Steiner’s minimum tree.” (from the abstract to “Smart network solutions in an amoeboid organism,” which appeared in the magazine Biophysical Chemistry in January 2004).

But wait, there’s more! Klaus-Peter Zaune created a six-legged robot that was powered by this same kind of mold. Slime mold generally avoids the light; circuits underneath the slime could sense the stuff shying away, and that caused the robot to move away as well. The robot made its appearance at the Second International Workshop on Biologically Inspired Approaches to Advanced Information Technology in January 2006, according to New Scientist.

Finally, a January 2008 article back in Nature (“Cellular memory hints at the origins of intelligence” by Philip Ball) states that Slime Mold remembers the things it’s done:

When the amoeba Physarum polycephalum is subjected to a series of shocks at regular intervals, it learns the pattern and changes its behaviour in anticipation of the next one to come.

Poor slime mold, getting those shocks. I’m sure it would prefer the carrot.