Test your mental skills with an Artificial Intelligence tool called Cognitron

The first artificial intelligence (AI) designed to model human mental skills has been developed by researchers from Imperial College London.

The AI, called Cognitron, has been developed by a team of psychologists, neuroscientists and engineers at the College.

The team are inviting members of the public to take part in citizen science research, where Cognitron will pose brain teasers for them.

Participants are put through a series of customised tests, which takes just over 30 minutes. After the person takes the test Cognitron will tell them how well they did.

Put your intelligence to the ultimate test and see how you fare compared to others.

The information of around 200 people at a time can be simultaneously analysed by Cognitron and, through a Bayesian process, which is an important statistical analysis technique, the AI progressively improves its brain teasers.

The aim of the study is to see if an AI machine can learn the major components of human intelligence using large amounts of data gathered from participants. Identifying these components remains one of the biggest challenges in the field of intelligence.

Cognitron will use the data that it gathers to build a more sophisticated intelligence test that will have a wide variety of applications, including analyses of gene-brain-cognitive associations, and clinical assessment of cognitively impaired patients.

Ultimately, the AI could benefit scientific research far more broadly. It may be able to erase subjectivity from research and lead to machines designing and running experiments on their own, which may improve the way fundamental science is conducted.

Dr Adam Hampshire, the study lead, from the Department of Medicine, said: “No one wants to eliminate the human element from science, but limitations and biases in how human scientists think are a major barrier for research. We believe that Cognitron may one day be able to enhance the research process for the scientific community.

"We are living in an age of big data and we need AI systems that can help us to process this huge avalanche of information. Imagine if a scientist could set a group of AIs to tackle multiple challenges they are trying to answer in their research field. It may speed up the painstaking trial and error process that fundamental research is based on and lead to unbiased findings and critical breakthroughs much quicker.”

Computational, Cognitive and Clinical Neuroimaging Laboratory

The work is being carried out by researchers at Imperial’s Computational, Cognitive and Clinical Neuroimaging Laboratory (C3NL), involving Dr Peter Hellyer, from the Department of Bioengineering, Dr Robert Leech, from the Department of Medicine, and Dr Romy Lorenz, a recent PhD graduate, now EPSRC postdoctoral fellow from the Centre for Neurotechnology. 

This story is adapted from a blog post by Roger Highfield, Director of External Affairs at the Science Museum.