RMIT PhD researcher Scarlett Howard set out to test the honeybee on its understanding, marking individual honeybees for easy identification and luring them to a specially-designed testing apparatus.
The bees were trained to choose an image with the lowest number of elements in order to receive a reward of sugar solution.
For example, the bees learned to choose three elements when presented with three vs. four; or two elements when presented with two vs. three.
When Howard periodically tested the bees with an image that contained no elements versus an image that had one or more, the bees understood that the set of zero was the lower number - despite never having been exposed to an “empty set”.
Dyer, a researcher in the Bio Inspired Digital Sensing-Lab (BIDS-Lab) in RMIT’s Digital Ethnography Research Centre, said the findings opened the door to new understandings of how different brains could represent zero.
“This is a tricky neuroscience problem,” he said.
“It is relatively easy for neurons to respond to stimuli such as light or the presence of an object but how do we, or even an insect, understand what nothing is?
“How does a brain represent nothing? Could bees and other animals that collect lots of food items, have evolved special neural mechanisms to enable the perception of zero?
“If bees can learn such a seemingly advanced maths skill that we don’t even find in some ancient human cultures, perhaps this opens the door to considering the mechanism that allows animals and ourselves to understand the concept of nothing.”
One of the problems in the development of artificial intelligence is enabling robots to operate in very complex environments, Dyer said.
“Crossing a road is simple for adult humans, we understand if there are no approaching cars, no bikes or trams, then it is probably ok to cross,” he said.
“But what is zero, how do we represent this for so many complex object classes to make decisions in complex environments?
“If bees can perceive zero with a brain of less than a million neurons, it suggests there are simple efficient ways to teach AI new tricks.”
The research was conducted in both Australia and France and involved many control experiments to validate the findings.
Study co-author, Dr Aurore Avarguès-Weber from the University of Toulouse in France, said: “The discovery that bees can show such elaborated understanding of numbers was really surprising given their tiny brain.”
“Large brains are thus not necessary to play with numbers. This capacity is therefore probably shared by many other animals.”
The paper “Numerical ordering of zero in honeybees” (DOI: 10.1126/science.aar4975) is published in Science today.
Story: Gosia Kaszubska
