LONDON – A new competition will benchmark the generalization capabilities of different AIs against animals to better understand progress towards artificial general intelligence.
Founded by leading AI academics from Imperial College London, the University of Cambridge, and the University of Valencia, the Animal-AI Olympics aims to test AI algorithms on tasks they have never encountered before.
The crux of the competition is to evaluate the progress of AI systems away from ‘narrow’ cognitive tasks towards artificial general intelligence. The benchmark? Animal cognition tasks.
“This idea came out of conversations with animal-intelligence researchers. You can take an animal, put it in an environment it’s never seen before, and give it a problem to solve, like getting through some contraption,” explained Dr. Matthew Crosby, a member of the organizing committee. “Whereas if you train an AI to be great at a specific task, it doesn’t even make sense to put it in a new environment. It won’t even try to solve the problem. It just fails to behave.”
Some of the goals of the competition include:
- Benchmarking current AI algorithms against multiple animal species
- Creating an ongoing benchmark and data repository for artificial cognition
- Determining which aspects of intelligence are challenging for current AI – and which approaches are most promising
- Creating new experiments to feed back into the animal cognition community
“Winning this competition will require an AI system that can behave robustly and generalize to unseen cases. A perfect score will require a breakthrough in AI, well beyond current capabilities,” the competition’s website explains. “However, even small successes will show that it is possible, not just to find useful patterns in data, but to extrapolate from these to an understanding of how the world works.”
The full details of the competition will be released in April along with information about entering. Find out more