AI has been used to test the validity of Müllerian Mimicry, a theory that states separate species lower down on the food chain will evolve similar warning signals - often skin or wing patterns - for scaring off predators.
The machine learning study, carried out on two similar butterfly species, validated the theory as well as pointed to new discoveries.
The team behind the study says this new method "allows discoveries which simply weren't possible before."
RELATED: EVOLUTION OF THE THEORY OF EVOLUTION
Some fear that AI will be able to mimic us to the point that it is capable of conscious thought - it's newfound sentience turning it into an apex predator like Skynet in Terminator.
In a new study by scientists from England and Japan, AI was used to test the way insects mimic each other in order to outsmart predators - useful information for Skynet, no doubt.
The theory of Müllerian mimicry posits that species - often similar ones - mimic each other, or co-evolve, for mutual benefit. For example, if one species of butterfly has a warning pattern on its wings that's effective at warding off predators, another species of butterfly will mimic this pattern - survival of the copycat if you will.
The theory was first proposed by German naturalist Fritz Müller, only two decades after Charles Darwin published On the Origin of Species.
Machine learning with wings
Testing the evolutional similarity of different patterns of different butterfly species would be a painstaking undertaking. The team of researchers found a machine learning solution.
The team, from the University of Cambridge, the University of Essex, the Natural History Museum, UK, and the Tokyo Institute of Technology in Japan used a machine-learning algorithm to test whether butterfly species do indeed co-evolve similar wing patterns for mutual benefit.
"We can now apply AI in new fields to make discoveries which simply weren't possible before," Cambridge University's Jennifer Hoyal Cuthill, study lead author, said in a press release.
"We wanted to test Müller's theory in the real world. Did these species converge on each other's wing patterns and if so, how much? We haven't been able to test mimicry across this evolutionary system before because of the difficulty in quantifying how similar two butterflies are."
Using 2400 photographs (examples above) from the Natural History Museum, the team trained their algorithm - called ButterflyNet - to record variations in patterns of butterfly wings.
ButterflyNet was then set to work on Heliconius butterflies, a prime example of Müllerian mimicry - more than 30 recognizable pattern types make them an ideal candidate.
"We found that these butterfly species borrow from each other, which validates Müller's hypothesis of mutual co-evolution," said Hoyal Cuthill.
"In fact, the convergence is so strong that mimics from different species are more similar than members of the same species."
New patterns, new findings
The researchers also discovered that Müllerian mimicry can generate entirely new patterns in butterflies by combining features from different lineages. Evolution, effectively, looks for the most effective combination of different patterns.
"Intuitively, you would expect that there would be fewer wing patterns where species are mimicking each other, but we see exactly the opposite, which has been an evolutionary mystery," said Hoyal Cuthill.
"Our analysis has shown that mutual co-evolution can actually increase the diversity of patterns that we see, explaining how evolutionary convergence can create new pattern feature combinations and add to biological diversity," noted Cuthill. "By harnessing AI, we discovered a new mechanism by which mimicry can produce evolutionary novelty. Counterintuitively, mimicry itself can generate new patterns through the exchange of features between species which mimic each other."
She continued: "Thanks to AI, we are now able to quantify the remarkable diversity of life to make new scientific discoveries like this: it might open up whole new avenues of research in the natural world."
The researcher's paper was published in the Journal, Science Advances.