Artificial Intelligence (AI) machines can already do several remarkable things: they are far better than humans at performing complex calculations, and they’re pretty good at playing chess . Researchers have once again tested the limits of AI by putting one of the world’s most intelligent AI machines through its paces with an IQ test, and the results are in: it has the same IQ as an average four-year-old child, as reported by MIT Technology Review .
Measuring intelligence through an IQ test is thought to be the best way to determine the intellectual capacity of people from a huge range of human cultures. A team of researchers, led by Stellan Ohlsson at the University of Illinois, decided to apply this concept to an intelligence outside of any normal human culture: an AI machine developed by the Massachusetts Institute of Technology (MIT).
The intelligent machine, dubbed ConceptNet 4, was given a verbal reasoning examination calibrated for four-year-old children. Known as the Wechsler Preschool and Primary Scale of Intelligence , it calculates a child’s IQ by asking a selection of questions from five categories.
The vocabulary category contains questions such as “What is a cat?”. The information category asks questions such as “Where can you find a tiger?”, and the word reasoning section asks the child to identify an object after being given three clues as to its identity. The comprehension category tests the child’s ability to understand the motivation behind actions, such as querying why people say hello or shake hands. Finally, the similarities category asks the child to understand the link between two objects, such as “Rain and snow are both made of _ ?”
After modifying ConceptNet 4’s programming to be able to deal with the questions it was going to be asked, the researchers gave it the same IQ test. The answers it gave were strongly linked to how it dealt with the language in the question, so more straightforward, concrete questions were handled well. Consequently, it did very well in the vocabulary and similarities segments, while doing averagely in the information question.
When concepts with inherent meaning or intent had to be handled, however, it dropped the ball. For example, when asked “why do people shake hands?” it interpreted the question as asking “what is the reason people’s hands shake?”. As a result, it decided that people shake hands because they are having an epileptic fit. As you can imagine, the AI scored poorly on the comprehension questions.
It also fared disastrously in the word reasoning category, giving truly bizarre answers unlike any child would ever use. When given the clues “This animal has a mane if it is male, it lives in Africa, and it is a yellowish-brown cat,” its five most common answers were “dog,” “cat,” “home,” “creature,” and “farm.”
As Ohlsson told MIT Technology Review , “if the clues say it is a cat, then types of cat are the only alternatives to be considered,” so this kind of misstep is currently inexplicable.
All categories considered, the AI’s measured verbal IQ was indeed that of an intellectually-average four-year-old child taking the same test. Stephen Hawking recently told BBC News he thought that artificial intelligence (AI) could lead to the extinction of humankind. Although this is entirely plausible, AI clearly has a long way to go to get to the point where it can stage a robot uprising.
