Scientists have built an artificial intelligence system that has performed as well as in intelligence tests as a young child.
The ConceptNet 4 system has been put through a verbal IQ test, getting a score that is seen as “average for a four-year-old child”.
The AI is built using natural language processing tools, which allow to it to understand words in the same way that humans do. That was combined with some other software that allowed it to understand and then answer questions.
In the introduction to a paper on their results, the scientists write: “The ConceptNet system scored a WPPSI-III VIQ that is average for a four-year-old child, but below average for 5 to 7 year-olds. Large variations among subtests indicate potential areas of improvement.
“In particular, results were strongest for the Vocabulary and Similarities subtests, intermediate for the Information subtest, and lowest for the Comprehension and Word Reasoning subtests. Comprehension is the subtest most strongly associated with common sense.”
The authors caution against seeing the test as a demonstration of real intelligence. The large variations in the results mean that we should not see the findings as proof that “ConceptNet has the verbal abilities a four-year-old”, they say, but it does mean that children’s IQ tests can be one way of measuring and comparing how good artificial intelligence systems are.
Join our new commenting forum
Join thought-provoking conversations, follow other Independent readers and see their replies