Olympic of Artificial Intelligence is all set to happen on June 2019, and it will be called an Animal-AI Olympics. In this competition, Artificial Intelligence will be going to treat like a crow or rat against concrete challenges.
This competition will test the capabilities of Artificial Intelligence against tasks that were originally designed to test animal cognition, in order to find out how close we are with machines that have common sense.
What I mean by machine with common sense is that the machines must be able to generate its very own logic against all odds and challenges in order to solve the problem, in the same way, the crow did, the crow about whom we heard in our school books.
The story of the crow goes like this, there was a thirsty crow who finds a pitcher with a small amount of water beyond the reach of its beak. After failing to push the pitcher over, the crow came up with a different idea.
The crow drops pebbles in pitcher one by one until the water level rises, allowing the bird to have a drink. For many people, the fable might look simple but it showed the superiority of intelligence over brute strength.
The aim of Animal-AI Olympics is to find such machines that can pass this kind of Aesop’s ancient intelligence test.
In June, researchers will train algorithms to master a suite of tasks that have traditionally been used to test animal cognition and the team at Cambridge will run them through 100 tests separated into 10 categories.
This will be the Animal-AI Olympics, with a share in a $10,000 prize pool on offer.
Matthew Crosby, a postdoctoral researcher at the Leverhulme Center, says that at this stage the tests are being kept the secret so that participants can’t teach the agents specific skills before the competition starts.
Though AI has been found highly successful in the various different field, but there is one thing that goes against them. AI that has been trained to find certain disease or to show relative products, works very well in their niche.
But when you apply the same AI systems to a totally different task, they become generally hopeless.
That is why, in the Animal-AI Olympics, the same AI will be subjected to 100 previously unseen tasks, that are typically used as intelligence tests for animals.
The things that Animal-AI Olympics will test through all of these tasks, is how well a machine/agent is able to adapt to diverse environments.
What is being tested here is the capacity to quickly adapt to new situations or translate skills from one type of activity to another, which is a good indicator of general intelligence.
This all tasks will ultimately showcase us a limited form of generalized intelligence a type of common sense that AI will need if it ever wants to succeed in our homes or in our daily lives.
Currently, the application of AI is niche-oriented and that’s why even the competition organizers accept that none of the AI systems will be able to adapt perfectly to every circumstance or post a perfect score.
But they hope that the best systems will be able to adapt to tackle the different problems they face.
There will be about 100 previously unseen tasks, some might be as basic as requiring the agent to retrieve food from an environment with no obstacles or hard as finding a hidden object or walking in dark environment.
According to Crosby, the most challenging aspect of the competition is that the machine will have to be good at all the tests across the board, the winning agent will be the one that shows good performance on average, rather than just an ability to master hard tasks.
The Animal-AI Olympics is the creation of a team of researchers at the Leverhulme Centre for the Future of Intelligence in Cambridge, England, along with GoodAI, a Prague-based research institute.
This competition is part of a bigger project at the Leverhulme Centre called Kinds of Intelligence, which brings together an interdisciplinary team of animal cognition researchers, computer scientists, and philosophers to consider the differences and similarities between human, animal, and mechanical ways of thinking.
More in AI :