Rethinking artificial intelligence



Posted by Trent Nouveau

In the 1800s, intelligence was typically associated with the ability to memorize facts and formulas. Today, intelligence is measured via IQ tests, with the average individual weighing in at about 100.



And a computer? Well, researchers at the Department of Philosophy, Linguistics and Theory of Science at the University of Gothenburg, Sweden, have coded a program capable of achieving an IQ score of 150.

Rethinking artificial intelligenceAs researcher Claes Strannegård notes, IQ tests are based on two types of problems: progressive matrices, which test the ability to perceive patterns in pictures, and numerical sequences, which measure the ability to detect patterns in numbers. 



Interestingly enough, the most common math computer programs score below 100 on IQ tests with number sequences. For Strannegård, this was reason enough to try and design "smarter" computer programs.

"We're trying to make [software] that can discover the same types of pattern humans can see," he explained.

Essentially, Strannegård and his research group hypothesized that number sequence problems were only partly a matter of mathematics – as psychology remains a critical factor.

"1, 2, and...? What comes next? Most people would say 3, but it could also be a repeating sequence like 1, 2, 1 or a doubling sequence like 1, 2, 4," said Strannegård.



"Neither of these alternatives is more mathematically correct than the others. What it comes down to is that most people have learned the 1-2-3 pattern."

As such, Strannegård's team integrated a psychological model or algorithm of human patterns in their computer program - complete with a mathematical paradigm that emulates human-like problem solving. It is now capable of acing the standardized tests, implying an IQ of at least 150.

"Our [software is] beating the conventional math programs because we are combining mathematics and psychology. Our method can potentially be used to identify patterns in any data with a psychological component, such as financial data. But it is not as good at finding patterns in more science-type data, such as weather data, since then the human psyche is not involved.

"[Nevertheless], we have developed a pretty good understanding of how the tests work. Now we want to divide them into different levels of difficulty and design new types of tests, which we can then use to design computer programs for people who want to practice their problem solving ability," Strannegård added.